Accuracy Assessment of different classifiers for Sustainable Development in Landuse and Landcover mapping using Sentinel SAR and Landsat-8 data

Q3 Engineering

EAI Endorsed Transactions on Energy Web Pub Date : 2023-10-13 DOI:10.4108/ew.4141

K. Kanmani, Vasanthi Padmanabhan, P. Pari

{"title":"Accuracy Assessment of different classifiers for Sustainable Development in Landuse and Landcover mapping using Sentinel SAR and Landsat-8 data","authors":"K. Kanmani, Vasanthi Padmanabhan, P. Pari","doi":"10.4108/ew.4141","DOIUrl":null,"url":null,"abstract":"Sentinel satellites make use of Synthetic Aperture Radar (SAR) which produces images with backscattered signals at fine spatial resolution from 10 m to 50 m. This study is mainly focused on evaluating and assessing the accuracy of various supervised classifiers like Random Forest classifier, Minimum Distance to mean classifier, KDTree KNN classifier, and Maximum Likelihood classifier for landuse / landcover mapping in Maduranthakam Taluk, Kancheepuram district, Tamilnadu, India. These classifiers are widely used for classifying the Sentinel SAR images. The SAR images were processed using speckle and terrain correction and converted to backscattered energy. The training datasets for the landcover classes, such as vegetation, waterbodies, settlement, and barren land, were collected from Google Earth images in high-resolution mode. These collected training datasets were given as input for the various classifiers during the classification. The obtained classified output results of various classifiers were analyzed and compared using the overall classification accuracy. The overall accuracy achieved by the Random Forest classifier for the polarization VV and VH was 92.86%, whereas the classified accuracy of various classifiers such as KDTree KNN, Minimum distance to mean, and Maximum Likelihood are found to be 81.68%, 83.17%, and 85.64% respectively. The random forest classifier yields a higher classification accuracy value due to its greater stability in allocating the pixels to the right landuse class. In order to compare and validate the results with sentinel data, the random classifier is applied with optical Landsat-8 satellite data. The classification accuracy obtained for Landsat-8 data is 84.61%. It is clearly proved that the random forest classifier with sentinel data gives the best classification accuracy results due to its high spatial resolution and spectral sensitivity. Thus accurate landuse and landcover mapping promote sustainable development by supporting decision-making at local, regional, and national levels.","PeriodicalId":53458,"journal":{"name":"EAI Endorsed Transactions on Energy Web","volume":"130 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"EAI Endorsed Transactions on Energy Web","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.4108/ew.4141","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"Engineering","Score":null,"Total":0}

引用次数: 0

Abstract

Sentinel satellites make use of Synthetic Aperture Radar (SAR) which produces images with backscattered signals at fine spatial resolution from 10 m to 50 m. This study is mainly focused on evaluating and assessing the accuracy of various supervised classifiers like Random Forest classifier, Minimum Distance to mean classifier, KDTree KNN classifier, and Maximum Likelihood classifier for landuse / landcover mapping in Maduranthakam Taluk, Kancheepuram district, Tamilnadu, India. These classifiers are widely used for classifying the Sentinel SAR images. The SAR images were processed using speckle and terrain correction and converted to backscattered energy. The training datasets for the landcover classes, such as vegetation, waterbodies, settlement, and barren land, were collected from Google Earth images in high-resolution mode. These collected training datasets were given as input for the various classifiers during the classification. The obtained classified output results of various classifiers were analyzed and compared using the overall classification accuracy. The overall accuracy achieved by the Random Forest classifier for the polarization VV and VH was 92.86%, whereas the classified accuracy of various classifiers such as KDTree KNN, Minimum distance to mean, and Maximum Likelihood are found to be 81.68%, 83.17%, and 85.64% respectively. The random forest classifier yields a higher classification accuracy value due to its greater stability in allocating the pixels to the right landuse class. In order to compare and validate the results with sentinel data, the random classifier is applied with optical Landsat-8 satellite data. The classification accuracy obtained for Landsat-8 data is 84.61%. It is clearly proved that the random forest classifier with sentinel data gives the best classification accuracy results due to its high spatial resolution and spectral sensitivity. Thus accurate landuse and landcover mapping promote sustainable development by supporting decision-making at local, regional, and national levels.

查看原文本刊更多论文

基于Sentinel SAR和Landsat-8数据的可持续发展土地利用和土地覆盖制图中不同分类器的精度评估

哨兵卫星使用合成孔径雷达(SAR)，产生10米至50米精细空间分辨率的后向散射信号图像。本研究主要对印度泰米尔纳德邦Maduranthakam Taluk、Kancheepuram地区的土地利用/土地覆盖制图中，随机森林分类器、最小均值距离分类器、KDTree KNN分类器和最大似然分类器等多种监督分类器的准确性进行了评价和评估。这些分类器被广泛用于Sentinel SAR图像的分类。对SAR图像进行散斑和地形校正，并转换为后向散射能量。土地覆盖类别的训练数据集，如植被、水体、聚落和荒地，以高分辨率模式从谷歌地球图像中收集。这些收集的训练数据集在分类过程中作为各种分类器的输入。对各种分类器得到的分类输出结果进行综合分类精度分析和比较。随机森林分类器对极化VV和VH的总体准确率为92.86%，而KDTree KNN、Minimum distance to mean和Maximum Likelihood等分类器的分类准确率分别为81.68%、83.17%和85.64%。由于随机森林分类器在将像素分配到正确的土地利用类别方面具有更高的分类精度值。为了与前哨数据进行比较和验证，将随机分类器应用于Landsat-8光学卫星数据。Landsat-8数据的分类精度为84.61%。结果表明，基于前哨数据的随机森林分类器具有较高的空间分辨率和光谱灵敏度，分类精度最高。因此，准确的土地利用和土地覆盖测绘通过支持地方、区域和国家各级的决策来促进可持续发展。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

EAI Endorsed Transactions on Energy Web Energy-Energy Engineering and Power Technology

CiteScore

2.60

自引率

0.00%

发文量

审稿时长

10 weeks

期刊介绍： With ICT pervading everyday objects and infrastructures, the ‘Future Internet’ is envisioned to undergo a radical transformation from how we know it today (a mere communication highway) into a vast hybrid network seamlessly integrating knowledge, people and machines into techno-social ecosystems whose behaviour transcends the boundaries of today’s engineering science. As the internet of things continues to grow, billions and trillions of data bytes need to be moved, stored and shared. The energy thus consumed and the climate impact of data centers are increasing dramatically, thereby becoming significant contributors to global warming and climate change. As reported recently, the combined electricity consumption of the world’s data centers has already exceeded that of some of the world''s top ten economies. In the ensuing process of integrating traditional and renewable energy, monitoring and managing various energy sources, and processing and transferring technological information through various channels, IT will undoubtedly play an ever-increasing and central role. Several technologies are currently racing to production to meet this challenge, from ‘smart dust’ to hybrid networks capable of controlling the emergence of dependable and reliable green and energy-efficient ecosystems – which we generically term the ‘energy web’ – calling for major paradigm shifts highly disruptive of the ways the energy sector functions today. The EAI Transactions on Energy Web are positioned at the forefront of these efforts and provide a forum for the most forward-looking, state-of-the-art research bringing together the cross section of IT and Energy communities. The journal will publish original works reporting on prominent advances that challenge traditional thinking.