Search

• CNNs and SVMs outperform in detecting and measuring road surface damages (RSD). • Smartphones, cameras, and accelerometers are frequent in RSD detection. • Server receiving images and backed by ML tools are potential in detecting RSD. • Sensor-based studies are less effective in detecting road surface cracks. • Engaging road users' community is an emerging technique to grab images for RSD detection. This review critically analyzes available tools and techniques for road damage assessment and monitoring and identifies further research areas. After the review, several are found for image and sensor data collection, such as smartphones, cameras, accelerometers, and Unnamed Aerial Vehicles (UAVs). Of these, smartphones are the most frequently used among the data processing tools, including Machine Learning (ML) algorithms, Convolution Neural Networks (CNNs), and Support Vector Machines (SVMs): they outperform in detecting and measuring various forms of road surface damage, including potholes, bumps, ruts, and patches. It is also found that the study of crack identification is comparatively lacking. Moreover, some studies attempt to capitalize on crowdsourcing or road users in data collection connected to a server while driving over road networks. However, physical observation and smartphone-based image posting to the server backed by appropriate ML tools for automatic and dynamic road damage detection are yet to be demonstrated or developed and are, therefore, of further research potential. Finally, from a national perspective, India and China are ahead of other countries. Thus, future research can focus on image processing-based automatic road performance monitoring in any country with a transportation system that is highly dependent on paved roads. [ABSTRACT FROM AUTHOR]

• This study presents an extensive systematic literature review on WD recognition. • A set of 94 research articles published between 2005–2022 has been selected for analysis. • A framework has been proposed to classify the research articles into four categories. Due to wheat diseases (WD), the global rate of wheat production is decreasing by 3.6% annually. With the help of computer vision technology, WD recognition is not a challenging task but motivates the concepts of image processing, image segmentation, feature extraction, and AI-based recognition models. The objective of this study is to review and systematically analyze studies that have been published between 2005 and 2022. The authors make an effort to determine the important developments in image segmentation models, tools, datasets, and comparative analysis for the accuracy of the recognition model. The current study follows the standard systematic literature review (SLR) approach and selects 638 studies from five different web source databases. Among 638 studies, 544 studies were discarded in the study extraction process. A total number of 94 studies have been published in 45 reputed journals, and 49 conferences that have been identified with evaluation, validation, proposed, and philosophical criteria. After analysis, ten types of image segmentation models were identified. The most prominent clustering-based image segmentation technique (34.78%) is used for powdery mildew and stripe rust WD recognition. During WD recognition, the accuracy performance parameter is found to be most prominent. China and India are the two countries on the Asian continent that contribute to WD recognition. The current study summarizes the findings of WD research and highlights the need for standard datasets and accuracy. It highlights the importance of exploring and developing more precise and hybrid segmentation classification models for WD recognition. [ABSTRACT FROM AUTHOR]

The maximum reflectance of vitrinite or graphite microcrystals in Yongan Coalfield, Fujian Province, China, ranges from 5.33% to 10.21%, comprising meta-anthracite, semi-graphite, and graphite. The morphological characteristics of aromatic fringes during different evolution stages were quantitatively analyzed using high-resolution transmission electron microscopy and image processing techniques. The results show that a nonlinear evolution of aromatic fringe morphology in the process of coal graphitization is manifested in four aspects, involving the rapid connection and growth, enhancement of ordering alignment, increase of stacking and development of straightening. In meta-anthracite samples, 3 × 3 aromatic rings representing short aromatic fringes are dominated, a majority of aromatic fringes were curved, disjoint or separate, and show a turbostratic layers, with poorly aligned. The spacing between adjacent aromatic fringes exceeded the range of graphitized carbon, and structural defects, such as nanopores were generally visible. In semi-graphite samples, in addition to 3 × 3 aromatic rings, the proportion of >8 × 8 aromatic rings represented longer aromatic fringes increasing significantly, along with ordering alignment and straightening. Stacking with two more layers increased and the spacing between adjacent aromatic fringes decreased to the range of graphitized carbon, and the structural defects decreased considerably. In graphite sample, the proportion of >8 × 8 aromatic rings increase significantly, the length, stacking, and ordering alignment of aromatic fringes increased abruptly, the spacing between adjacent aromatic fringes remained within the range of natural graphite, and a rapid graphite-like structure with flat aromatic fringes formed. A stage evolution of aromatic fringe morphology, whereas remaining almost constant of mainly organic elements, which imply the evolution of aromatic fringes in meta-anthracite originated from ring condensation dominated by late coalification, graphitization starts at semi-graphite stage and enhanced during graphite stage. The making up, originating from linkup and pileup, considered as a special mechanism for the graphitization of aromatic fringes. [Display omitted] • A nonlinear evolution of length, direction, stacking and curvature of aromatic fringes. • Coal graphitization starts in semi-graphite and show an abruptly change. • Making up of BSUs is the main mechanism of coal graphitization. [ABSTRACT FROM AUTHOR]

• The all-sky airglow imager and Digisonde data were used to investigate plasma bubbles. • Optical image processing was performed to obtain image products with high resolution. • Plasma bubbles appeared like long stripe and bifurcated vertically. • The distance between adjacent plasma bubbles was in the range of 100–400 km. • Plasma bubbles generally drifted eastward and the average drift velocity was 90 m/s. Ionospheric plasma bubbles and scintillation phenomena often occur at low latitudes in China. The all-sky airglow imager located at Fuke station in Hainan (19.5°N, 109.2°E) is part of the Meridian project in China. In this work, we used mainly the observation data obtained by the imager and a digisonde to study plasma bubbles. Optical image processing—including image enhancement, azimuth correction and image projection—was performed on the original airglow observation images, and clear high-resolution images of the bubbles were obtained. Based on these, the morphological features, spatial and temporal variations, spatial scale, drift velocity of the plasma bubbles in the F region of the ionosphere were investigated and analyzed. The plasma bubbles appeared as long, dark stripes on the bright airglow emission background. Their large-scale structures often bifurcated vertically, and were usually tilted westward with north-south extension along the magnetic field lines. The east-west scale of the plasma bubbles was usually greater than 200 km and these increased gradually with time. Moreover, the distance between the adjacent plasma bubbles also increased over time, and was in the 100–400 km range. Plasma bubbles generally drifted eastward with an average drift velocity of approximately 90 m/s. The occurrence and duration of the plasma bubbles were compared with digisonde observations, and it was shown that plasma bubbles lasted for nearly four hours. Additionally, the main characteristics of the plasma bubbles were obtained using both optical and radio observations and the study of which can be beneficial for monitoring and predicting ionospheric scintillations. [ABSTRACT FROM AUTHOR]

Global warming has caused severe regional water security risks, and from the theory of the hydrological cycle, the daily extreme temperature could also bring an impact on the streamflow volume, which could be even more important than the average temperature. Therefore, based on the level of the maximum or minimum temperature warming scenarios, a variety of meteorological datasets were selected to assess the asymmetric effect of increased daily extreme temperature on the streamflow from a multi-scale perspective by using the Variable Infiltration Capacity (VIC) model. Model simulations indicate that the streamflow experiences more significant changes in response to the maximum temperature than in response to the minimum temperature, and the relationships of streamflow with both the maximum and minimum temperatures show an upwards parabolic response function, but the response function varies with the type of warming. Additionally, the seasonal and monthly duration curves results show that the increases in both the maximum and minimum temperatures demonstrate a similar response that leads the proportion of the flood period streamflow to be increased (Tmin : 0.16–0.53%/°C; Tmax : 0.11–0.51%/°C). When the minimum temperature increases, the higher the original temperature in the region, the greater the proportion of the flood period streamflow increases. However, when the maximum temperature increases, the opposite effect occurs. • This paper included a numerical analysis of the sensitivity of streamflow to daily extreme temperature increases. • This paper discusses in detail the regional asymmetric effect of daily extreme temperature increase to streamflow. • A relatively complete picture of the hydrological processes caused by daily extreme temperature rise is presented. [ABSTRACT FROM AUTHOR]

Real-time unmanned aerial vehicles (UAVs)-based thermal infrared images processing, due to high spatial resolution and knowledge of the various infrared radiant energy level distribution of solid bodies, has important applications such as monitoring and control of the various phenomena in different natural situations. One of these applications is monitoring the ground vehicles in cities by using detection or semantic segmentation of them in the thermal images. In this research, our purpose is to improve the performance of deep learning combined model by using Gaussian-Bernoulli Restricted Boltzmann Machine (GB-RBM) specifications for the segmentation of the ground vehicles from UAV-based thermal infrared imagery. The proposed model is studied in three steps. First, designing the proposed model by using an encoder-decoder structure and addition of extracted features from convolutional layers and restricted Boltzmann machine in the network. Second, the implementation of the research goals on four sets of UAV-based thermal infrared imagery named NPU_CS_UAV_IR_DATA that was collected from some streets of China by using FLIR TAU2 thermal infrared sensor in 2017. Finally, analyzing the performance of the proposed model by using five state-of-the-art models in semantic segmentation. The results evaluated the performance of the proposed model as a robust model with the average precision and average processing time of approximately 0.97, and 19.73 s for all datasets, respectively. [ABSTRACT FROM AUTHOR]

• A surrogate model for distributed hydrological models is developed using deep networks. • The surrogate model can handle high-dimensional input–output maps of VIC. • The surrogate model can capture the dynamic relationship between the time-varying input and output. • The surrogate model outperforms LSTM in approximating spatial patterns of VIC output. Surrogate modelling methods have been used to estimate spatially distributed parameters of land surface hydrological models due to their computational efficiency. However, traditional surrogate techniques have trouble in modelling high-dimensional input–output maps. In addition, their applications to a distributed model over large spatial and temporal domains are very unlikely to produce consistent spatial pattern and dynamic relationship of model input and output, leading to unbalanced model performance across different spatiotemporal domains. To address these issues, we employ a deep autoregressive neural network to construct an accurate and reliable surrogate system of distributed dynamic model outputs. The network uses a deep convolutional encoder-decoder architecture to take full advantage of the deep convolutional layers in high-dimensional image processing and spatial feature extraction. Moreover, the autoregressive strategy makes the network good at handling dynamic relationship between the time-varying model input and output. We apply this deep learning (DL) network to building a surrogate model of the Variable Infiltration Capacity (VIC) model simulated soil moisture over the Huaihe river basin of China. The results show that the adopted deep autoregressive neural network can provide an accurate surrogate model of the high-dimensional dynamic relationship between 15 input fields and 1 output field, each covering 408 grids, over multiple years. This surrogate model significantly outperforms the popular long short-term memory (LSTM) model, achieving an average mean squared error (MSE) of 0.26 and an average R 2 of 0.76 using the test datasets. The surrogate's MSE and R 2 performance represents an average improvement of 43% and 33%, respectively, compared to the LSTM model. In addition, it shows better spatial performance than the LSTM model, demonstrating its superior performance in approximating spatial patterns of VIC simulated soil moisture. The accuracy of this surrogate method and its ability to handle high-dimensional data is promising for advancing parameter uncertainty quantification of distributed land surface hydrological models, which usually have high dimensionality due to the large number of variables involved, including distributed parameters, input forcing variables, and output variables. [ABSTRACT FROM AUTHOR]

Pedestrian dynamics in emergent situations such as stampedes and fire hazards have been widely studied, while few studies are focusing on human behavior in terrorist attack events. The research is vitally important for guiding efficient evacuation of the attacked crowd and reducing casualties. However, the research is insufficient due to the lack of researching materials. Besides, it is unconvincing for researchers to carry out these types of experiments. In this paper, the recorded video of the 3 ⋅ 1 Kunming terrorist attack event in China is analyzed. The instantaneous velocity field of the attacked crowd in the ticket hall is extracted from the video material, which is based on an image processing method. By analyzing the statistical characteristics of crowd velocity, crowd movement frequency and crowd entropy, several phenomena are found including crowd oscillating phenomenon, self-organizing grouping phenomenon. These discoveries will benefit the understanding of crowd dynamic in close-range attacks which is essential for the evacuation guidance efficiency, better design of architecture layout, and better defending strategies against the terrorists. • Crowd velocity peak is oscillating back and forth within the attacked crowd. • The velocity entropy of crowd is shown oscillating around the mean value 6.7. • The characteristic frequency of crowd movement is about 0.4 Hz. • Crowd in different area who have different movement behavior patterns show grouping phenomenon. [ABSTRACT FROM AUTHOR]

Highlights • Time-related factors and distances-related factors affect the drivers' deceleration rate choice in Vehicle-Pedestrian conflict have been analyzed. • The drivers' effective decelerating zone with sight angle has been proposed with the consideration of longitudinal & lateral conflict behavior. • Intercultural comparison has been made to address the significant differences in driver behavior in China and Germany. Abstract To analyse the drivers' decelerating behavior in conflict avoidance process under Vehicle-Pedestrian conflict situation is accordingly a surrogate method for the guidance of driver evasive manoeuvre and improving the active driving safety. Field observation data has been collected by video recording and image processing at urban unsignalized midblock crosswalks in Beijng, China and Munich, Germany. The relationship between deceleration rate and vehicle speed has been statistically described. By analyzing how the time-related measures (Vehicle Time to Conflict Point and Pedestrian Time to Conflict Point) and the distance-related measures (Vehicle longitudinal distance to Conflict Point and Pedestrian Lateral distance to Conflict Point) affect the deceleration rate choice, the drivers' effective decelerating zone with sight angle has been proposed with the consideration of longitudinal & lateral conflict behavior. Minimum Vehicle-Pedestrian conflict distance is calculated and a 2 m minimum distance is suggested as the threshold for collision discrimination. Comparisons which address the needs to make the intercultural adaptation when introducing Driver Assistance Systems to developing countries have also been made to show the significant differences between China and Germany. The whole work could hopefully be the groundwork in the target of conflict pre-scanning by decelerating zones under various conflict situations and the design of vehicle active safety systems. [ABSTRACT FROM AUTHOR]

Abstract Under a background of a significantly warming climate, current aeolian desertification in China's northeastern Tibetan Plateau (NETP) has been widely studied, but disagreements related to its causes continue to persist. In order to understand the present through the past, this study determined the various phases of aeolian desertification in the NETP based on the statistical analysis of 40 optically stimulated luminescence (OSL) aeolian sand dates from 23 aeolian sections. Because the redistribution of aeolian sand by wind processes can reduce the number of older sand dates, this research only affects exactly the past millennium. Following this, we analyzed the relationships between aeolian desertification phases in conjunction with reconstructed climatic data. Results revealed three distinct aeolian desertification phases that approximately correspond to 1480 CE, 1680 CE, and 1800 CE, respectively. By comparing these dates to well-established paleoclimate records, we determined that these three aeolian desertification phases all occurred under a cold and dry climate background. Accordingly, aeolian desertification in the NETP today under its current warm and wet climate background can therefore be attributable to irrational human activities rather than rising temperatures. Accordingly, we can infer that under the current warmer and wetter climate background, aeolian desertification in the NETP today can be reversed as long as human activities remain moderate, even without the implementation of artificial revegetation initiatives. Graphical abstract Unlabelled Image Highlights • The past aeolian desertification in the Tibetan Plateau was studied. • The purpose is to understand the present through the past. • Results revealed that current aeolian desertification is attributable to human activities. • Negative effects of rising temperatures on cold alpine ecosystems have been overestimated. [ABSTRACT FROM AUTHOR]

Quantifying the distribution of caliche nodules in a catchment is a basis toward understanding soil erosion and soil degradation as well as soil-water processes in the soil containing caliche nodules. In this study, field surveys and soil sampling were coupled with spatial and statistical analyses along with image processing to investigate the spatial distribution of caliche nodules and their influencing factors in the Liudaogou catchment, the northern Loess Plateau of China. Results showed that more caliche nodules were present at the top of the mountain, mountain ridges, roadsides, edges of ravines, and other high elevation regions, as represented by comparatively higher caliche nodule coverage (CNC, 11.26–18.28%). Caliche nodules were rarely present in low elevation areas such as the terraced cropland and check dam, as represented by lower CNC (<2.55%). More than 65% of the total caliche nodules had a diameter of 10–50 mm. The caliche nodule coverage and diameter showed strong spatial dependence. The CNC was significantly related to the slope gradient and vegetation cover (P < 0.01), and the caliche nodule diameter (CND) was significantly related to the elevation (P < 0.01). Two simple models were established to predict the CNC and CND values given the parameters of slope gradient and vegetation cover. The validation of the models showed relatively high accuracy. These findings quantify caliche nodule distribution in a catchment for the first time, and may benefit the assessment of soil erosion and soil degradation in this special soil in the Loess Plateau. [ABSTRACT FROM AUTHOR]

Affected by groundwater fluctuations, the loess around the hydro-fluctuation zone near the groundwater table is frequently subjected to periodic dry-wet cycles. The investigation on the creep characteristics and microstructural evolution associated with the mechanical response of loess after the application of dry-wet cyclings is of great significance to better understand the loess landslide mechanism. Triaxial creep tests and scanning electron microscopy (SEM) were conducted on intact loess specimens obtained from Heifangtai platform, Gansu, China. The micrographs were processed using IPP software and the microstructural information of soil pores were obtained. Results of triaxial creep test revealed that the application of dry-wet cyclings causes the creep deformation of loess soils to increase significantly, and they showed that the long-term strength and dry-wet cycle follow an exponentially decreasing function, indicating that the reduction in the long-term strength of loess soils is greater in the first five dry-wet cycling compared with that in the subsequently dry-wet cyclings. The microstructural analysis showed that aggregates were found to cluster upon the application of drying-wetting cycling, followed by the rearrangement of soil skeletons, transforming the initial relatively dense structure into a loosen one with a greater percentage of macro-pores which are more sensitive to creep deformation of loess soils. Additionally, it is observed that the microstructural evolution is dependent on the number of dry-wet cyclings. As the dry-wet cycles increase, the flatter and irregular macro-pores increased gradually at the expense of meso-pores. In addition, the mechanism of irrigation-induced loess landslides is revealed from the perspective of variation in creep characteristics of loess after the dry-wet cycles, which can be summarized into three stages, namely, the dry-wet cycle stage of the loess around the hydro-fluctuation zone; the long-term creep-liquefaction stage of the deep loess, and the slip surface transfixion stage. • Reduction in q L is the greatest in the first five dry-wet cycling compared with that in the subsequently dry-wet cycling. • The microstructure of loess samples subjected to dry-wet cycling after yielding was investigated by SEM. • Flatter and irregular macro-pores increased gradually at the expense of meso-pores with an increase in the dry-wet cycling. [ABSTRACT FROM AUTHOR]

• A Convolutional Long Short Term Memory Network is used to predict the flood events based on deep learning techniques. • The spatial and time characteristics of floods in China are well modeled to overcome the shortcomings generated by merely relying on time-series analysis. • Different from traditional methods, the hydrological area is gridded into different watersheds for future processing using image processing methods. Floods cause substantial damage across the world every year. Accurate and timely prediction of floods can significantly minimize the loss of life and property. Recently, numerous machine learning models have been used for flood prediction, showing that their performance is preferable to traditional statistical models. However, the existing models neglect the spatial features of floods, which drive flood generation and concentration. In this paper, the area of interest is divided into grids based on longitude and latitude, and the rainfall and discharge collected by stations are combined into tensors according to station coordinates. Different from one-dimensional time series, our input feature is a two-dimensional time series with spatial information. Hence, combining a Convolutional Neural Network (CNN) with a Long Short Term Memory Network (LSTM), we propose the convolution LSTM (ConvLSTM) to extract spatiotemporal features of hydrological information. The methodology is demonstrated using the hydrological data collected at the Xi County stations, located on the Huai River in Henan Province, China. Numerical results indicate that the relative error of arrival time is within 30%, and the relative error of peak discharge is within 20%, satisfying the 2005 Chinese Water Resource Standard on flood prediction permit error. The experiments also show that the ConvLSTM outperforms the recent models in terms of flood arrival time and peak discharge, thereby proving a promising alternative. [ABSTRACT FROM AUTHOR]

Abstract: Remote diagnose of fish diseases for farmers is unrealized in China, but use of mobile phones and remote analysis based on image processing can be feasible due to the widespread use of mobile phones with camera features in rural areas. This paper presents a novel method of classifying species of fish based on color and texture features and using a multi-class support vector machine (MSVM). Fish images were acquired and sent by smartphone, and the method utilized was comprised of the following stages. Color and texture subimages of fish skin were obtained from original images. Color features, statistical texture features and wavelet-based texture features of the color and texture subimages were extracted, and six groups of feature vectors were composed. LIBSVM software was tested using leave-one-out cross validation to find the best group for classification in feature selection procedure. Two multi-class support vector machines based on a one-against-one algorithm were constructed for classification. The feature selection results showed that the Bior4.4 wavelet filter in HSV color space achieved greater accuracy than the other feature groups. The classification results indicate that only the DAGMSVM meets the requirement of time efficiency for the system. The results of this study suggest that the best classification model for fish species recognition is composed of a wavelet domain feature extractor with Bior4.4 wavelet filter in HSV color space and a one-against-one algorithm based DAGMSVM classifier. [Copyright &y& Elsevier]

Abstract: In this paper a method based on in situ measured reflectance spectra was developed for accurately mapping salt fields in the Taibei Salt Farm in Lianyungang City of East China. After radiometric correction and geometric rectification, six multispectral ETM+ bands were fused with the panchromatic band via principal components analysis (PCA). The fused data were used to map salt fields in the farm. Salt fields were mapped 91.95% correctly in comparison with ground statistics from raw multispectral bands. This accuracy level rose to 96.4% with the use of the panchromatic band (15m resolution). However, PCA-fused data produced the highest accuracy of 98.8%. At the zone level, coarse resolution data resulted in an accuracy of 98.93% for crystallization ponds, but only 82.68% for condensation ponds. Use of the panchromatic band alone improved the accuracy for condensation ponds but decreased the accuracy for crystallization and evaporation ponds. Results derived from PCA-fused data are highly consistent with the ground statistics at a discrepancy between 0.35% for evaporation zone and 1.80% for condensation zone. It is concluded that a fine spectral resolution is important to separation of ponds of a varying salt content while a fine spatial resolution is important to their mapped area. Ponds of a lower salt content can be mapped more accurately from fused data than from raw spectral bands. [Copyright &y& Elsevier]

Abstract: The just noticeable difference (JND) in black level (BL), white level (WL), and chroma as visible in natural images under practical viewing conditions is determined. The same experiments are conducted in China and in the Netherlands to evaluate possible ethnic differences in the size of JNDs. In both countries, there is a strong effect of image content on the JND of BL and WL, while for most images used in this experiment no content dependency is found for the JND of CS. For most images, there is no difference in JND of WL and CS between China and the Netherlands. For some images, we found a difference in JND of BL between China and the Netherlands, but this difference most probably is related to the fact that the viewers looked to a different part in the image, rather than to an ethnic difference. [Copyright &y& Elsevier]

• An automatic winter-jujube grading robot based on machine vision was designed. • A method combining YOLOv3 algorithm and hand-engineered features was developed to calculate the maturity of winter-jujube. • This method can detect overlapping winter-jujubes accurately and has a high stability under different lighting conditions. • The grading device has high ability to sort winter-jujube into three maturity categories. Winter-jujube is a kind of fresh fruit in China. After harvest, winter-jujubes require to grade into different categories according to their maturity levels. Mature winter-jujube can be recognized by their red colour. In this study, a winter-jujube grading robot is designed. Moreover, a method combining YOLOv3 algorithm and hand-engineered features is developed to calculate the maturity of winter-jujube. The grading robot is composed of a transmission unit, an image acquisition unit, and an actuator unit. Based on YOLOv3 algorithm, a detection model is trained, and compared with SSD and Faster R-CNN algorithms. When the IoU are 0.7, 0.8, and 0.9, the F1 scores of the model are 100%, 100%, and 93.66%, respectively. The mAP (IoU = 0.50:0.05:0.95) of the model is 94.78%, and the detection time of single image is 0.042 s. The detection model exhibits a high stability under different lighting conditions. In addition, overlapping winter-jujubes in the image can be detected accurately. After image distortion correction and object detection, an image processing flow for spatial positioning, size measurement and maturity calculation for winter-jujubes is designed. Finally, a real-time grading device for winter-jujube is built to perform grading experiments. The maturity grading accuracy is 97.28%, and the average grading time of each winter-jujube is 1.39 s. [ABSTRACT FROM AUTHOR]

The development of computer tomography and image processing has made it possible to establish virtual 3D reconstruction and non-invasive dissection of fossil specimens. We used these methods to reconstruct a virtual 3D skull of a Pleistocene cheetah skull from the Tuozi cave, Tangshan, Nanjing, and virtually dissected it for anatomic studies, and measured the volumes of different parts of the endocranium. The endocranium of the cheetah skull has showed that its frontal sinus is beehive-like, the frontal lobe of cerebra is relatively large but the temporal lobe is relatively small, its cerebral sulcus and gyrus are more complicated than those of the domestic cat, similar to those of the domestic dog, but simpler than those of giant panda, pig, cattle and horse. The technology of virtual 3D reconstruction and non-invasive dissection of fossil specimens can extend the morphological study from the exterior to the interior, and it can also help to study fragile specimens and virtually backup rare and precious specimens. [ABSTRACT FROM AUTHOR]

China frequently experiences natural disasters, of which flooding is the most serious. How to monitor and control natural disasters, assess damage, and provide relief is the most urgent problem for the Chinese government and disaster experts. A national integrated system using remote sensing, geographic information systems, the Global Positioning System, and other technology for monitoring and evaluating flood disasters has been assembled and tried out for 3 years. The system has played an important role in flood mitigation during the trial and has become a key part of the flood management system at China''s National Flood Control Headquarters. This paper presents an overview of the system and its use in China. [Copyright &y& Elsevier]

• An automated Kashin Beck Disease (KBD) diagnosis algorithm was proposed based on deep learning. • The method focuses on multi-features fusion for detection. The features are extracted from hand X-ray images by DCNN, including global and local features. • Global features represent structures of hand bones, and local features represent subtle edge information from critical regions of metaphysis based on the domain knowledge of KBD. • The model can effectively diagnose KBD and provide substantial benefits to reduce largescale screening costs and missed diagnosis rate. Kashin-Beck Disease (KBD) is a serious endemic bone disease leading to short stature. The early radiological examinations are crucial for potential patients. However, many children in rural China cannot be diagnosed in time due to the shortage of professional orthopedists. In this paper, an algorithm is developed to automatically screening KBD based on hand X-ray images of subjects, which can help the government reducing human resources investment and assisting the poor precisely. The KBD diagnosis method focuses on multi-feature fusion for classification. Two kinds of features presented in X-ray images are extracted by a deep convolutional neural network (DCNN). One is the global features that represent shapes and structures of the whole hand bone. The other is local features that represent edge and texture information from critical regions of the metaphysis. The global features tend to sketch the major informative parts, whereas other fine local features can provide supplementary information. Then both kinds of features are combined and fed into the KBD classifier of a fully connected neural network (FCNN) to obtain diagnostic results. Our research team collected 960 samples in KBD endemic areas of Tibet from 2017 to 2018. The dataset contains 219 KBD positive images and 741 negative images. Experiments indicate that the method based on multi-feature achieves the best average accuracy and sensitivity rate of of 98.5% and 97.6% for diagnosis, which is 4.0% and 7.6% higher than the method with only the global features respectively. The KBD diagnosis method shows that our proposed multi-feature fusion helps to achieve higher diagnosis performance and stability compared with only using global features for detection. The automated KBD diagnosis algorithm provides substantial benefits to reduce large-scale screening costs and missed diagnosis rate. [ABSTRACT FROM AUTHOR]

• An arc approximation algorithm for cucumber curvature is proposed. • The center line of the cucumber coincides with the approximated arc curve. • The new algorithm has higher classification advantages than elliptic approximation. The cucumber is one of the most important consumer vegetables, and due to high freshness and appearance standards, it must be graded according to quality. One key metric for grading cucumbers is curvature (arch height) relative to length. To date, this classification standard has not been implemented in a commercial automatic system because the curvature of the cucumber is related to its length. This paper presents and tests a circular arc approximation algorithm for measuring the curvature of a cucumber using image processing. By fitting the central axis of each cucumber to establish a curve equation, the approximate eigenvalue radius, R, and the radius of the section perpendicular to the curvature of cucumber, r, could be obtained quickly and efficiently. The curve of the cucumber was then transformed into a new characteristic, R' (R' = R – r), which was used as a threshold to classify each cucumber. The method was verified by theory and experiment, and the average error of 0.71 % corresponding to the coefficient R2 = 0.9958 of the similarity index between the curve and central axis of the cucumber was verified. It was found that the fitted curve and the central axis of the cucumber were coincident. Moreover, we designed an algorithm that can detect S-shaped cucumbers based on the analysis of the contour, which can avoid the influence of S-shaped cucumbers on the results. In our tests, 148 cucumbers were classified using the circular arc approximation method, the elliptic approximation method, and a manual method. The results of the classifications were compared. According to China's standard evaluation, the classification error rate of the circular arc approximation method was 10.1 % lower than that of the elliptic approximation method. The incidence of cucumbers being classified into an adjacent grade was 0.7 % using the circular approximation method, and there were no cases of cucumbers being classified into non-adjacent grades. The classification error rate of the elliptic approximation was 10.8 %, of which 2.7 % was classified into non-adjacent grades. Under the European standard, the cucumber classification accuracy using the circular arc approximation algorithm reached 100 %. Therefore, the proposed method offers a more accurate classifier than the elliptic approximation method, and the circular arc approximation algorithm can be fully applied in the commercial cucumber classification process. [ABSTRACT FROM AUTHOR]