Low Resolution Satellite Research Articles

Using very‐high‐resolution satellite imagery and deep learning to detect and count African elephants in heterogeneous landscapes

AbstractSatellites allow large‐scale surveys to be conducted in short time periods with repeat surveys possible at intervals of <24 h. Very‐high‐resolution satellite imagery has been successfully used to detect and count a number of wildlife species in open, homogeneous landscapes and seascapes where target animals have a strong contrast with their environment. However, no research to date has detected animals in complex heterogeneous environments or detected elephants from space using very‐high‐resolution satellite imagery and deep learning. In this study, we apply a Convolution Neural Network (CNN) model to automatically detect and count African elephants in a woodland savanna ecosystem in South Africa. We use WorldView‐3 and 4 satellite data –the highest resolution satellite imagery commercially available. We train and test the model on 11 images from 2014 to 2019. We compare the performance accuracy of the CNN against human accuracy. Additionally, we apply the model on a coarser resolution satellite image (GeoEye‐1) captured in Kenya, without any additional training data, to test if the algorithm can generalize to an elephant population outside of the training area. Our results show that the CNN performs with high accuracy, comparable to human detection capabilities. The detection accuracy (i.e., F2 score) of the CNN models was 0.78 in heterogeneous areas and 0.73 in homogenous areas. This compares with the detection accuracy of the human labels with an averaged F2 score 0.77 in heterogeneous areas and 0.80 in homogenous areas. The CNN model can generalize to detect elephants in a different geographical location and from a lower resolution satellite. Our study demonstrates the feasibility of applying state‐of‐the‐art satellite remote sensing and deep learning technologies for detecting and counting African elephants in heterogeneous landscapes. The study showcases the feasibility of using high resolution satellite imagery as a promising new wildlife surveying technique. Through creation of a customized training dataset and application of a Convolutional Neural Network, we have automated the detection of elephants in satellite imagery with accuracy as high as human detection capabilities. The success of the model to detect elephants outside of the training data site demonstrates the generalizability of the technique.

Unsupervised Video Satellite Super-Resolution by Using Only a Single Video

Recent studies have shown that deep-learning (DL)-based methods lead to improved performance in video satellite super-resolution (SR). However, the vast majority of prior work is supervised, which is restricted to artificially generated training data (e.g., predetermined bicubic downsampling). Unfortunately, in the real world, the low-resolution (LR) satellite video frames rarely obey these restrictions. To solve this problem, we resort to unsupervised learning and propose a video satellite SR method by using only a single video. The single video SR (SingleVSR) method takes advantage of the power of DL without relying on prior high-resolution (HR) and LR pairs. In the training phase, the LR frames are alternately processed by both downsampling network (i.e., Net <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LR</sub> ) and upsampling network (i.e., Net <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">HR</sub> ). The losses obtained by LR frames and network outputs are used to optimize both Net <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">LR</sub> and Net <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">HR</sub> . In the testing phase, the trained Net <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">HR</sub> is applied to generate the SR results of LR frames. In contrast to the existing video satellite SR methods, our SingleVSR does not require any assumption on degradation or any additional training data except for the single video to be tested. Experiments performed on Jilin-1 and OVS-1 satellite videos demonstrate the superiority of the proposed method.

Providing Satellite Data to R/V Akademik Mstislav Keldysh during a Mission in the Southern Ocean

During cruise 79 of R/V Akademik Mstislav Keldysh in the Atlantic Antarctic, information is supplied using optical and radar high-, mid-, and low-resolution satellite data, in combination with geoinformation technologies. During field studies from December 2019 to February 2020, preliminarily processed information obtained by a multisatellite Earth remote sensing (ERS) system on sea surface temperature, sea ice concentration, icebergs, altimetry, and the concentration of near-surface layer matter (chlorophyll, particulate organic carbon, inorganic suspended particulate matter, and total suspended matter) is transmitted to the vessel. A description is given of the information infrastructure that provides the research vessel with routine satellite data that is automatically integrated into the user’s geoinformation environment as unified spatiotemporal distributions of important water environment parameters. Aspects of using satellite information when planning a ship’s route and making measurements are described.

Bed topography of Princess Elizabeth Land in East Antarctica

Abstract. We present a topographic digital elevation model (DEM) for Princess Elizabeth Land (PEL), East Antarctica. The DEM covers an area of ∼900 000 km2 and was built from radio-echo sounding data collected during four campaigns since 2015. Previously, to generate the Bedmap2 topographic product, PEL's bed was characterized from low-resolution satellite gravity data across an otherwise large (&gt;200 km wide) data-free zone. We use the mass conservation (MC) method to produce an ice thickness grid across faster flowing (&gt;30 m yr−1) regions of the ice sheet and streamline diffusion in slower flowing areas. The resulting ice thickness model is integrated with an ice surface model to build the bed DEM. Together with BedMachine Antarctica and Bedmap2, this new bed DEM completes the first-order measurement of subglacial continental Antarctica – an international mission that began around 70 years ago. The ice thickness data and bed DEMs of PEL (resolved horizontally at 500 m relative to ice surface elevations obtained from the Reference Elevation Model of Antarctica – REMA) are accessible from https://doi.org/10.5281/zenodo.4023343 (Cui et al., 2020a) and https://doi.org/10.5281/zenodo.4023393 (Cui et al., 2020b).

Satellite Imagery Classification with Deep Learning : A Survey

Object detection from satellite images has been a challenging problem for many years. With the development of effective deep learning algorithms and advancement in hardware systems, higher accuracies have been achieved in the detection of various objects from very high-resolution satellite images. In the past decades satellite imagery has been used successfully for weather forecasting, geographical and geological applications. Low resolution satellite images are sufficient for these sorts of applications. But the technological developments in the field of satellite imaging provide high resolution sensors which expands its field of application. Thus, the High-Resolution Satellite Imagery (HRSI) proved to be a suitable alternative to aerial photogrammetric data to provide a new data source for object detection. Since the traffic rates in developing countries are enormously increasing, vehicle detection from satellite data will be a better choice for automating such systems. In this research, a different technique for vehicle detection from the images obtained from high resolution sensors is reviewed. This review presents the recent progress in the field of object detection from satellite imagery using deep learning.

Combined Model Color-Correction Method Utilizing External Low-Frequency Reference Signals for Large-Scale Optical Satellite Image Mosaics

Optical satellites are affected by factors such as seasonal and atmospheric variation, illumination, and sensor distortion. Thus, satellite images covering large-scale area often show conspicuous color differences, resulting in poor color continuity of the mosaicked satellite image. This study proposes a novel combined model color correction (CMCC) method for high-resolution optical satellite images, which constructively combines a defogging model with a radiation correction model. First, this study analyzed the feasibility of using easily available low-resolution satellite images as external references to correct the color of high-resolution images and describes the selection criteria for external references. Second, considering the negative effects of atmosphere on the color and clarity of remote sensing images, we proposed an optical satellite image enhancement method, which is based on the content characteristics of remote sensing images and the dark channel prior defogging method. Finally, we designed a two-stage color correction process: 1) correcting the color of downsampled images via low-frequency modeling and replacement and 2) mapping the color of downsampled images to original images through local modeling and super-resolution color correction. Furthermore, this study proposes an indicator of quality considered mean absolute error (QCMAE) for quantitative evaluation of the color correction result. We selected 328 Gaofen-1 (GF-1) high-resolution images for the experiments. Visual effects and statistical results of images after being processed by the proposed CMCC are both superior to the three state-of-the-art methods, which verifies the effectiveness and reliability of the proposed method.

A paradigm of GIS and remote sensing for crop water deficit assessment in near real time to improve irrigation distribution plan

This study comprehensively used Remote Sensing (RS) and GIS for independently monitoring crop water adequacy in response to existing irrigation management system in Bari Doab, Pakistan. Canals’ rotation plan is developed before start of each cropping season to distribute water on fixed turn basis. The study combines the moderate (Landsat-8) and low resolution (MODIS) satellite imagery to assess near real time crop water deficit in canal command areas (CCAs) of the study area. Landsat imagery were used to develop crop classification and determine cropping pattern using time series of NDVI. MODIS’ NDVI product was employed to monitor the current crop health and develop annual reference NDVI cycle based on last 13 years’ imagery. Climatic data of WMO stations combined with DEM have also been used to determine reference evapotranspiration based on Penman-Monteith method. These inputs were combined to estimate average crop water deficit in CCAs for ongoing cropping season on 08-day interval, the interval parallel to the canal rotation plan. RS based reflectance crop coefficients (KCR) of current and reference crop cycle were developed to calculate potential and actual evapotranspiration, and then translated their difference into crop water deficit. Lag time of 08–10 days is limitation of the study due to satellites overpass time, but it can still prove very helpful considering the crop season spanned over 160–180 days. Results revealed that maximum crop water deficit (up to 9 mm) occurred in CCAs during crop growing (mid) stage. Crop water deficit showed inverse relation when plotted against intense rainfall months. Crop water deficit results could be helpful for irrigation managers to modify canals’ rotation plan in near real time according to crop health and growing stage. It would not only help in achieving equitable water distribution but also assuring supplies when required by the crops.

Comparative Assessment of Vegetation Indices in Downscaling of MODIS Satellite Land Surface Temperature

In thermal remote sensing, the freely available land surface temperature (LST) at high-resolution data is essential. The present study aims to downscale the low-resolution (1000 m) MODIS satellite’s LST data. LST downscaling technique was developed using the statistical relationship between Earth’s surface vegetation indices (VI) and LST. The MODIS satellite’s three fundamental VI [soil-adjusted vegetation index (SAVI), normalized difference vegetation index (NDVI), and fractional vegetation cover (fc)] were taken as supporting regression variables, in the case study of Jaipur city, India. The NDVI was showing the highest correlation coefficient (R2) of 0.72, 0.62, and 0.82 and outperformed SAVI (0.57, 0.39, 0.65) and fc (0.48, 0.43, 0.69), in winter, summer, and monsoon season, respectively. The (15 ≤ CV ≤ 25%) sampled data has shown much higher R2 (0.74, 0.63, and 0.85) compared with full data (0.58, 0.39, and 0.60), in winter, summer, and monsoon season, respectively. The downscaled LST values were validated by surface temperature recorded by thermal data loggers. The physical regression-based downscaling model was able to predict LST accurately up to 200-m resolution, without significant errors (< 1 °C). The downscaled finer resolution LST can be used for applications such as fire detection, thermal comfort monitoring, soil moisture mapping, and detection/visualization of urban centers.

Investigating the Reliability of Irrigated Area Coverages from Fragmented Lands: The case of Tigray, Northern Ethiopia.

Consistent and well-timed information regarding the extent of irrigated areas is very important for the development of the agricultural sector. This study investigated the reliability of traditional estimates of irrigated area coverages from three micro-dam irrigation schemes. High resolution Google Earth (GE) and medium resolution satellite imagery coupled with an intensive ground survey were employed to determine irrigation area coverages. Landsat-8 image pre-processing and classification was carried out in ENVI 5.0 software. In comparison to the Landsat-8 based estimates, the traditional method over-estimated 87% of the total 14 cases considered. The level of discrepancies in scheme area coverages range from a minimum of -20% (underestimated) to a maximum of 224%. Besides this, the traditional estimates were ranging from -1.8 to 238% of GE-based estimates. Whereas, the Landsat-8 based estimates slightly overestimated it by 1.8% -16% of GE-based estimates. On the other hand, the traditional estimates tend to underestimate/ignore crops with smaller area coverages and overestimate crops with relatively larger area coverages. The results hinted that the using medium and low resolution satellite imagery could improve the data collection and reporting from irrigated agriculture in developing countries.

CIRCULAR-SHAPED OBJECT DETECTION IN LOW RESOLUTION SATELLITE IMAGES

Abstract. Circle and ellipse detection have been a widely discussed topic among the computer vision community. Applying circle detection methods to satellite images can bring valuable information on urban or industrial areas. A limitation of such methods lies in the resolution of the object to detect. The smaller they are in the image, the harder it becomes to detect them accurately. In this paper, we explore several circle detection methods adapted to low resolution satellite images. An algorithm based on level-lines detection and classification will be presented in this paper along with other well known algorithms. The methods are evaluated in the context of oil tank detection in Sentinel-2 images where those circular objects can have an observed diameter as small as 2 pixels.

STOCKPILE MONITORING USING LINEAR SHAPE-FROM-SHADING ON PLANETSCOPE IMAGERY

Abstract. The storage and management of stockpiles of materials is a fundamental process in large scale activities such as mining, civil engineering, and in the management of waste landfill sites. Following the evolution of stocks has always been important, and advancements in remote sensing technology are not only facilitating this, but also making it possible in near real-time. Nowadays, this monitoring appears to be performed almost exclusively using UAV based techniques. This paper proposes to apply a simple Shape-from-Shading method on low resolution satellite images provided by PlanetScope to monitor the evolution of the volume of stockpiles. The proposed Shape-from-Shading formulation makes it possible to handle occluding objects in the scene. The loss in accuracy due to the low resolution of PlanetScope images is well compensated by the daily revisit frequency and by the fact that spaceborne acquisitions require no human supervision. With satellites it is also easy to follow simultaneously several stocking sites all over the world. We test our method on two coal storage sites and demonstrate that the stockpiles are well detected.

A Novel Classification Extension-Based Cloud Detection Method for Medium-Resolution Optical Images

Accurate cloud detection using medium-resolution multispectral satellite imagery (such as Landsat and Sentinel data) is always difficult due to the complex land surfaces, diverse cloud types, and limited number of available spectral bands, especially in the case of images without thermal bands. In this paper, a novel classification extension-based cloud detection (CECD) method was proposed for masking clouds in the medium-resolution images. The new method does not rely on thermal bands and can be used for masking clouds in different types of medium-resolution satellite imagery. First, with the support of low-resolution satellite imagery with short revisit periods, cloud and non-cloud pixels were identified in the resampled low-resolution version of the medium-resolution cloudy image. Then, based on the identified cloud and non-cloud pixels and the resampled cloudy image, training samples were automatically collected to develop a random forest (RF) classifier. Finally, the developed RF classifier was extended to the corresponding medium-resolution cloudy image to generate an accurate cloud mask. The CECD method was applied to Landsat-8 and Sentinel-2 imagery to test the performance for different satellite images, and the well-known function of mask (FMASK) method was employed for comparison with our method. The results indicate that CECD is more accurate at detecting clouds in Landsat-8 and Sentinel-2 imagery, giving an average F-measure value of 97.65% and 97.11% for Landsat-8 and Sentinel-2 imagery, respectively, as against corresponding results of 90.80% and 88.47% for FMASK. It is concluded, therefore, that the proposed CECD algorithm is an effective cloud-classification algorithm that can be applied to the medium-resolution optical satellite imagery.

Determination of Water Surface and Vegetation Change in Akkaya Reservoir Basin Using Remote Sensing Method

In the study, the temporal change analysis of Akkaya Reservoir in Niğde was made. The Reservoir was established in 1964. The change of reservoir between 1999, 2009 and 2019 was analyzed using remote sensing method. Satellite images used in the study belong to August (1999, 2009, 2019). MNDWI (Modified Normalized Difference Water Index) analysis was used. With the results of this analysis; the amount of water surfaces in the reservoir basin was calculated by years. After this analysis, NDVI (Normalized Difference Vegetation Index) analysis was used and land cover types were determined by CORINE. It has been determined that the water surface map of 2009 covers a larger area compared to other years. When only the area where the reservoir is located is evaluated by the polynomial equation (2nd degree), there may be a decrease of approximately 26% in the reservoir water surface area within 15 years. As a result of the study, it has been determined that Landsat satellite data can be used to determine the water surface amounts. If study in smaller areas will be done, use the lower resolution satellite images are suggested. In this study, MNDWI analysis was found more successful than NDVI analysis to find water surface area. Multi-band satellite images provide important qualitative information in evaluating the functions used in ecosystem-based planning studies.

Evaluation of the hydraulic and morfometric connections of the riverbed with using GIS

Currently, in the country developed models with using on GIS technologies are mainly based on linear schemes of location and mapping of water resources, and there is no database for determining the state of water resources, hydraulic structures, and river and riverbed deformation. The main problem with the use of GIS models in water management is their low accuracy, the fact that created models were based on low-resolution satellite imagery and used old analytical techniques. Models in irrigation systems which used GIS technologies are one of the problems waiting their solution. In this article analyzed possibility of GIS estimation of deformation processes in the river bed (in case of Sokh River). The processes in the riverbed were analyzed based on satellite data and compared with field experiments. Given conclusions based on analysis.

Aerial Imagery Based on Commercial Flights as Remote Sensing Platform.

Remote sensing is commonly performed via airborne platforms such as satellites, specialized aircraft, and unmanned aerial systems (UASs), which perform airborne photography using mounted cameras. However, they are limited by their coverage (UASs), irregular flyover frequency (aircraft), and/or low spatial resolution (satellites) due to their high altitude. In this paper, we examine the utilization of commercial flights as an airborne platform for remote sensing. Namely, we simulate a situation where all aircraft on commercial flights are equipped with a mounted camera used for airborne photography. The simulation is used to estimate coverage, the temporal and spatial resolution of aerial imagery acquired this way, as well as the storage capacity required for storing all imagery data. The results show that Europe is 83.28 percent covered with an average of one aerial photography every half an hour and a ground sampling distance of 0.96 meters per pixel. Capturing such imagery results in 20 million images or four petabytes of image data per day. More detailed results are given in the paper for separate countries/territories in Europe, individual commercial airlines and alliances, as well as three different cameras.

Efficient agricultural yield prediction using metaheuristic optimized artificial neural network using Hadoop framework

The low-resolution imagery of satellite is used extensively for monitoring crops and forecasting of yield which has a major role to play in the operational systems. A combination of high levels of temporal frequency along with an extended coverage was connected with lower costs per each area unit making the images a choice that is convenient at the national level and the regional level scales. There are various quantitative and qualitative approaches for low-resolution satellite imagery to be used for the primary predictor of the final yield of crops. But, very little work is done on the yield prediction that is based on environmental and satellite data. To handle such satellite images may be very challenging owing to large data amounts. Big data analysis is efficient in handling a large amount of data generated for predicting agricultural yield. In this work, a neural network is used for prediction and to enhance its performance; a population-based incremental learning technique is proposed for optimizing the weights. The results of the experiment proved that the method proposed has better results compared to that of the other methods.

Crop and Weed Leaf Area Index Mapping Using Multi-Source Remote and Proximal Sensing

Site specific management rationalizes farm inputs and mitigate environmental impacts. Traditionally, low resolution satellite imagery and soil maps are employed for site specific decisions in large scale farms. However, these approaches are not good at sub-field level due to low spatial resolution. To overcome this problem, either manual scouting is employed or extensive high resolution data collection platforms are used. In both cases, the cost outweighs the expected returns. Consequently, variable rate applications are not preferred in large fields. Leaf Area Index (LAI) is a useful measure to monitor crop growth and health for site specific management. In this paper we propose an accurate and scalable process where multispectral remote sensing and proximal sensing data is used to estimate LAI. Crop LAI (CLAI) and Weed LAI (WLAI) are estimated from limited high resolution ground image samples using semantic segmentation. These limited LAIs are extended to the whole field using remote sensing and proximal sensing data. We find that LAIs are spatially related with Soil, Water and Topography (SWAT) maps and are field specific. With increasing weed population in the fields, correlation of WLAI with the SWAT zone increases. However, CLAI remains comparatively consistent across SWAT zones due to variable rate seeding and fertilizer application based on soil variance. Our results demonstrate that LAIs can be predicted accurately from limited high resolution ground imagery, satellite imagery, SWAT, and soil properties maps.

Mapping Regional Ecosystem Functional Types Based on Sentinel-2 Satellite Imagery

Ecosystem functional types (EFTs) are the patches of land surface showing similar in carbon dynamics. EFTs are not defined by the structure and composition of vegetation and represent the spatial heterogeneity of ecosystem functions. Identifying EFTs based on low-resolution satellite remote sensing data cannot satisfy the needs of fine-scale characterization of regional ecosystem functional patterns. Here, taking Zhenlai County, Northeast China as an example, the heterogeneity in ecosystem functions was characterized by identifying EFTs from Sentinel-2 time series data using ISODATA algorithm. Ecosystem functional attributes derived from dynamics of the normalized difference vegetation index (NDVI), the fraction of absorbed photosynthetically active radiation (FAPAR), and canopy water content (CWC) in the growing season were calculated. The correspondence analysis (CA) method was used to reveal relationships between the EFTs and land cover types. Our results showed that the nine selected remotely sensed variables indicating carbon and water flux of the regional ecosystems could be adopted in ecosystem functions classification. The obtained EFTs based on Sentinel-2 images reflected the internal structure of carbon balance well and the distribution pattern of ecosystem functional diversity a fine scale. This study helps to understand the functional heterogeneity pattern of temperate terrestrial ecosystems.

Finer Classification of Crops by Fusing UAV Images and Sentinel-2A Data

Accurate crop distribution maps provide important information for crop censuses, yield monitoring and agricultural insurance assessments. Most existing studies apply low spatial resolution satellite images for crop distribution mapping, even in areas with a fragmented landscape. Unmanned aerial vehicle (UAV) imagery provides an alternative imagery source for crop mapping, yet its spectral resolution is usually lower than satellite images. In order to produce more accurate maps without losing any spatial heterogeneity (e.g., the physical boundary of land parcel), this study fuses Sentinel-2A and UAV images to map crop distribution at a finer spatial scale (i.e., land parcel scale) in an experimental site with various cropping patterns in Heilongjiang Province, Northeast China. Using a random forest algorithm, the original, as well as the fused images, are classified into 10 categories: rice, corn, soybean, buckwheat, other vegetations, greenhouses, bare land, water, roads and houses. In addition, we test the effect of UAV image choice by fusing Sentinel-2A with different UAV images at multiples spatial resolutions: 0.03 m, 0.10 m, 0.50 m, 1.00 m and 3.00 m. Overall, the fused images achieved higher classification accuracies, ranging between 10.58% and 16.39%, than the original images. However, the fused image based on the finest UAV image (i.e., 0.03 m) does not result in the highest accuracy. Instead, the 0.10 m spatial resolution UAV image produced the most accurate map. When the spatial resolution is less than 0.10 m, accuracy decreases gradually as spatial resolution decreases. The results of this paper not only indicate the possibility of combining satellite images and UAV images for land parcel level crop mapping for fragmented landscapes, but it also implies a potential scheme to exploit optimal choice of spatial resolution in fusing UAV images and Sentinel-2A, with little to no adverse side-effects.

Arctic greening associated with lengthening growing seasons in Northern Alaska

Many studies have reported that the Arctic is greening; however, we lack an understanding of the detailed patterns and processes that are leading to this observed greening. The normalized difference vegetation index (NDVI) is used to quantify greening, which has had largely positive trends over the last few decades using low spatial resolution satellite imagery such as AVHRR or MODIS over the pan-Arctic region. However, substantial fine scale spatial heterogeneity in the Arctic makes this large-scale investigation hard to interpret in terms of vegetation and other environmental changes. Here we focus on one area of the northern Alaskan Arctic using high spatial resolution (4 m) multispectral satellite imagery from DigitalGlobe™ to analyze the greening trend near Utqiaġvik (formerly known as Barrow) over 14 years from 2002 to 2016. We found that tundra vegetation has been greening (τ = 0.65, p = 0.01, NDVI increase of 0.01 yr−1) despite no overall change in vegetation community composition. The greening is most closely correlated to the number of thawing degree days (R2 = 0.77, F = 21.5, p < 0.001) which increased in a similar linear trend over the 14 year study period (1.79 ± 0.50 days per year, p < 0.01, τ = −0.56). This suggests that in this Arctic ecosystem, greening is occurring due to a lengthening growing season that appears to stimulate plant productivity without any significant change in vegetation communities. We found that vegetation communities in wetter locations greened about twice as fast as those found in drier conditions supporting the hypothesis that these communities respond more strongly to warming. We suggest that in Arctic environments, vegetation productivity may continue to rise, particularly in wet areas.