Search

This study uses a small unmanned aircraft system equipped with a multispectral sensor to assess various vegetation indices (VIs) for their potential to monitor iron deficiency chlorosis (IDC) in a grain sorghum (Sorghum bicolor L.) crop. IDC is a nutritional disorder that stunts a plants' growth and causes its leaves to yellow due to an iron deficit. The objective of this project is to find the best VI to detect and monitor IDC. A series of flights were completed over the course of the growing season and processed using Structure‐from‐Motion photogrammetry to create orthorectified, multispectral reflectance maps in the red, green, red‐edge, and near‐infrared wavelengths. Ground data collection methods were used to analyze stress, chlorophyll levels, and grain yield, correlating them to the multispectral imagery for ground control and precise crop examination. The reflectance maps and soil‐removed reflectance maps were used to calculate 25 VIs whose separability was then calculated using a two‐class distance measure, determining which contained the largest separation between the pixels representing IDC and healthy vegetation. The field‐acquired data were used to conclude which VIs achieved the best results for the dataset as a whole and at each level of IDC (low, moderate, and severe). It was concluded that the MERIS terrestrial chlorophyll index, normalized difference red‐edge, and normalized green (NG) indices achieved the highest amount of separation between plants with IDC and healthy vegetation, with the NG reaching the highest levels of separability for both soil‐included and soil‐removed VIs. Core Ideas: A small unmanned aircraft system was utilized to collect multispectral imagery.Soil background noise was assessed and removed via image thresholding and maximum likelihood classification.A total of 25 vegetation indices were evaluated with and without soil.A two‐class distance measure assessed which vegetation index best detected iron deficiency chlorosis.Normalized green, normalized difference red‐edge, and MERIS terrestrial chlorophyll indices performed best. [ABSTRACT FROM AUTHOR]

UAV swarms have multiple real-world applications but operate in a dynamic environment where disruptions can impede performance or stop mission progress. Ideally, a UAV swarm should be resilient to disruptions to maintain the desired performance and produce consistent outputs. Resilience is the system's capability to withstand disruptions and maintain acceptable performance levels. Scientists propose novel methods for resilience integration in UAV swarms and test them in simulation scenarios to gauge the performance and observe the system response. However, current studies lack a comprehensive inclusion of modeled disruptions to monitor performance accurately. Existing approaches in compartmentalized research prevent a thorough coverage of disruptions to test resilient responses. Actual resilient systems require robustness in multiple components. The challenge begins with recognizing, classifying, and implementing accurate disruption models in simulation scenarios. This calls for a dedicated study to outline, categorize, and model interferences that can be included in current simulation software, which is provided herein. Wind and in-path obstacles are the two primary disruptions, particularly in the case of aerial vehicles. This study starts a multi-step process to implement these disruptions in simulations accurately. Wind and obstacles are modeled using multiple methods and implemented in simulation scenarios. Their presence in simulations is demonstrated, and suggested scenarios and targeted observations are recommended. The study concludes that introducing previously absent and accurately modeled disruptions, such as wind and obstacles in simulation scenarios, can significantly change how resilience in swarm deployments is recorded and presented. A dedicated section for future work includes suggestions for implementing other disruptions, such as component failure and network intrusion. [ABSTRACT FROM AUTHOR]

Cultivation and grazing since the mid-nineteenth century in Texas has caused dramatic changes in grassland vegetation. Among these changes is the encroachment of native and introduced brush species. The distribution and quantity of brush can affect livestock production and water holding capacity of soil. Still, at the same time, brush can improve carbon sequestration and enhance agritourism and real estate value. The accurate identification of brush species and their distribution over large land tracts are important in developing brush management plans which may include herbicide application decisions. Near-real-time imaging and analyses of brush using an Unoccupied Aerial System (UAS) is a powerful tool to achieve such tasks. The use of multispectral imagery collected by a UAS to estimate the efficacy of herbicide treatment on noxious brush has not been evaluated previously. There has been no previous comparison of band combinations and pixel- and object-based methods to determine the best methodology for discrimination and classification of noxious brush species with Random Forest (RF) classification. In this study, two rangelands in southern Texas with encroachment of huisache (Vachellia farnesianna [L.] Wight & Arn.) and honey mesquite (Prosopis glandulosa Torr. var. glandulosa) were studied. Two study sites were flown with an eBee X fixed-wing to collect UAS images with four bands (Green, Red, Red-Edge, and Near-infrared) and ground truth data points pre- and post-herbicide application to study the herbicide effect on brush. Post-herbicide data were collected one year after herbicide application. Pixel-based and object-based RF classifications were used to identify brush in orthomosaic images generated from UAS images. The classification had an overall accuracy in the range 83–96%, and object-based classification had better results than pixel-based classification since object-based classification had the highest overall accuracy in both sites at 96%. The UAS image was useful for assessing herbicide efficacy by calculating canopy change after herbicide treatment. Different effects of herbicides and application rates on brush defoliation were measured by comparing canopy change in herbicide treatment zones. UAS-derived multispectral imagery can be used to identify brush species in rangelands and aid in objectively assessing the herbicide effect on brush encroachment. [ABSTRACT FROM AUTHOR]

Automatically detecting the wet/dry shoreline from remote sensing imagery has many benefits for beach management in coastal areas by enabling managers to take measures to protect wildlife during high water events. This paper proposes the use of a modified HED (Holistically-Nested Edge Detection) architecture to create a model for automatic feature identification of the wet/dry shoreline and to compute its elevation from the associated DSM (Digital Surface Model). The model is generalizable to several beaches in Texas and Florida. The data from the multiple beaches was collected using UAS (Uncrewed Aircraft Systems). UAS allow for the collection of high-resolution imagery and the creation of the DSMs that are essential for computing the elevations of the wet/dry shorelines. Another advantage of using UAS is the flexibility to choose locations and metocean conditions, allowing to collect a varied dataset necessary to calibrate a general model. To evaluate the performance and the generalization of the AI model, we trained the model on data from eight flights over four locations, tested it on the data from a ninth flight, and repeated it for all possible combinations. The AP and F1-Scores obtained show the success of the model's prediction for the majority of cases, but the limitations of a pure computer vision assessment are discussed in the context of this coastal application. The method was also assessed more directly, where the average elevations of the labeled and AI predicted wet/dry shorelines were compared. The absolute differences between the two elevations were, on average, 2.1 cm, while the absolute difference of the elevations' standard deviations for each wet/dry shoreline was 2.2 cm. The proposed method results in a generalizable model able to delineate the wet/dry shoreline in beach imagery for multiple flights at several locations in Texas and Florida and for a range of metocean conditions. [ABSTRACT FROM AUTHOR]

In this study, the potential of raw samples of digitized echo waveforms collected by full-waveform (FW) terrestrial laser scanning (TLS) for point cloud classification is investigated. Two different TLS systems are employed, both equipped with a waveform digitizer for access to the raw waveform and online waveform processing which assigns calibrated waveform attributes to each point measurement. Point cloud classification based on samples of the raw single-peak echo waveform is compared with point cloud classification based on the calibrated online waveform attributes. A deep convolutional neural network (DCNN) is designed for the supervised classification. Random forest classifier is used as a benchmark to evaluate the performance of the proposed DCNN model. In addition, feature importance and temporal stability of the raw waveform samples versus the calibrated waveform attributes for point cloud classification are reported. Classification results are evaluated at two study sites, a built environment on a university campus and a coastal wetland environment. Results show that direct classification of the raw waveform samples outperforms classification based on the set of waveform attributes at both study sites. Results also show that the contribution of the range, as the only geometric attribute in the raw waveform feature vector, significantly increases the classification performance. Finally, the performance of the DCNN for filtering ground points to generate a digital terrain model (DTM) based on classification of the raw waveform samples is assessed and compared to a DTM generated from a progressive morphological filter and to real-time kinematic (RTK) GNSS survey data. [ABSTRACT FROM AUTHOR]

The use of unmanned aircraft systems (UAS) delivering imaging technologies in agricultural settings has become more prevalent over the past five years and is growing in pest management programs. Here, spectral data from a three-band consumer-grade camera with a filter to obtain Near Infrared (NIR) data, mounted on a fixed-winged UAS, was used to assess the ability to detect cotton fleahopper, Pseudatomoscelis seriatus (Reuter) (Hemiptera: Miridae), injury to immature fruiting bodies on cotton. In a small plot experiment conducted two years and two planting periods each year, cotton fleahopper densities were manipulated with insecticide. Variable populations of cotton fleahopper across the plots were achieved in 2015, ranging between 0 and 3.5 cotton fleahopper-days over a five-week period when squares were forming. Derived from spectral data of multiple UAS flights, unexpected but inconsistent trends (by regression analysis) of increasing Normalized Difference Vegetation Index (NDVI), values with increasing cotton fleahopper days were detected in both plantings and years (five of 12 regressions were significant). Our preliminary data suggest that differences in cotton fleahopper activity on cotton may be reflected in NDVI values using a modified consumer-grade camera in-season. But the interpretation of NDVI may be complicated by the feeding site of cotton fleahopper, leading to unexpected and inconsistent regressions. Exploration of image resolution and bandwidth to define optical sensor needs appears important for cotton fleahopper, given its feeding habitat and injury to cotton. The application of UAS-derived remotely sensed data to detect insect-induced plant stress continues to have merit, but a merging of best suited UAS technology to the needs of detecting insect-induced cotton stress will be a researchintensive endeavor. [ABSTRACT FROM AUTHOR]

The main objective of this study is the development of a simulation and optimization method for wide-area terrain mapping with terrestrial laser scanning (TLS). The problem can be stated as follows: given a prior digital surface model (DSM) of a region of interest (e.g. from airborne lidar or structure-from-motion photogrammetry), determine the minimum number of scan locations required to seamlessly scan the terrain in the region for a given scanner range and angular field-of-view (FOV). An optimization method for measurement setup is developed using multiple viewshed analysis and simulated annealing (SA) constrained by the system performance characteristics and survey specifications. The method is evaluated at a sediment and erosion control facility with hilly terrain by comparing random scan locations versus optimized three to six scan locations. Statistical results illustrate that average visibility for random sampling increases gradually with scan locations. However, random sampling clearly underperforms in terms of scan visibility relative to five or six optimized scan locations with an average visibility of 100%. Similar patterns in optimized scan locations demonstrate that certain terrain morphometry at the study site is an essential factor for TLS survey design. Finally, an optimized solution is compared to a brute-force manual solution for determining four scan locations for conducting surveys at the study site. Results show the effectiveness of the optimization method for selecting combinations of scan locations that enable more efficient TLS survey coverage over a wider terrain area compared to manual selection. Furthermore, results demonstrate the adaptability of the method to take into consideration different scan parameters and survey conditions, such as pre-determined scan locations that may be required (e.g. a survey control monument). [ABSTRACT FROM AUTHOR]

Unmanned aircraft system (UAS) measured canopy height has frequently been determined by means of digital surface models (DSMs) derived from structure-from-motion (SfM) photogrammetry without examining specific metrics in detail. Multiple geospatial factors to be considered for the purpose of generating an accurate height estimation were characterized and summarized in this letter using UAS-SfM photogrammetry over an experimental maize field trial. This particular study demonstrated that: 1) the 99th percentile height in a 25 cm-wide crop row polygon provided the best canopy height estimation accuracy; 2) the height difference between using a rasterized DSM and direct three-dimensional (3D) point cloud was minor yet steadily increased when the DSM resolution value grew; and 3) the accuracy of the DSM-based canopy height estimation dropped significantly after the DSM resolution became coarser than 12 cm. Results also suggested that the cost function introduced in this letter has the potential to be used for optimizing the height estimation accuracy of various crop types given ground truth. [ABSTRACT FROM AUTHOR]

Accurate characterization of marsh elevation and landcover evolution is important for coastal management and conservation. This research proposes a novel unsupervised clustering method specifically developed for segmenting dense terrestrial laser scanning (TLS) data in coastal marsh environments. The framework implements unsupervised clustering with the well-known K-means algorithm by applying an optimization to determine the ";k" clusters. The fundamental idea behind this novel framework is the application of multi-scale voxel representation of 3D space to create a set of features that characterizes the local complexity and geometry of the terrain. A combination of point- and voxel-generated features are utilized to segment 3D point clouds into homogenous groups in order to study surface changes and vegetation cover. Results suggest that the combination of point and voxel features represent the dataset well. The developed method compresses millions of 3D points representing the complex marsh environment into eight distinct clusters representing different landcover: tidal flat, mangrove, low marsh to high marsh, upland, and power lines. A quantitative assessment of the automated delineation of the tidal flat areas shows acceptable results considering the proposed method is unsupervised with no training data. Clustering results based on K-means are also compared to results based on the Self Organizing Map (SOM) clustering algorithm. Results demonstrate that the developed multi-scale voxelization approach and representative feature set are transferrable to other clustering algorithms, thereby providing an unsupervised framework for intelligent scene segmentation of TLS point cloud data in marshes. [ABSTRACT FROM AUTHOR]

Lodging has been recognized as one of the major destructive factors for crop quality and yield, resulting in an increasing need to develop cost-efficient and accurate methods for detecting crop lodging in a routine manner. Using structure-from-motion (SfM) and novel geospatial computing algorithms, this study investigated the potential of high resolution imaging with unmanned aircraft system (UAS) technology for detecting and assessing lodging severity over an experimental maize field at the Texas A&M AgriLife Research and Extension Center in Corpus Christi, Texas, during the 2016 growing season. The method was proposed to not only detect the occurrence of lodging at the field scale, but also to quantitatively estimate the number of lodged plants and the lodging rate within individual rows. Nadir-view images of the field trial were taken by multiple UAS platforms equipped with consumer grade red, green, and blue (RGB), and near-infrared (NIR) cameras on a routine basis, enabling a timely observation of the plant growth until harvesting. Models of canopy structure were reconstructed via an SfM photogrammetric workflow. The UAS-estimated maize height was characterized by polygons developed and expanded from individual row centerlines, and produced reliable accuracy when compared against field measures of height obtained from multiple dates. The proposed method then segmented the individual maize rows into multiple grid cells and determined the lodging severity based on the height percentiles against preset thresholds within individual grid cells. From the analysis derived from this method, the UAS-based lodging results were generally comparable in accuracy to those measured by a human data collector on the ground, measuring the number of lodging plants (R² = 0.48) and the lodging rate (R² = 0.50) on a per-row basis. The results also displayed a negative relationship of ground-measured yield with UAS-estimated and ground-measured lodging rate. [ABSTRACT FROM AUTHOR]

Detection of near-ground objects occluded by above-ground vegetation from airborne light detection and ranging (lidar) measurements remains challenging. Our hypothesis is that the probability of obstruction due to objects above ground at any location in the forest environment can be reasonably characterized solely from airborne lidar data. The essence of our approach is to develop a data-driven learning scheme that creates high-resolution two-dimensional (2-D) probability maps for obstruction in the under-canopy environment. These maps contain information about the probabilities of obstruction (clutter map) and lidar undersampling (uncertainty map) in the near-ground space. Airborne and terrestrial lidar data and field survey data collected within the forested mountainous environment of Shenandoah National Park, Virginia, USA are utilized to test and evaluate the proposed approach in this work. A newly developed individual tree detection algorithm is implemented to estimate the undersampled stem contributions to the probability of obstruction. Results show the effectiveness of the tree detection algorithm with an accuracy index (AI) of between 61.5% and 80.7% (tested using field surveys). The estimated clutter maps are compared to the maps created from terrestrial scans (i.e., ground truth) and the results show the root-mean-square error (RMSE) of 0.28, 0.32, and 0.34 at three study sites. The overall framework in deriving near-ground clutter and uncertainty maps from airborne lidar data would be useful information for the prediction of line-of-sight visibility, mobility, and above-ground forest biomass. [ABSTRACT FROM PUBLISHER]

The world's beaches hold an appeal that draws millions of people to live and, each year, millions more to relax and play. However, in the face of this allure fly the strain and consequence of overuse. Increasingly, the imperative to find ways to effectively manage this fragile, finite resource compels science to learn more about this complex, multifaceted system. It is through such understanding that the best hope for effective management lies. In this paper, we look at ways in which researchers study the shore. Specifically, we'll examine this active corner of geomorphology as seen through the lens of the geoscientist, with particular focus on two geomorphic features: the shoreline and the beach. Further emphasis is placed on how investigators have historically, and until today, applied concepts, tools, and methods borrowed from the spatial sciences and, in more recent times, geographic information technologies to the study of the shoreline and the beach. We begin this exploration with a first principal: the definition of shoreline-the boundary where land and water meet. Next, we examine ways in which researchers over the years have worked to generate a suitable shoreline analog or proxy for study. We then follow with a look at how shoreline position is measured-an area where much recent research attention has been placed. Taking matters a step further, we explore the beach, and beach change in three dimensions, looking at how investigators are using geospatial technologies to characterize and analyze change. Finally, we look at various ways in which researchers predict shore and beach changes, by combining spatial analysis and technology with numerical and statistical models. © 2012 The Author. Geography Compass © 2013 Blackwell Publishing Ltd [ABSTRACT FROM AUTHOR]

Current laser scanning (Lidar, light detection and ranging) technologies span a wide range of survey extent and resolutions, from regional airborne Lidar mapping and terrestrial Lidar field surveys to laboratory systems utilizing indoor three-dimensional (3D) laser scanners. Proliferation in Lidar technology and data collection enables new approaches for monitoring and analysis of landscape evolution. For example, repeat Lidar surveys that generate a time series of point cloud data provide an opportunity to transition from traditional, static representations of topography to terrain abstraction as a 3D dynamic layer. Three case studies are presented to illustrate novel techniques for landscape evolution analysis based on time series of Lidar data: (1) application of multi-year airborne Lidar surveys to a study of a dynamic coastal region, where the change is driven by eolian sediment transport, wave-induced beach erosion, and human intervention; (2) monitoring of vegetation growth and the impact of landscape structure on overland flow in an agricultural field using terrestrial laser scanning; and (3) investigation of landscape design impacts on overland water flow and other physical processes using a tangible geospatial modeling system. The presented studies demonstrate new insights into landscape evolution in different environments that can be gained from Lidar scanning spanning 1.0-0.001 m resolutions with geographic information system analysis capabilities. [ABSTRACT FROM AUTHOR]

The deep convolutional neural network (DCNN) has recently been applied to the highly challenging and ill-posed problem of single image super-resolution (SISR), which aims to predict high-resolution (HR) images from their corresponding low-resolution (LR) images. In many remote sensing (RS) applications, spatial resolution of the aerial or satellite imagery has a great impact on the accuracy and reliability of information extracted from the images. In this study, the potential of a DCNN-based SISR model, called enhanced super-resolution generative adversarial network (ESRGAN), to predict the spatial information degraded or lost in a hyper-spatial resolution unmanned aircraft system (UAS) RGB image set is investigated. ESRGAN model is trained over a limited number of original HR (50 out of 450 total images) and virtually-generated LR UAS images by downsampling the original HR images using a bicubic kernel with a factor × 4 . Quantitative and qualitative assessments of super-resolved images using standard image quality measures (IQMs) confirm that the DCNN-based SISR approach can be successfully applied on LR UAS imagery for spatial resolution enhancement. The performance of DCNN-based SISR approach for the UAS image set closely approximates performances reported on standard SISR image sets with mean peak signal-to-noise ratio (PSNR) and structural similarity (SSIM) index values of around 28 dB and 0.85 dB, respectively. Furthermore, by exploiting the rigorous Structure-from-Motion (SfM) photogrammetry procedure, an accurate task-based IQM for evaluating the quality of the super-resolved images is carried out. Results verify that the interior and exterior imaging geometry, which are extremely important for extracting highly accurate spatial information from UAS imagery in photogrammetric applications, can be accurately retrieved from a super-resolved image set. The number of corresponding keypoints and dense points generated from the SfM photogrammetry process are about 6 and 17 times more than those extracted from the corresponding LR image set, respectively. [ABSTRACT FROM AUTHOR]

Deep learning has already been proved as a powerful state-of-the-art technique for many image understanding tasks in computer vision and other applications including remote sensing (RS) image analysis. Unmanned aircraft systems (UASs) offer a viable and economical alternative to a conventional sensor and platform for acquiring high spatial and high temporal resolution data with high operational flexibility. Coastal wetlands are among some of the most challenging and complex ecosystems for land cover prediction and mapping tasks because land cover targets often show high intra-class and low inter-class variances. In recent years, several deep convolutional neural network (CNN) architectures have been proposed for pixel-wise image labeling, commonly called semantic image segmentation. In this paper, some of the more recent deep CNN architectures proposed for semantic image segmentation are reviewed, and each model's training efficiency and classification performance are evaluated by training it on a limited labeled image set. Training samples are provided using the hyper-spatial resolution UAS imagery over a wetland area and the required ground truth images are prepared by manual image labeling. Experimental results demonstrate that deep CNNs have a great potential for accurate land cover prediction task using UAS hyper-spatial resolution images. Some simple deep learning architectures perform comparable or even better than complex and very deep architectures with remarkably fewer training epochs. This performance is especially valuable when limited training samples are available, which is a common case in most RS applications. [ABSTRACT FROM AUTHOR]

Flood modeling and analysis has been a vital research area to reduce damages caused by flooding and to make urban environments resilient against such occurrences. This work focuses on building a framework to simulate and visualize flooding in 3D using position-based fluids for real-time flood spread visualization and analysis. The framework incorporates geographical information and takes several parameters in the form of friction coefficients and storm drain information, and then uses mechanics such as precipitation and soil absorption for simulation. The preliminary results of the river flooding test case were satisfactory, as the flood extent was reproduced in 220 s with a difference of 7%. Consequently, the framework could be a useful tool for practitioners who have information about the study area and would like to visualize flooding using a particle-based approach for real-time particle tracking and flood path analysis, incorporating precipitation into their models. [ABSTRACT FROM AUTHOR]

Dense three-dimensional (3D) point cloud data sets generated by Terrestrial Laser Scanning (TLS) and Unmanned Aircraft System based Structure-from-Motion (UAS-SfM) photogrammetry have different characteristics and provide different representations of the underlying land cover. While there are differences, a common challenge associated with these technologies is how to best take advantage of these large data sets, often several hundred million points, to efficiently extract relevant information. Given their size and complexity, the data sets cannot be efficiently and consistently separated into homogeneous features without the use of automated segmentation algorithms. This research aims to evaluate the performance and generalizability of an unsupervised clustering method, originally developed for segmentation of TLS point cloud data in marshes, by extending it to UAS-SfM point clouds. The combination of two sets of features are extracted from both datasets: "core" features that can be extracted from any 3D point cloud and "sensor specific" features unique to the imaging modality. Comparisons of segmented results based on producer's and user's accuracies allow for identifying the advantages and limitations of each dataset and determining the generalization of the clustering method. The producer's accuracies suggest that UAS-SfM (94.7%) better represents tidal flats, while TLS (99.5%) is slightly more suitable for vegetated areas. The users' accuracies suggest that UAS-SfM outperforms TLS in vegetated areas with 98.6% of those points identified as vegetation actually falling in vegetated areas whereas TLS outperforms UAS-SfM in tidal flat areas with 99.2% user accuracy. Results demonstrate that the clustering method initially developed for TLS point cloud data transfers well to UAS-SfM point cloud data to enable consistent and accurate segmentation of marsh land cover via an unsupervised method. [ABSTRACT FROM AUTHOR]

Digital elevation models (DEMs) have become ubiquitous and remarkably effective in the field of earth sciences as a tool to characterize surface topography. All DEMs have a degree of inherent error and uncertainty that is propagated to subsequent models and analyses, which can lead to misinterpretation and inaccurate estimates. A new method was developed to estimate local DEM errors and implement corrections while quantifying the uncertainties of the implemented corrections. The method is based on the flexibility and ability to model complex problems with ensemble neural networks (ENNs). The method was developed to be applied to any DEM created from a corresponding set of elevation points (point cloud) and a set of ground truth measurements. The method was developed and tested using hyperspatial resolution terrestrial laser scanning (TLS) data (sub-centimeter point spacing) collected from a marsh site located along the southern portion of the Texas Gulf Coast, USA. ENNs improve the overall DEM accuracy in the study area by 68% for six model inputs and by 75% for 12 model inputs corresponding to root mean square errors (RMSEs) of 0.056 and 0.045 m, respectively. The 12-input model provides more accurate tolerance interval estimates, particularly for vegetated areas. The accuracy of the method is confirmed based on an independent data set. Although the method still underestimates the 95% tolerance interval, 8% below the 95% target, results show that it is able to quantify the spatial variability in uncertainties due to a relationship between vegetation/land cover and accuracy of the DEM for the study area. There are still opportunities and challenges in improving and confirming the applicability of this method for different study sites and data sets. [ABSTRACT FROM AUTHOR]

Recent deep-learning counting techniques revolve around two distinct features of data—sparse data, which favors detection networks, or dense data where density map networks are used. Both techniques fail to address a third scenario, where dense objects are sparsely located. Raw aerial images represent sparse distributions of data in most situations. To address this issue, we propose a novel and exceedingly portable end-to-end model, DisCountNet, and an example dataset to test it on. DisCountNet is a two-stage network that uses theories from both detection and heat-map networks to provide a simple yet powerful design. The first stage, DiscNet, operates on the theory of coarse detection, but does so by converting a rich and high-resolution image into a sparse representation where only important information is encoded. Following this, CountNet operates on the dense regions of the sparse matrix to generate a density map, which provides fine locations and count predictions on densities of objects. Comparing the proposed network to current state-of-the-art networks, we find that we can maintain competitive performance while using a fraction of the computational complexity, resulting in a real-time solution. [ABSTRACT FROM AUTHOR]