Search

Wetlands play a vital role in regulating the global carbon cycle, providing biodiversity, and reducing flood risks. These functions maintain ecological balance and ensure human well-being. Timely, accurate monitoring of wetlands is essential, not only for conservation efforts, but also for achieving Sustainable Development Goals (SDGs). In this study, we combined Sentinel-1/2 images, terrain data, and field observation data collected in 2020 to better understand wetland distribution. A total of 22 feature variables were extracted from multi-source data, including spectral bands, spectral indices (especially red edge indices), terrain features, and radar features. To avoid high correlations between variables and reduce data redundancy, we selected a subset of features based on recursive feature elimination (RFE) and Pearson correlation analysis methods. We adopted the random forest (RF) method to construct six wetland delineation schemes and incorporated multiple types of characteristic variables. These variables were based on remote sensing image pixels and objects. Combining red-edge features, terrain data, and radar data significantly improved the accuracy of land cover information extracted in low-mountain and hilly areas. Moreover, the accuracy of object-oriented schemes surpassed that of pixel-level methods when applied to wetland classification. Among the three pixel-based schemes, the addition of terrain and radar data increased the overall classification accuracy by 7.26%. In the object-based schemes, the inclusion of radar and terrain data improved classification accuracy by 4.34%. The object-based classification method achieved the best results for swamps, water bodies, and built-up land, with relative accuracies of 96.00%, 90.91%, and 96.67%, respectively. Even higher accuracies were observed in the pixel-based schemes for marshes, forests, and bare land, with relative accuracies of 98.67%, 97.53%, and 80.00%, respectively. This study's methodology can provide valuable reference information for wetland data extraction research and can be applied to a wide range of future research studies. [ABSTRACT FROM AUTHOR]

The Inner Niger Delta (IND), one of the largest floodplain systems in Africa, sustains the livelihoods of more than three million people and is a driver of the rural economy of Mali as far as agriculture, fish production, and livestock are concerned. Because the IND ecosystem and economy are flood-dependent, it is important to monitor seasonal flooding variations. Many attempts to accomplish this task have relied on detailed datasets, such as daily discharge, daily rainfall, and evapotranspiration, which are not easily accessible for data-sparse areas. Additionally, because the area is large, this remains a challenging task. In this study, the interannual variability of seasonal inundation in the IND was investigated by leveraging the computing power of the Google Earth Engine and its large catalogue of open datasets. The main objective was to analyse the temporal and spatial distributions of the inundation extent during the last 13 years. A collection of Landsat 5, 7, 8, and 9 images were composited and different bands were used with various water and vegetation indices in a pixel-based supervised classification to detect the flood extent between 2010 and 2022. A significant improvement in classification accuracy was observed thanks to the different indices. The results suggest a general increasing trend in the maximum annual inundation extent. Throughout the study period, the maximum inundated area varied between 15,209 km2 in autumn 2011 and 21,536 km2 in autumn 2022. The upstream water intake led to a decrease of about 6–10% of the inundated area. Similar fluctuations in the inundated area, precipitation, and river discharge were observed. The proposed approach demonstrates a great potential for monitoring annual inundation, especially for large areas such as the IND, where in situ measurements are sparse. [ABSTRACT FROM AUTHOR]

A series of earthquakes occurred in Kumamoto, Japan, in April 2016, which caused numerous landslides. In this study, high-resolution pre-event and post-event optical imagery, plus bi-temporal Synthetic Aperture Radar (SAR) data are paired with geospatial data to train a pixel-based machine learning classification algorithm using logistic regression to identify landslides occurred because of the Kumamoto earthquakes. The geospatial data used include a categorical variable (surficial geology), and six continuous variables including elevation, slope, aspect, curvature, annual precipitation, and landslide probability derived by the USGS preferred geospatial model which incorporates ground shaking in the input parameters. Grayscale index change and vegetation index change are also calculated from the optical imagery and used as input variables, in addition to temporal differences in HH (horizontally transmitted and horizontally received polarization) and HV (horizontally transmitted and vertically received polarization) amplitudes of SAR data. A detailed human-drawn landslide occurrence inventory was used as ground-truth for model development and testing. The selection of optimal features was done using a supervised feature ranking method based on the Receiver Operating Characteristic (ROC) curve. To weigh the benefit of combining different types of imagery, temporal change information and geospatial environmental indicators for landslide mapping after earthquakes, five different combinations of features were tested, and the results showed that adding data of selected geospatial parameters (landslide probability, slope, curvature, precipitation, and geology) plus selected change indices (grayscale index change, vegetation index change, and HV amplitude difference of SAR data) to the imagery (post event optical) lead to the highest classification accuracy of 86.5% on class-balanced independent testing data. A comparative analysis was conducted to evaluate the performance of the proposed method with five other commonly used machine learning classification methods, and the results have shown the superiority of the logistic regression method, followed by support vector machines. [ABSTRACT FROM AUTHOR]

Wetlands play a vital role in regulating the global carbon cycle, providing biodiversity, and reducing flood risks. These functions maintain ecological balance and ensure human well-being. Timely, accurate monitoring of wetlands is essential, not only for conservation efforts, but also for achieving Sustainable Development Goals (SDGs). In this study, we combined Sentinel-1/2 images, terrain data, and field observation data collected in 2020 to better understand wetland distribution. A total of 22 feature variables were extracted from multi-source data, including spectral bands, spectral indices (especially red edge indices), terrain features, and radar features. To avoid high correlations between variables and reduce data redundancy, we selected a subset of features based on recursive feature elimination (RFE) and Pearson correlation analysis methods. We adopted the random forest (RF) method to construct six wetland delineation schemes and incorporated multiple types of characteristic variables. These variables were based on remote sensing image pixels and objects. Combining red-edge features, terrain data, and radar data significantly improved the accuracy of land cover information extracted in low-mountain and hilly areas. Moreover, the accuracy of object-oriented schemes surpassed that of pixel-level methods when applied to wetland classification. Among the three pixel-based schemes, the addition of terrain and radar data increased the overall classification accuracy by 7.26%. In the object-based schemes, the inclusion of radar and terrain data improved classification accuracy by 4.34%. The object-based classification method achieved the best results for swamps, water bodies, and built-up land, with relative accuracies of 96.00%, 90.91%, and 96.67%, respectively. Even higher accuracies were observed in the pixel-based schemes for marshes, forests, and bare land, with relative accuracies of 98.67%, 97.53%, and 80.00%, respectively. This study’s methodology can provide valuable reference information for wetland data extraction research and can be applied to a wide range of future research studies.

ABSTRACTTimely and accurate wetland information is necessary for wetland resource management. Recent advances in machine learning and remote sensing have facilitated cost-effective monitoring of wetlands. However, reliable methods for fine-grained and rapid wetland mapping are still lacking. To address the issue, a wetland sample set with 20 categories for China was collected based on a sampling strategy that combines automatic sample generation and visual interpretation. Simultaneously, a novel multi-stage method for fine-grained wetland classification was proposed, which integrates pixel-based and object-based strategies using ensemble learning algorithms and multi-source remote sensing data. First, a pixel-based ensemble learning algorithm was implemented to classify five rough wetland categories and six non-wetland categories. Second, an object-based ensemble learning approach was designed to separate the water cover in the pixel-based classification results into eight detailed categories. Third, the merged pixel-based and object-based classification results were refined with knowledge-based post-processing procedures to identify 14 fine-grained wetland categories. Results using the Pixel Information Expert Engine (PIE-Engine) cloud platform proved the effectiveness of the proposed wetland classification method. The overall accuracy, kappa, and weighted F1 reached 87.39%, 82.80%, and 86.02%, respectively. The adopted ensemble learning algorithm yielded better performance than classifiers such as CatBoost, random forest, and XGBoost. The incorporation of spectral, texture, shape, topographic, and geographic features from multi-source data contributed to differentiating wetland categories. According to the relative contribution, spectral indexes (NDVI and NDWI), texture features (sum average and contrast), and topographic features (slope and elevation) were identified as important leading predictors for the first-stage pixel-based classification. Shape features (shape index and compactness) and auxiliary features (geographic location) were crucial predictors for the second-stage object-based classification. Compared with other products, our 10-m wetland mapping results for national wetland reserves were rich in detail and fine in categories. Overall, the constructed sample set and developed classification method show promise in laying a foundation for large-scale wetland mapping. The derived wetland maps can provide support for wetland protection and restoration.

Land use land cover (LULC) classification using remote sensing images is a valuable resource in various fields such as climate change, urban development, and land degradation monitoring. The city of Madurai in India is known for its diverse geographical elements and rich heritage, which includes the cultural sport of “Jallikattu”: whose main competitor, the zebusare deeply affected by the conversion of their waterbodies and pastures into concrete jungles. Hence, monitoring land degradation is vital in preserving the geography and cultural heritage of the study area, Madurai. The “Landsat 8 Operational Land Imager tier_2 collection_2 Level_2 Surface Reflectance” image was taken for this study. The LULC classification is performed based on the following classes: forest, agriculture, urban, water bodies, uncultivated land, and bare land. The objective of the study is to incorporate auxiliary features to spectral and textural features along with a simple non-iterative clustering (SNIC) segmentation algorithm and implement a boundary-specific two-level learning approach based on support vector machines (SVM) and k nearest neighbors (kNN) classification algorithms. The overall accuracy (OA) of 95.78% and 0.94 Kappa score (K) were obtained using a boundary-specific two-level model augmented with auxiliary feature and SNIC algorithm in comparison to PB, OB, and OBS, which achieve OA (K) of 81% (0.76), 91% (0.89), and 94.42% (0.92), respectively. The results demonstrate a notable enhancement in overall classification accuracy when augmenting the features and refining classification decisions using a boundary-specific two-level learning approach. [ABSTRACT FROM AUTHOR]

This article describes a comparison of the pixel-based classification methods used to distinguish ice from other land cover types. The article focuses on processing RGB imagery, as these are very easy to obtained. The imagery was taken using UAVs and has a very high spatial resolution. Classical classification methods (ISODATA and Maximum Likelihood) and more modern approaches (support vector machines, random forests, deep learning) have been compared for image data classifications. Input datasets were created from two distinct areas: The Pond Skříň and the Baroch Nature Reserve. The images were classified into two classes: ice and all other land cover types. The accuracy of each classification was verified using a Cohen's Kappa coefficient, with reference values obtained via manual surface identification. Deep learning and Maximum Likelihood were the best classifiers, with a classification accuracy of over 92% in the first area of interest. On average, the support vector machine was the best classifier for both areas of interest. A comparison of the selected methods, which were applied to highly detailed RGB images obtained with UAVs, demonstrates the potential of their utilization compared to imagery obtained using satellites or aerial technologies for remote sensing. [ABSTRACT FROM AUTHOR]

IntroductionOver the past 60 years, the use of plastic covers as a tool to increase the harvest and increase the yield of horticultural crops has steadily increased worldwide. Plastic greenhouses have turned desert areas into areas with modern agricultural development, the province of Almera in the south of Spain is a good example. It now has many greenhouses, making it a global model of agricultural development; but greenhouses contain large quantities of phthalates and cause harm to people, like hormonal disturbances, heart problems, cancer, etc. However, plastic greenhouses are widely built to produce vegetables and fruits near cities. As a result, several remote sensing methods have been developed to identify and monitor the distribution of plastic covered greenhouses in order to manage water resources, identify sites and quantities of plastic greenhouses. Remote detection is the only practical method to monitor plastic greenhouses in a vast geographic area. In the past few years, there have been few studies using high spatial resolution images. In one study, three main absorption ranges were identified that are unaffected by dust, washing and surface factors. In Spain, an artificial intelligence neural network has been proposed to identify greenhouse using Quick Bird images with a resolution of 1.5m. Studies based on medium spatial resolution imagery were also conducted, resulting in different results. In search on land cover classification using Landsat TM images, no favorable results were found. Research has proposed a new method for mapping greenhouses using Sentinel-2 dual-time images and 1D-CNN deep learning. The aim of the current research is to identify the length of the cultivation period of greenhouse crops using Sentinel-2 satellite images. Researchers used a variety of satellite imagery to identify and classify greenhouses. The innovative objective of the current research is to evaluate the Sentinel-2 satellite images in determining the length of the cultivation period of greenhouse crops; which was done for the first time within the country. Material and MethodsIn the first step, the cultivated area of greenhouses was identified using an aerial photograph. Then, useful bands were extracted in greenhouse studies using Sentinel-2 time series images. Next, vegetation and plastic cover indices were calculated for the greenhouse growing period. 1) Identify and determine the area under cultivation in greenhouses: By comparing the pixel-based classification algorithms and the object-oriented classification algorithms, the cultivated area of greenhouses was identified. ENVI and eCognition software were used for pixel-based classification and object-oriented classification, respectively. It should be noted that only one aerial photograph with three bands, blue, green and red, has been used in object-oriented classification. In order to classify the base pixel, learning samples were selected using expert knowledge from the study area. In some instances, Google Earth was used as well. Learning samples were selected scattered across the image to improve accuracy. For validation of the maps obtained from these algorithms, ground control points were used. To reduce human error, these points were also entered into Google Earth. 2) Identifying functional bands in the greenhouse study: The spectral behavior curve of different earth surface covers in Blue, Green, Red, SWIR, NIR bands was drawn in an image of the Sentinel-2 satellite. These spectral signatures were compared on May 20, 2020 i.e., when outdoor vegetation and greenhouse cultivation were at their peak. 3) Determining the duration of the greenhouse growing season: Vegetation indices of the NDVI, SAVI, OSAVI, MSAVI, GNDVI, RVI, DVI, RGVI, IPVI and plastic cover indices PGI, RPGI, PI, PMLI were compared using Sentinel-2 satellite time series images. Results and DiscussionGreenhouse cultivation is used to improve quality and increase food production. However, their development has many negative effects on the environment. Therefore, obtaining accurate and timely information on the distribution of greenhouses and the time of planting and harvesting crops under plastic covers can make a significant contribution to agricultural management, water management and soil protection. The present study is the first study to identify the length of the greenhouse cultivation period using Sentinel-2 satellite time series images. The results showed that among the object-oriented classification algorithms, two classification algorithms, Bayes and KNN, were more precise for the identification and determination of the cultivated area of greenhouses. The reflection of the plant below the plastic cover of the greenhouse in the NIR, Narrow NIR, Red Edge and SWIR bands increases by an equal amount in comparison with the reflection from the vegetation in the open space. Comparing vegetation cover and greenhouse cover indices showed that the indices designed based on SWIR and Red bands showed greater reflection during the hot season of the year. Indices based on NIR, Narrow NIR, and Red Edge bands were more reflective from October to late February. Based on the results obtained from the ground truth data, greenhouses are plastered during the warm season. This caused an increase in the reflectance in indices designed based on SWIR and Red bands and a decrease in reflectance in indices designed based on NIR, Narrow NIR, Red Edge bands, during the hot season. Therefore, combining all indices, two crop periods were observed: the first started in early March, peak cover was reached in April, and harvest continued. Until the middle of August. The second harvest started at the end of August and peaked in December, until the end of harvest in mid-February.

The Inner Niger Delta (IND), one of the largest floodplain systems in Africa, sustains the livelihoods of more than three million people and is a driver of the rural economy of Mali as far as agriculture, fish production, and livestock are concerned. Because the IND ecosystem and economy are flood-dependent, it is important to monitor seasonal flooding variations. Many attempts to accomplish this task have relied on detailed datasets, such as daily discharge, daily rainfall, and evapotranspiration, which are not easily accessible for data-sparse areas. Additionally, because the area is large, this remains a challenging task. In this study, the interannual variability of seasonal inundation in the IND was investigated by leveraging the computing power of the Google Earth Engine and its large catalogue of open datasets. The main objective was to analyse the temporal and spatial distributions of the inundation extent during the last 13 years. A collection of Landsat 5, 7, 8, and 9 images were composited and different bands were used with various water and vegetation indices in a pixel-based supervised classification to detect the flood extent between 2010 and 2022. A significant improvement in classification accuracy was observed thanks to the different indices. The results suggest a general increasing trend in the maximum annual inundation extent. Throughout the study period, the maximum inundated area varied between 15,209 km2 in autumn 2011 and 21,536 km2 in autumn 2022. The upstream water intake led to a decrease of about 6–10% of the inundated area. Similar fluctuations in the inundated area, precipitation, and river discharge were observed. The proposed approach demonstrates a great potential for monitoring annual inundation, especially for large areas such as the IND, where in situ measurements are sparse.

In recent times, the development of innovative processes permits the application of a circular economy approach to the management and exploitation of mining waste with respect to human health and environment, such that society is changing its fundamentally negative perception of the mining sector. This study presents the opportunities and challenges of supplying raw materials from waste using a remote sensing technique, mycorrhizal-assisted phytoremediation, and hydrometallurgical techniques to transform mining waste from a problem to a resource. Soil/mine wastes from the Sierra Pintada mine (Mendoza, Argentina) were mineralogically and chemically analyzed, and then, a mapping of the mining waste was carried out by Sentinel-2A images to identify areas with similar characteristics. The bioaccumulation of HMs by autochthonous shrubs was also determined to select accumulator plant species, and to evaluate their potential for phytoremediation of mine soils at different technological scales, when they were inoculated with arbuscular mycorrhizal fungi originated from a mining-impacted area. RMs were recovered from plant biomass by scaling in bioreactors, the depuration module, and hydrometallurgical techniques. The encouraging results highlight that this multidisciplinary approach can be applied to meet the increasing demand for RMs supply and, at the same time, to protect the environment and public health. [ABSTRACT FROM AUTHOR]

Pixel and texture based classification using a well-defined and efficient architecture is considered as a major challenge. Nowadays, a large number of satellite images are received within a fraction of seconds, however processing such images to identify the land cover and land use is considered as a tedious process. To achieve this objective with high accuracy, an algorithm of cellular automata (ACA) is introduced in this proposed approach. The pixel-based classification is carried out with parallelepiped and maximum likelihood classifier, whereas the texture-based classification is accomplished using Softmax regression (SR) classifier. By incorporating ACA, the accuracy of these classification techniques is improved and the performance is then evaluated. This overall classification process is performed to understand the land cover and land use of Kerala. The classification accuracy attained using ACA-based parallelepiped, maximum likelihood and SR is found higher than classical parallelepiped, maximum likelihood, and SR algorithms. The final result reveals that the texture-based ACA classification provides a higher classification accuracy rate (96.8%) than the pixel-based ACA classification (90.98%). [ABSTRACT FROM AUTHOR]

• The proposed MCTNet integrates the advantages of CNN and Transformer for crop mapping. • MCTNet has a fewer number of parameters and better performance compared to other advanced models. • Features extracted by the CNN and Transformer sub-modules of MCTNet are complementary. • ALPE module makes the proposed model robust to missing values in time-series images. Deep learning approaches have provided state-of-the-art performance in crop mapping. Recently, several studies have combined the strengths of two dominant deep learning architectures, Convolutional Neural Networks (CNNs) and Transformers, to classify crops using remote sensing images. Despite their success, many of these models utilize patch-based methods that require extensive data labeling, as each sample contains multiple pixels with corresponding labels. This leads to higher costs in data preparation and processing. Moreover, previous methods rarely considered the impact of missing values caused by clouds and no-observations in remote sensing data. Therefore, this study proposes a lightweight multi-stage CNN-Transformer network (MCTNet) for pixel-based crop mapping using time-series Sentinel-2 imagery. MCTNet consists of several successive modules, each containing a CNN sub-module and a Transformer sub-module to extract important features from the images, respectively. An attention-based learnable positional encoding (ALPE) module is designed in the Transformer sub-module to capture the complex temporal relations in the time-series data with different missing rates. Arkansas and California in the U.S. are selected to evaluate the model. Experimental results show that the MCTNet has a lightweight advantage with the fewest parameters and memory usage while achieving the superior performance compared to eight advanced models. Specifically, MCTNet obtained an overall accuracy (OA) of 0.968, a kappa coefficient (Kappa) of 0.951, and a macro-averaged F1 score (F1) of 0.933 in Arkansas, and an OA of 0.852, a Kappa of 0.806, and an F1 score of 0.829 in California. The results highlight the importance of each component of the model, particularly the ALPE module, which enhanced the Kappa of MCTNet by 4.2% in Arkansas and improved the model's robustness to missing values in remote sensing data. Additionally, visualization results demonstrated that the features extracted from CNN and Transformer sub-modules are complementary, explaining the effectiveness of the MCTNet. [ABSTRACT FROM AUTHOR]

Nowadays, remote sensing image analysis is needed in various important tasks such as city planning, land-use classification, agriculture monitoring, military surveillance, and many other applications. In this context, hyperspectral images can play a useful role, but require specific handling. This paper presents a convolutional neural network based on one-dimensional support vector machine (SVM) convolution operations (1D-CSVM) for the analysis of hyperspectral images. SVM-based CNN (CSVM) was introduced first for the classification of high spatial resolution RGB images. It relies on linear SVMs to create filter banks in the convolution layers. In this work, the network is modified to cope with one-dimensional hyperspectral signatures and perform pixel-based classification. It thus analyzes each pixel spectrum independently from the pixel spatial neighborhood. Experiments were carried out on four benchmark hyperspectral datasets, Salinas-A, Kennedy Space Center (KSC), Indian Pines (IP) and Pavia University (Pavia-U). Compared to state-of-the-art models, the proposed network produces promising results for all tested datasets, with an accuracy up to 99.76%.

Riparian forests in Iran are valuable ecosystems which have many ecological values. Because of destruction of these forests in recent years, providing spatio-temporal information from area and distribution of these ecosystems has been receiving much attention. This study was performed for mapping distribution, area and density of riparian forests in southern Iran using Sentinel-2A, Google Earth, and field data. First Sentinel-2A satellite image of the study area was provided. The field work was performed to take the training areas and to assess the forest density of riparian forests in Khuzestan province. In the first part of this study, after selecting training areas as pixel-based samples on the Sentinel-2A satellite image, supervised classification of image was performed using support vector machine (SVM) algorithm to classify the distribution of riparian forests. After classification of Sentinel-2A satellite image, the boundary of riparian forests map was checked and corrected on Google Earth images. In the second part of this study, field data, Normalized Difference Vegetation Index (NDVI), and regression model were used to assess the density of riparian forests. Finally, the accuracy of the final riparian forest map (showing both distribution and density of riparian forests) was assessed using Google Earth images. Results showed that the final riparian forest map (showing both distribution and density of riparian forests) with overall accuracy 89% and kappa index 0.81 had a good accuracy for classifying the distribution and density of riparian forests in Khuzestan province. These results demonstrate the accuracy of SVM algorithm for classifying the distribution of riparian forests and also capability of NDVI for classifying the density of riparian forests in this study. Results also showed that regression model (R2 = 0.97) is reliable for estimating riparian forest density. The results demonstrated that there are 68447.18 ha of riparian forest around the main rivers in Khuzestan province, mainly distributed in the northwest and southeast of the province. From this area, 54694.15 ha have been covered by dense forests and 13753.03 ha by sparse forests. Results of this research have created the useful data of area, distribution and density of riparian forests in 10-m spatial resolution which is necessary for conservation and management of these forests in southern Iran. It is suggested that mapping area, distribution and density of these forests would be performed using SVM algorithm and NDVI in the certain temporal periods for protective management of these ecosystems in time series. [ABSTRACT FROM AUTHOR]

Salt marshes are vulnerable to sea-level rise, sediment deficits, and storm impacts. To remain vertically resilient, salt marshes must accrete sediment at rates greater or equal to sea-level rise. Ice-rafted debris (IRD), sediment that has been moved and deposited from ice sheets, is one of many processes that contribute to salt marsh sediment accretion in northern latitudes. On 4 January 2018, a winter storm caused major ice mobilization in the Plum Island Estuary (PIE), Massachusetts, USA, which led to large deposits of ice-rafted sediment. We aimed to quantify the volume and mass of deposited sediment, and evaluate the significance of IRD to sediment supply in Plum Island using pixel-based land-cover classification of aerial imagery collected by an Unmanned Aircraft System (UAS) and a Digital Elevation Model. Field measurements of patch thickness, and the area of IRD determined from the classification were used to estimate annual sediment accretion from IRD. Results show that IRD deposits are localized in three areas, and estimates show that IRD contributes an annual sediment accretion rate of 0.57 ± 0.14 mm/y to the study site. New England salt marsh accretion rates typically vary between 2–10 mm/y, and the average PIE sediment accretion rate is 2.5–2.7 mm/y. Therefore, this event contributed on average 20% of the annual volume of material accreted by salt marshes, although locally the deposit thickness was 8–14 times the annual accretion rate. We show that pixel-based classification can be a useful tool for identifying sediment deposits from remote sensing. Additionally, we suggest that IRD has the potential to bring a significant supply of sediment to salt marshes in northern latitudes and contribute to sediment accretion. As remotely sensed aerial imagery from UASs becomes more readily available, this method can be used to efficiently identify and quantify deposited sediment. [ABSTRACT FROM AUTHOR]

This study focuses on the comparison of machine learning algorithms for classifying the mangrove species of Malad creek, Mumbai, India using pixel-based and object-based approaches. As mangrove forests are complex in structure and require high resolution remotely sensed images for classifying them at the species level, traditional classification methods of remote sensing may not be suitable. In order to find an appropriate method, pixel-based random forest and K Nearest Neighbor classifiers and object-based random forest and K Nearest Neighbor classifiers were tested and compared using WorldView – 2 images. In addition, the best resolution RGB images from Google Earth were classified using random forest and K Nearest Neighbor algorithms in object-based approach and compared. The classification results showed that both object-oriented and pixel-based methods could identify the major mangrove species at the community level. However, the performance of object-based random forest classifier was better than other classification approaches in both WorldView-2 and Google Earth images with overall accuracies of 92.53% and 80.72% and kappa coefficients of 0.89 and 0.73. The overall results showed the potential of random forest classifier in object-based method for classifying the mangroves at the species level. [ABSTRACT FROM AUTHOR]

Crop supply and management is a global issue, particularly in the context of global climate change and rising urbanization. Accurate mapping and monitoring of specific crop types are crucial for crop studies. In this study, we proposed: (1) a methodology to map two main winter crops (winter wheat and winter barley) in the northern region of Finistère with high-resolution Sentinel-2 data. Different classification approaches (the hierarchical classification and the classical direct extraction), and classification methods (pixel-based classification (PBC) and object-based classification (OBC)) were performed and evaluated. Subsequently, (2) a further study that involved monitoring the phenology of the winter crops was carried out, based on the previous results. The aim is to understand the temporal behavior from sowing to harvesting, identifying three important phenological statuses (germination, heading, and ripening, including harvesting). Due to the high frequency of precipitation in our study area, crop phenology monitoring was performed using Sentinel-1 C-band SAR backscatter time series data using the Google Earth Engine (GEE) platform. The results of the classification showed that the hierarchical classification achieved a better accuracy when it is compared to the direct extraction, with an overall accuracy of 0.932 and a kappa coefficient of 0.888. Moreover, in the hierarchical classification process, OBC reached a better accuracy in cropland mapping, and PBC was proven more suitable for winter crop extraction. Additionally, in the time series backscatter coefficient of winter wheat, the germination and ripening (harvesting) phases can be identified at VV and VH/VV polarizations, and heading can be identified in both VV and VH polarizations. Secondly, we were able to detect the germination phase of winter barley in VV and VH, ripening with both polarizations and VH/VV, and finally, heading in VV and VH polarizations. [ABSTRACT FROM AUTHOR]

Globally, native semiarid grasslands and savannas have experienced a densification of woody plant species—leading to a multitude of environmental, economic, and cultural changes. These encroached areas are unique in that the diversity of tree species is small, but at the same time the individual species possess diverse phenological responses. The overall goal of this study was to evaluate the ability of very high resolution drone imagery to accurately map species of woody plants encroaching on semiarid grasslands. For a site in the Edwards Plateau ecoregion of central Texas, we used affordable, very high resolution drone imagery to which we applied maximum likelihood (ML), support vector machine (SVM), random forest (RF), and VGG-19 convolutional neural network (CNN) algorithms in combination with pixel-based (with and without post-processing) and object-based (small and large) classification methods. Based on test sample data (n = 1000) the VGG-19 CNN model achieved the highest overall accuracy (96.9%). SVM came in second with an average classification accuracy of 91.2% across all methods, followed by RF (89.7%) and ML (86.8%). Overall, our findings show that RGB drone sensors are indeed capable of providing highly accurate classifications of woody plant species in semiarid landscapes—comparable to and even greater in some regards to those achieved by aerial and drone imagery using hyperspectral sensors in more diverse landscapes. [ABSTRACT FROM AUTHOR]

Convolutional Neural Networks (CNN) have become the core of modern machine learning approaches. In addition to its inspiring interior design idea, the success of CNN depends mainly on two factors, the first is the availability of training data and the second is the computing power of the used devices. In the field of remote sensing, data availability is difficult and expensive. Furthermore, processing large remote sensing data to accommodate different models is a laborious process. At the same time, training data is often collected in the form of points distributed over crop fields rather than regions, which results in the scarcity of training data. To specifically address the scarcity of training points, in this paper we present a sparse pixel-based training of U-Net convolutional neural networks for land cover classification. Training images are reconstructed from the points’ collection in a random manner, they are used as an input for the convolution networks. Based on this proposed method, the amount of training data is reproduced from the different spectral signals for each land cover. We conducted extensive experiments on eight classes, using ground truth data collected from several locations in Fayoum Governorate, Egypt. The obtained results showed the superiority of the proposed method over other methods.

This paper presents a thorough evaluation of twenty-one state-of-the-art widely-used crop segmentation algorithms, motived by their significance in vision tasks for further analysis. An ideal crop segmentation algorithm can effectively extract crop information, thus providing an important precondition for the application of intelligent agriculture analytics. In order to enable researchers in this field to fully understand various crop segmentation methods, this paper proposes a new classification strategy of object segmentation by dividing the algorithms into pixel-based and region-based approaches at first, and then systematically evaluating various crop segmentation methods with a unified data benchmark and four common metrics. A new dataset which incorporates crop variety, environment condition and observation distance into consideration is constructed for demonstrating the experiments and comparisons. The effectiveness and robustness of these algorithms were evaluated by three sets of comparative experiments. Based on the quantitative results, we summarize the advantages and disadvantages of the evaluated algorithms from the segmentation performances with four metric indicators. Furthermore, the discussion and evaluation results will provide great support for precision agriculture analysis.

Water body extraction plays a key role in water resource management and development which is one of the key issues in research topics related to remote sensing in recent decades. Pixel-based; supervised and unsupervised classification methods and water indices have been developed for extracting water body from satellite imagery while for some cases they are unable to differentiate water body from low albedo features with different spectral characteristics. The object of this paper is to evaluate the efficiency of object-based, pixel-based and water indices for water body extraction. Landsat-8 imagery used for this study which has 30m resolution in Visible, NIR (Near-Infrared), SWIR (Shortwave Infrared) spectrum and panchromatic band with 15m resolution. The Edge-Based Segmentation (EBS) algorithm and the Support Vector Machine (SVM) classification method has been used for object-based approach, on the other hand, Maximum Likelihood and K-means methods has been selected for pixel-based classification. As a water index, Normalized Difference Water Index (NDWI) has been selected to extract water body from satellite imagery. The study area selected in both urban and mountainous regions, with different characteristics, which located in Turkey. Results show that object-based water body extraction is more accurate than the other methods tested.

Sentinel-2 MSI is one of the core missions of the Copernicus Earth Observation programme of the European Union. This mission shows great potential to map the regional high-resolution spatio-temporal dynamics of land use and land cover. In tropical regions, despite the high revisiting time of 5 days including both Sentinel-2A and 2B satellites, the frequent presence of clouds, cloud-shadows, haze and other atmospheric contaminants are precluding the visibility of the Earth surface up to several months. In this paper we present four annual pan-tropical cloud-free composites computed and exported from Google Earth Engine (GEE) by making use of available Sentinel-2 L1C collection for the period spanning from 2015 to 2020. We furthermore propose empirical approaches to reduce the BRDF effect over tropical forest areas by showing pros and cons of image-based versus swath-based methodologies. Additionally, we provide a dedicated web-platform offering a fast and intuitive way to browse and explore the proposed annual composites as well as layers of potential annual changes as a ready-to-use means to visually identify and verify degradation and deforestation activities as well as other land cover changes.

Pixel-based classification is a complex but well-known process widely used for satellite imagery classification. This paper presents a supervised multi-classifier pipeline that combined multiple Earth Observation (EO) data and different classification approaches to improve specific land cover type identification. The multi-classifier pipeline was tested and applied within the SCO-Live project that aims to use olive tree phenological evolution as a bio-indicator to monitor climate change. To detect and monitor olive trees, we classify satellite images to precisely locate the various olive groves. For that first step we designed a multi-classifier pipeline by the concatenation of a first classifier which uses a temporal Random-Forest model, providing an overall classification, and a second classifier which uses the result from the first classification. IOTA2 process was used in the first classifier, and we compared Multi-layer Perceptron (MLP) and One-class Support Vector Machine (OCSVM) for the second. The multi-classifier pipelines managed to reduce the false positive (FP) rate by approximately 40% using the combination RF/MLP while the RF/OCSVM combination lowered the FP rate by around 13%. Both approaches slightly raised the true positive rate reaching 83.5% and 87.1% for RF/MLP and RF/OCSVM, respectively. The overall results indicated that the combination of two classifiers pipeline improves the performance on detecting the olive groves compared to pipeline using only one classifier.

Understanding the composition and the changes of New Zealand’s woody vegetation communities is important for effective management. However, past national-scale mapped classifications emphasised mature rather than seral vegetation communities and forests were mapped in relative coarse spatial resolution. The integration of Sentinel-2 and PlanetScope imagery provides an opportunity for forest mapping with low cost and high accuracy.This study aims to investigate the feasibility of the integrated image for detailed forest mapping. Free satellite data (Sentinel-2, PlanetScope, fused data) were compared with commercial data (WorldView-2, and WorldView-2 resampled to Sentinel-2 and PlanetScope spatial resolutions) by conducting pixel-based classification with three machine learning classifiers (Support Vector Machine radial basis function kernel, Random Forest, Artificial Neural Network). The combinations of imagery type and classifier were assessed on their potential for mapping nine land cover classes in podocarp forest in New Zealand’s central north island, including: conifer, low layer vegetation, broadleaf evergreen, highland softwood, wetland vegetation, water, dead tree, lowland softwood, and low-density vegetation and bare soil. Spectral features (single bands and indices), textural features, and an 8 m resolution digital terrain model (DTM) were used in classifications; the relative importance of these input features was also assessed.In this study, it was found that the overall classification accuracy was dependent on the combination of classifier and imagery, with different combinations resulting in a range of accuracies between 0.669 and 0.956. The best overall accuracy was achieved by integrating Sentinel-2 and PlanetScope imagery (0.956) which was even greater than that of WorldView-2 (0.951). The digital terrain model was the most important feature for all scenarios; Gray-Level Co-Occurrence Matrix-Mean was the most important texture variable for WorldView-2 and integrated images. Original bands, as well as GI, Norm-G, and SR-NIRR, were also crucial for vegetation classification.

Salt marshes are vulnerable to sea-level rise, sediment deficits, and storm impacts. To remain vertically resilient, salt marshes must accrete sediment at rates greater or equal to sea-level rise. Ice-rafted debris (IRD), sediment that has been moved and deposited from ice sheets, is one of many processes that contribute to salt marsh sediment accretion in northern latitudes. On 4 January 2018, a winter storm caused major ice mobilization in the Plum Island Estuary (PIE), Massachusetts, USA, which led to large deposits of ice-rafted sediment. We aimed to quantify the volume and mass of deposited sediment, and evaluate the significance of IRD to sediment supply in Plum Island using pixel-based land-cover classification of aerial imagery collected by an Unmanned Aircraft System (UAS) and a Digital Elevation Model. Field measurements of patch thickness, and the area of IRD determined from the classification were used to estimate annual sediment accretion from IRD. Results show that IRD deposits are localized in three areas, and estimates show that IRD contributes an annual sediment accretion rate of 0.57 ± 0.14 mm/y to the study site. New England salt marsh accretion rates typically vary between 2–10 mm/y, and the average PIE sediment accretion rate is 2.5–2.7 mm/y. Therefore, this event contributed on average 20% of the annual volume of material accreted by salt marshes, although locally the deposit thickness was 8–14 times the annual accretion rate. We show that pixel-based classification can be a useful tool for identifying sediment deposits from remote sensing. Additionally, we suggest that IRD has the potential to bring a significant supply of sediment to salt marshes in northern latitudes and contribute to sediment accretion. As remotely sensed aerial imagery from UASs becomes more readily available, this method can be used to efficiently identify and quantify deposited sediment.

Crop supply and management is a global issue, particularly in the context of global climate change and rising urbanization. Accurate mapping and monitoring of specific crop types are crucial for crop studies. In this study, we proposed: (1) a methodology to map two main winter crops (winter wheat and winter barley) in the northern region of Finistère with high-resolution Sentinel-2 data. Different classification approaches (the hierarchical classification and the classical direct extraction), and classification methods (pixel-based classification (PBC) and object-based classification (OBC)) were performed and evaluated. Subsequently, (2) a further study that involved monitoring the phenology of the winter crops was carried out, based on the previous results. The aim is to understand the temporal behavior from sowing to harvesting, identifying three important phenological statuses (germination, heading, and ripening, including harvesting). Due to the high frequency of precipitation in our study area, crop phenology monitoring was performed using Sentinel-1 C-band SAR backscatter time series data using the Google Earth Engine (GEE) platform. The results of the classification showed that the hierarchical classification achieved a better accuracy when it is compared to the direct extraction, with an overall accuracy of 0.932 and a kappa coefficient of 0.888. Moreover, in the hierarchical classification process, OBC reached a better accuracy in cropland mapping, and PBC was proven more suitable for winter crop extraction. Additionally, in the time series backscatter coefficient of winter wheat, the germination and ripening (harvesting) phases can be identified at VV and VH/VV polarizations, and heading can be identified in both VV and VH polarizations. Secondly, we were able to detect the germination phase of winter barley in VV and VH, ripening with both polarizations and VH/VV, and finally, heading in VV and VH polarizations.

Important factors in planning and urban management, especially in order to achieve sustainable development and optimal use of land in urban areas is the availability of timely and accurate information about the status of land use in urban areas. In this dirction, the aim of present study is to compare two methods of pixel-based and object-oriented classification in land use mapping with using ETM+ image in Ilam city, Ilam province using Idrisi Selva software. After the supply of related image and implementing geometric and radiometric corrections on image, we applied two classification methods to land use mapping. To assess the accuracy of classification methods, we used of indices of overall accuracy, kappa coefficient, producer accuracy and user accuracy. The results show that the object-oriented classification method has more resolution than the pixel-based classification method. The results of accuracy show that method of object-oriented in two indices of overall accuracy and kappa coefficient with (respectively) 94 percent and 91 percent was more accurate then pixel-based classification method with overall and Kappa coefficient 88 percent and 86 percent. The result of this study suggests that from object-oriented classification method use to production of land use map.

While forest roads are important to forest managers in terms of facilitating the exploitation of wood and timber, their role is far more multifunctional. They permit access to emergency services in the case of forest fires as well as acting as fire breaks, enhance biodiversity, and provide access to the public to enjoy recreational activities. Detailed maps of forest roads are an essential tool for better and more timely forest management and automatic/semi-automatic tools allow not only the creation of forest road databases, but also enable these to be updated. In Spain, LiDAR data for the entire national territory is freely available, and the capture of higher density data is planned in the next few years. As such, the development of a forest road detection methodology based on LiDAR data would allow maps of all forest roads to be developed and regularly updated. The general objective of this work was to establish a low density LiDAR data-based methodology for the semi-automatic detection of the centerline of forest roads on steep terrain with various types of canopy cover. Intensity and slope images were generated using the currently available LiDAR data of the study area (0.5 points m-2). Two image classification approaches were evaluated: pixel-based and object-oriented classification (OBIA). The LiDAR-derived centerlines obtained with the two approaches were compared with the real centerlines which had previously been digitized in the field. The road width, type of surface and type of vegetation cover were also recorded. The effectiveness of the two approaches was evaluated through three quality indicators: correctness, completeness and quality. In addition, the accuracy of the LiDAR-derived centerlines was also evaluated by combining GIS analysis and statistical methods. The pixel-based approach obtained higher values than OBIA for two of the three quality measures (correctness: 93% compared to 90%; and quality: 60% compared to 56%) as well as in terms of positional accuracy (± 5.5 m vs. ± 6.8 for OBIA). The results obtained in this study demonstrate that producing road maps is among the most valuable and easily attainable products of LiDAR data analysis.

Globally, native semiarid grasslands and savannas have experienced a densification of woody plant species—leading to a multitude of environmental, economic, and cultural changes. These encroached areas are unique in that the diversity of tree species is small, but at the same time the individual species possess diverse phenological responses. The overall goal of this study was to evaluate the ability of very high resolution drone imagery to accurately map species of woody plants encroaching on semiarid grasslands. For a site in the Edwards Plateau ecoregion of central Texas, we used affordable, very high resolution drone imagery to which we applied maximum likelihood (ML), support vector machine (SVM), random forest (RF), and VGG-19 convolutional neural network (CNN) algorithms in combination with pixel-based (with and without post-processing) and object-based (small and large) classification methods. Based on test sample data (n = 1000) the VGG-19 CNN model achieved the highest overall accuracy (96.9%). SVM came in second with an average classification accuracy of 91.2% across all methods, followed by RF (89.7%) and ML (86.8%). Overall, our findings show that RGB drone sensors are indeed capable of providing highly accurate classifications of woody plant species in semiarid landscapes—comparable to and even greater in some regards to those achieved by aerial and drone imagery using hyperspectral sensors in more diverse landscapes.

Accurate maps of the spatial distribution of tropical tree species provide valuable insights for ecologists and forest management. The discrimination of tree species for economic, ecological, and technical reasons is usually necessary for achieving promising results in tree species mapping. Most of the data used in tree species mapping normally have some degree of imbalance. This study aimed to assess the effects of imbalanced data in identifying and mapping trees species under threat in a selectively logged sub-montane heterogeneous tropical forest using random forest (RF) and support vector machine with radial basis function (RBF-SVM) kernel classifiers and WorldView-2 multispectral imagery. For comparison purposes, the original imbalanced dataset was standardized using three data sampling techniques: oversampling, undersampling, and combined oversampling and undersampling techniques in R. The combined oversampling and undersampling technique produced the best results: F1-scores of 68.56 ± 2.6% for RF and 64.64 ± 3.4% for SVM. The balanced dataset recorded improved classification accuracy compared to the original imbalanced dataset. This research observed that more separable classes recorded higher F1-scores. Among the species, Syzygium guineense and Zanthoxylum gilletii were the most accurately mapped whereas Newtonia buchananii was the least accurately mapped. The most important spectral bands with the ability to detect and distinguish between tree species as measured by random forest classifier, were the Red, Red Edge, Near Infrared 1, and Near Infrared 2.

Instance and variable selection involve identifying a subset of instances and variables such that the learning process will use only this subset with better performances and lower cost. Due to the huge amount of data available in many fields, data reduction is considered as an NP-hard problem. In this paper, we present a simultaneous instance and variable selection approach based on the Random Forest-RI ensemble methods in the aim to discard noisy and useless information from the original data set. We proposed a selection principle based on two concepts: the ensemble margin and the importance variable measure of Random Forest-RI. Experiments were conducted on cytological images for the automatic segmentation and recognition of white blood cells WBC (nucleus and cytoplasm). Moreover, in order to explore the performance of our proposed approach, experiments were carried out on standardized datasets from UCI and ASU repository, and the obtained results of the instances and variable selection by the Random Forest classifier are very encouraging. [ABSTRACT FROM AUTHOR]

Nuclear power plants accidents should be seen as very complex issues, once they encompass environmental and social aspects that must be integrated to sustain appropriate management measures in a potential disaster event. Therefore, emergency plans require sounds spatial information knowledge and the interactions of their environments. The herein study aims to contribute to the emergency plans, especially to the possible evacuation routes identification for a Nuclear Power Plant located in Angra dos Reis, Brazil. The study was conducted within 120 km radius around the power plant. Results indicate that evacuation plans should be elaborated for the urban cores with low roads availability, or even for high population density areas, that in spite of being well served by alternative routes, exhibit increased road traffic and population flow. These results present important information to provide support to local emergency management actions and guidelines, playing a key role to assist on the evacuation planning. [ABSTRACT FROM AUTHOR]

Glaciers are important and sensitive part of our environment which can be used as indicator of global warming and climate change. Glacier facies represent distinct regions of a glacier surface characterized by near surface structure and density that develop as a function of spatial variations in surface melt and accumulation. The facies mapping aids in delineating different zones of the glacier, which are useful in computing glacier mass balance and modeling. In this study we tested traditional and advanced classification techniques on the Edithbreen glacier situated in Ny-Ålesund, Svalbard, using WorldView-3 and Landsat 8 OLI. The comparison of the accuracy was conducted using error matrices. Six measures of accuracy were derived from the error matrices and were compared with each other to find the method delivering the most adequate output for facies mapping. The pixel-based approach applied to Landsat-8 data yielded higher accuracies (>80%) when compared to that. The object-oriented classification revealed a much better accuracy and high kappa coefficient for both low and high-resolution datasets. The study clearly indicates that the object-oriented classification provides better results for glacier facies classification when high spatial resolution is used, but for lower spatial resolution, pixel-based methods are adequate. [ABSTRACT FROM AUTHOR]

The purpose of this study is to examine the use of multi-resolution object-based classification methods for the classification of Unmanned Aircraft Systems (UAS) images of wetland vegetation and to compare its performance with pixel-based classification approaches. Three types of classifiers (Support Vector Machine, Artificial Neural Network and Maximum Likelihood) were utilized to classify the object-based images, the original 8-cm UAS images and the down-sampled (30 cm) version of the image. The results of the object-based and two pixel-based classifications were evaluated and compared. Object-based classification produced higher accuracy than pixel-based classifications if the same type of classifier is used. Our results also showed that under the same classification scheme (i.e. object or pixel), the Support Vector Machine classifier performed slightly better than Artificial Neural Network, which often yielded better results than Maximum Likelihood. With an overall accuracy of 70.78%, object-based classification using Support Vector Machine showed the best performance. This study also concludes that while UAS has the potential to provide flexible and feasible solutions for wetland mapping, some issues related to image quality still need to be addressed in order to improve the classification performance.

Quality tree species information gathering is the basis for making proper decisions in forest management. By applying new technologies and remote sensing methods, very high resolution (VHR) satellite imagery can give sufficient spatial detail to achieve accurate species-level classification. In this study, the influence of pansharpening of the WorldView-3 (WV-3) satellite imagery on classification results of three main tree species (Quercus robur L., Carpinus betulus L., and Alnus glutinosa (L.) Geartn.) has been evaluated. In order to increase tree species classification accuracy, three different pansharpening algorithms (Bayes, RCS, and LMVM) have been conducted. The LMVM algorithm proved the most effective pansharpening technique. The pixel- and object-based classification were applied to three pansharpened imageries using a random forest (RF) algorithm. The results showed a very high overall accuracy (OA) for LMVM pansharpened imagery: 92% and 96% for tree species classification based on pixel- and object-based approach, respectively. As expected, the object-based exceeded the pixel-based approach (OA increased by 4%). The influence of fusion on classification results was analyzed as well. Overall classification accuracy was improved by the spatial resolution of pansharpened images (OA increased by 7% for pixel-based approach). Also, regardless of pixel- or object-based classification approaches, the influence of the use of pansharpening is highly beneficial to classifying complex, natural, and mixed deciduous forest areas.

Narcotic plants contain substances that cause unusual excitation and subsequent depression of the central nervous system. Many narcotic plants contain substances that have medicinal properties and are used primarily as pain relievers. Opium poppy (Papaver somniferum) is one of the most cultivated narcotic plants. The planning of Opium poppy growing is controlled by United Nations and Drugs and Medicines Control Program. This study was conducted to establish a basis for determining the traceability of poppy cultivating areas using remote sensing in the large plain. A very high-resolution QuickBird-2 satellite image was used in the study. The software of ERDAS Imagine and eCognition Developer was performed for image processing and classification. A variety of classification methods were performed on the satellite image. ArcGIS software was used for accuracy assessment, ground control, and map production of the poppy cultivating area. The accuracy of classification methods was compared. The producer–user accuracy was estimated as 97.99% using spectral difference sub-segmentation process of multiresolution segmentation algorithm in object-based classifications method. This algorithm can be a practical approach for determining of opium poppy parcels in the large plain. [ABSTRACT FROM AUTHOR]

The role of image classification based on multi-source, multi-temporal and multi-resolution remote sensed data is on the rise in the environmental studies due to the availability of new satellite sensors, easier access to aerial orthoimages and the automation of image analysis algorithms. The remote sensing technology provides accurate information on the spatial and temporal distribution of land use and land cover (LULC) classes. The presented study focuses on LULC change dynamics (especially secondary forest succession) that occurred between 1974 and 2010 in the Błędów Desert (an area of approx. 1210 ha; a unique refuge habitat – NATURA 2000; South Poland). The methods included: photointerpretation and on-screen digitalization of KH-9 CORONA (1974), aerial orthoimages (2009) and satellite images (LANDSAT 7 ETM+, 1999 and BlackBridge – RapidEye, 2010) and GIS spatial analyses. The results of the study have confirmed the high dynamic of the overgrowth process of the Błędów Desert by secondary forest and shrub vegetation. The bare soils covered 19.3% of the desert area in 1974, the initial vegetation and bush correspondingly 23.1% and 30.5%. In the years 2009/2010 the mentioned classes contained: the bare soils approx. 1.1%, the initial vegetation – 8.7% and bush – 15.8%. The performed classifications and GIS analyses confirmed a continuous increase in the area covered by forests, from 11.6% (KH-9) up to 24.2%, about 25 years later (LANDSAT 7) and in the following 11 years, has shown an increase up to 35.7% (RapidEye 2010).

The aim of this study was to investigate the possible use of geoinformatics tools and generally available geodata for mapping land cover/use on the reclaimed areas. The choice of subject was dictated by the growing number of such areas and the related problem of their restoration. Modern technology, including GIS, photogrammetry and remote sensing are relevant in assessing the reclamation effects and monitoring of changes taking place on such sites. The LULC classes mapping, supported with thorough knowledge of the operator, is useful tool for the proper reclamation process evaluation.

Progress in urban climate science is severely restricted by the lack of useful information that describes aspects of the form and function of cities at a detailed spatial resolution. To overcome this shortcoming we are initiating an international effort to develop the World Urban Database and Access Portal Tools (WUDAPT) to gather and disseminate this information in a consistent manner for urban areas worldwide. The first step in developing WUDAPT is a description of cities based on the Local Climate Zone (LCZ) scheme, which classifies natural and urban landscapes into categories based on climate-relevant surface properties. This methodology provides a culturally-neutral framework for collecting information about the internal physical structure of cities. Moreover, studies have shown that remote sensing data can be used for supervised LCZ mapping. Mapping of LCZs is complicated because similar LCZs in different regions have dissimilar spectral properties due to differences in vegetation, building materials and other variations in cultural and physical environmental factors. The WUDAPT protocol developed here provides an easy to understand workflow; uses freely available data and software; and can be applied by someone without specialist knowledge in spatial analysis or urban climate science. The paper also provides an example use of the WUDAPT project results.

The ever increasing need for accurate burned area mapping has led to a number of studies that focus on improving the mapping accuracy and effectiveness. In this work, we investigate the influence of derivative spectral and spatial features on accurately mapping recently burned areas using VHR IKONOS imagery. Our analysis considers both pixel and object-based approaches, using two advanced image analysis techniques: (a) an efficient feature selection method based on the Fuzzy Complementary Criterion (FuzCoC) and (b) the Support Vector Machine (SVM) classifier. In both cases (pixel and object-based), a number of higher-order spectral and spatial features were produced from the original image. The proposed methodology was tested in areas of Greece recently affected by severe forest fires, namely, Parnitha and Rhodes. The extensive comparative analysis indicates that the SVM object-based scheme exhibits higher classification accuracy than the respective pixel-based one. Additionally, the accuracy increased with the addition of derivative features and subsequent implementation of the FuzCoC feature selection (FS) method. Apart from the positive effect in the classification accuracy, the application of the FuzCoC FS method significantly reduces the computational requirements and facilitates the manipulation of the large data volume. In both cases (pixel and objet) the results confirmed that the use of an efficient feature selection method is a prerequisite step when extra information through higher-order features is added to the classification process of VHR imagery for burned area mapping.

The delimitation of urban space is conceptually elusive and fuzzy. Commonly, urban areas are delimited through administrative boundaries. These artificial, fixed boundaries, however, do not necessarily represent the actual built-up extent, the urban catchment, or the economic linkage within and across neighboring metropolitan regions. For an approach to spatially delimit an urban corridor--a generically defined concept of a massive urban area--we use the Boston to Washington (Boswash) region as an example. This area has been consistently conceptualized in literature as bounded urban space. We develop a method to spatially delimit the urban corridor using multi-source geodata (built-up extent, infrastructure and socioeconomic data) which are based on a grid rather than on administrative units. Threshold approaches for the input data serve to construct Boswash as varying connected territorial spaces, allowing us to investigate the variability of possible spatial forms of the area, i.e., to overcome the simple dichotomous classification in favor of a probability-based differentiation. Our transparent multi-layer approach, validated through income data, can easily be modified by using different input datasets while maintaining the underlying idea that the likelihood of an area being part of an urban corridor is flexible, i.e., in our case a factor of how many input layers return positive results. [ABSTRACT FROM AUTHOR]

Pixel-based and object-based classifications are two commonly used approaches in extracting land cover information from remote sensing images. However, they each have their own inherent merits and limitations. This study, therefore, proposes a new classification method through the integration of pixel-based and object-based classifications (IPOC). Firstly, it employs pixel-based soft classification to obtain the class proportions of pixels to characterize the land cover details from pixel-scale properties. Secondly, it adopts area-to-point kriging to explore the class spatial dependence between objects for each pixel from object-based soft classification results. Thirdly, the class proportions of pixels and the class spatial dependence of pixels are fused as the class occurrence of pixels. Last, a linear optimization model on objects is built to determine the optimal class label of pixels within each object. Two remote sensing images are used to evaluate the effectiveness of IPOC. The experimental results demonstrate that IPOC performs better than the traditional pixel-based hard classification and object-based hard classification methods. Specifically, the overall accuracy of IPOC is 7.64% higher than that of pixel-based hard classification and 4.64% greater than that of object-based hard classification in the first experiment, while the overall accuracy improvements in the second experiment are 3.59% and 3.42%, respectively. Meanwhile, IPOC produces less salt and pepper effect than the pixel-based hard classification method and generates more accurate land cover details and small patches than the object-based hard classification method. [ABSTRACT FROM AUTHOR]