Abhishek Narvaria, U. Kumar, Kanumuru Shree Jhanwwee, A. Dasgupta, Gurdeep Kaur
{"title":"Classification and Identification of Crops Using Deep Learning with UAV Data","authors":"Abhishek Narvaria, U. Kumar, Kanumuru Shree Jhanwwee, A. Dasgupta, Gurdeep Kaur","doi":"10.1109/InGARSS51564.2021.9792009","DOIUrl":"https://doi.org/10.1109/InGARSS51564.2021.9792009","url":null,"abstract":"Agriculture is a source of livelihood for the majority of Indian population. The current agricultural problems include timely estimation of crop yield, detection of crop disease, and assessment of damages due to natural disasters such as flood, drought, etc. Any proposed solution to these problems will require accurate classification of crops. In this perspective, UAV (unmanned aerial vehicle) data are proving to be crucial in providing images of the land surface in multispectral channels such as blue (B), green (G), red (R), near infrared, red edge and long wave infrared thermal bands. Once the UAV images have been obtained, segmentation and classification of crops can be performed using deep learning techniques. The objectives of this work is to apply U-Net convolutional neural network (CNN) architecture to (i) find the best combination of input spectral bands for various crop classification such as wheat, cotton, maize, grass, and soil, (ii) to analyse the role of including vegetation indices (NDVI and EVI) in crop classification, and (iii) to analyse the role of textural parameters in performing semantic segmentation of the multiple crops in the scene. Average accuracy of the trained model was 83.35% with RGB bands and 74.61% with CIR (Colour Infrared) composite images. Results indicated that inclusion of NDVI in CIR input dataset yield high segmentation accuracy of 83.85% while dealing with the five classes’ separation problem.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117204588","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"InGARSS 2021 Cover Page","authors":"","doi":"10.1109/ingarss51564.2021.9792077","DOIUrl":"https://doi.org/10.1109/ingarss51564.2021.9792077","url":null,"abstract":"","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127464555","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Identification of Bark Beetle Infestation in Part of Bohemian Forest using Sentinel-1 Time Series Insar","authors":"Mohammed Sultan Alshayef, M. Musthafa","doi":"10.1109/ingarss51564.2021.9792116","DOIUrl":"https://doi.org/10.1109/ingarss51564.2021.9792116","url":null,"abstract":"Continuous monitoring of forests by earth observation satellites has provided information on forest disturbance and aid in strategizing conservation measures. This study attempts to identify bark beetle insect infestation in the forest using time series Synthetic Aperture Radar (SAR) interferometric coherence. The analyses shows lower interferometric SAR coherence values for regions infested by bark beetle compared to the forest regions not infested.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126108778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Kousik Biswas, R. Panda, Pavan Kumar M., D. Chakravarty, Pabitra Mitra, A. Misra, Dibyendu Ghosh, Prosenjit Banerjee
{"title":"Estimation of Ground Displacement in San Francisco Bay Area Using a Spatio-Temporal Unwrapping Network Based Psinsar Algorithm","authors":"Kousik Biswas, R. Panda, Pavan Kumar M., D. Chakravarty, Pabitra Mitra, A. Misra, Dibyendu Ghosh, Prosenjit Banerjee","doi":"10.1109/ingarss51564.2021.9792114","DOIUrl":"https://doi.org/10.1109/ingarss51564.2021.9792114","url":null,"abstract":"For many years, ground surface deformations have been monitored using the DInSAR technique. Because of some major drawbacks, this method has been taken over by a more precise remote sensing mechanism, known as PSInSAR (Persistent Scatterer Interferometric SAR). The latter approach proves helpful as it offers better precision, noise reduction in wider coverage of the geographical area when compared with DInSAR. In this work, we have estimated the deformation rate and topographic error of the San Francisco Bay area using the STUN (Spatio-Temporal Unwrapping Network) algorithm with the pre-selected PS points. Further, the outliers were removed using the squared-norm of least-square residuals. Finally, we validate the estimated deformation rate with the results of StaMPS (Stanford Method for Persistent Scatterer), and they are in agreement with minimal discrepancies (mean 0.9, standard deviation = 2.8 mm/y).","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"128 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122906794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Manjunath B. Ujjinakoppa, U. Kumar, Rahisha Thottolil, A. Dasgupta
{"title":"Multimodal and Multitemporal Spatial Data Analysis in Google Earth Engine Cloud Computing Platform to Detect Human Settlements Without Electricity: A Case Study of Bangalore City","authors":"Manjunath B. Ujjinakoppa, U. Kumar, Rahisha Thottolil, A. Dasgupta","doi":"10.1109/ingarss51564.2021.9792041","DOIUrl":"https://doi.org/10.1109/ingarss51564.2021.9792041","url":null,"abstract":"Accurate information about the human settlements without electricity is essential for monitoring the areas deprived of access to electricity and to end the darkness. Motivated by the 2021 IEEE GRSS Data Fusion Contest organized by the Image Analysis and Data Fusion Technical Committee of the IEEE Geoscience and Remote Sensing Society (GRSS), the objective of this research was to assess the human settlements without electricity for areas in and around Bangalore City. We used multimodal and multitemporal data of the year 2019 with 27 layers such as Landsat-8 OLI bands, Sentinel-1 C Band (SAR data) with VV and VH polarization, spectral indices (EVI, NDVI, MNDWI, NDBI, NDMI, BSI, SAVI, IBI, BuEI and SoEI), Texture parameters (DISS, Entropy and Angular Second Moment), Topological data (slope and elevation), and land surface temperature to detect land use map with urban builtup, vegetation, water and barren land classes with a spatial resolution of 30 m using object-based Random Forest algorithm. To overcome the computational limitations, all the analyses were carried out in Google Earth Engine (GEE) cloud-based platform that has planetary-scale analysis capabilities. Overall, 39 experiments on classification were carried out with various combinations of feature vectors to obtain the most accurate land use map with 4 classes. Composite of Landsat bands and advantages of other spectral indices with thresholds rendered the highest classification accuracy of 93.49%. The final mapping results of human settlements without electricity was obtained by comparing binary classified maps with resampled VIIRS night-time light imagery. The results revealed that the total area of human settlements without electricity in Bangalore City is approximately 36.57 sq. km. accounting for 6.2% of the total study area.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129561076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
K. Mallick, Tian Hu, Yun Bai, Nishan Bhattarai, I. Trebs, M. Schlerf, G. Boulet, Tianxing Wang, C. R. Sanchez, R. Shortt, D. Baldocchi
{"title":"Thermal and Shortwave Infrared Remote Sensing of Ecosystem Processes: Opportunities, Synergies, and Challenges","authors":"K. Mallick, Tian Hu, Yun Bai, Nishan Bhattarai, I. Trebs, M. Schlerf, G. Boulet, Tianxing Wang, C. R. Sanchez, R. Shortt, D. Baldocchi","doi":"10.1109/InGARSS51564.2021.9791903","DOIUrl":"https://doi.org/10.1109/InGARSS51564.2021.9791903","url":null,"abstract":"Land surface temperature (LST) is one of the preeminent state variables that affects the ecosystem functioning and it also gets modulated due to biotic and abiotic ecosystem processes. Understanding the factors controlling LST and the impacts of LST on the ecosystem water-carbon exchanges are very often hindered due to (1) the effects of LST retrieval uncertainty and angular variation on different processes, (2) the lack of authentic physically-based models that link LST with ecosystem physiological properties. Moreover, the synergy of LST and water-sensitive shortwave infrared (SWIR) bands in ecosystem modelling is absent until now, which could enhance the application of LST. The planned Indo-French polar-orbiting mission TRISHNA (Thermal infraRed Imaging Satellite for High-Resolution Natural resource Assessment) with its unprecedented spatial resolution and temporal revisit (57 m and 3 days noon-night observation capacity), 6 VSWIR and 4 TIR Bands in 8 – 12 m, will bring new insights to overcome these challenges. The optical-thermal observations will be used to attain LST, emissivity and surface reflectance with high accuracy and precision, underlying future scientific breakthroughs in understanding the energy-water-carbon exchange process. This paper briefly describes an overview of challenges in estimation of LST, role of thermal directionality and two of the most important ecosystem processes (evaporation and photosynthesis). The utility of SWIR reflectance in tandem with LST as a ‘signal of opportunity’ is also described.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127966150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rajat, B. Nela, Gulab Singh, Virendra Singh Rathore, A. Glazovsky
{"title":"Retrieval of Mass Balance of Austre Grønfjordbreen in the Western Svalbard","authors":"Rajat, B. Nela, Gulab Singh, Virendra Singh Rathore, A. Glazovsky","doi":"10.1109/InGARSS51564.2021.9792005","DOIUrl":"https://doi.org/10.1109/InGARSS51564.2021.9792005","url":null,"abstract":"In this study, the geodetic method was used to infer glacier mass balance, and its spatial variation over glaciers, in the part of Svalbard (Arctic).The study proposed an adaptive mass balance inversion algorithm developed using different densities for ice and firn. Firn identification was achieved through the multi-temporal SAR approach. The performance of the mass balance retrieval algorithm was validated with the ground-based measured mass balance over the Austre Grønfjordbreen, Svalbard. Using elevation rate of change and density (900 ±10 kg/m3) for ice and firn (600 ±200 kg/m3) areas, mass balance is estimated. The mean of geodetic mass balance is found to be -1.47 ± 0.01 m w. e. a-1 with constant density and -1.40 ± 0.06 m w. e. a-1 with the consideration of adaptive density for ice and firn area and then compared with field data (-1.38 m w. e. a-1). The mean mass balance of the proposed method is near to the field based calculated mass balance.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132468885","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A Novel Statistical Preprocessing Approach for Hyperspectral Image Unmixing","authors":"Fatemeh Kowkabi, A. Keshavarz, L. Kumar","doi":"10.1109/InGARSS51564.2021.9792139","DOIUrl":"https://doi.org/10.1109/InGARSS51564.2021.9792139","url":null,"abstract":"Hyperspectral images have wide application in different commercial and military missions. However, these images suffer from the low spatial resolution, and hence, spectral unmixing has been proposed as a suitable technique to solve this deficiency. Recently, some preprocessing modules have been proposed which use the combination of spatial and spectral information of image pixels with the purpose of improving endmember extraction and spectral unmixing in hyperspectral images. In this paper, we propose a statistical preprocessing algorithm using weighted extended cumulative entropy (WECE) as the Edge Detection algorithm to identify spectrally pure pixels and KL distance estimation using the mean and covariance values of pure oversegments and their neighbors to find the spatially homogenous and spectrally pure pixels for the next endmember extraction algorithms. Moreover, the average of KL distances are applied due to the reduction of noise power and local spectral variability. The performance of the novel statistical preprocessing algorithm has been compared with the sate-of-the-art in terms of RMSE reconstruction and average SAD errors on two real hyperspectral images known as HYDICE urban and AVIRIS Indian Pines in combination with several endmember extraction (EE) algorithms and its efficiency is evaluated.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134399482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Fusion of Low-Cost UAV Point Cloud with TLS Point Cloud for Complete 3D Visualisation of a Building","authors":"I. Chauhan, Alok Rawat, Mps Chauhan, RD Garg","doi":"10.1109/InGARSS51564.2021.9792104","DOIUrl":"https://doi.org/10.1109/InGARSS51564.2021.9792104","url":null,"abstract":"3D modelling of buildings is an important task for getting geometrical knowledge of the building as well as for planning and monitoring purposes. The use of low-cost UAV (Unmanned Aerial Vehicle) derived point cloud is becoming a popular way for 3D model creation. To enhance the accuracy of such models generally researchers employ GCP’s(Ground Control Points), based on GNSS survey to georeference the point cloud. But using GCP’s has limitations like object’s accessibility & also it is time consuming. Another way to increase the accuracy of such point clouds is merging them with TLS(Terrestrial Laser Scanner) point cloud which is georeferenced using a geodetic GNSS. The research here is trying to achieve the same by fusing point clouds derived from low-cost UAV (Phantom 4 pro v.2) and TLS. This is achieved in 3 steps a) Metric based comparison b) C2C(Cloud to Cloud) & M3C2 (Multi scale Model to Model Cloud Comparison) algorithm-based comparison c) running linear regression for selected points on both the point clouds. An RMS(Root Mean Square) of 0.125 is achieved for Phantom pro metric-based comparison. The C2C & M3C2 algorithm-based comparison shows that most points in the facade of the building have almost zero error i.e., the point clouds are totally identical in these areas. Also, linear regression comparison gives very high R-squared value affirming the fact that point clouds obtained from UAV highly correlate to that obtained from TLS. Thus, based on above 3 comparison methods, it can be concluded that a UAV point cloud could be fused with a single view GNSS referenced TLS point cloud to obtain a complete 3D visualization of building.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124537316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Use of Shannon Information Entropy in Earthquake Nowcasting","authors":"S. Pasari, Priyesh Agarwal, N. Neha","doi":"10.1109/ingarss51564.2021.9792131","DOIUrl":"https://doi.org/10.1109/ingarss51564.2021.9792131","url":null,"abstract":"Earthquake nowcasting uses cumulative counts of interevent small earthquakes (say, 3.0≤M<6.0) between successive large earthquakes (M≥6.0) to estimate the current dynamic state of a geological fault system. As Shannon information entropy (SIE) can suitably quantify the amount of information preserved in a dataset, the present study implements the concept of SIE to the earthquake fault system to examine its efficacy in the computation of earthquake potential score (EPS). The EPS provides a unique measure to determine the current level of earthquake cycle progression in a defined area. Earthquake data from a large region in Southeast Asia is considered for illustration. A comparison of EPS obtained from conventional natural-time analysis and SIE-based natural time analysis shows that SIE has minor contribution in seismic nowcasting analysis.","PeriodicalId":260112,"journal":{"name":"2021 IEEE International India Geoscience and Remote Sensing Symposium (InGARSS)","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114325825","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}