期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2018
卷号：6
期号：2
页码：771
DOI：10.15680/IJIRCCE.2018.0602015
出版社：S&S Publications
摘要：Modernization of infrastructure mainly depends on the perfection of how the ecological entities such aslandscapes and the water bodies are laid out. With the development of Infrastructure of Smart City (ISC), IntelligentTransportation Systems (ITS) and the degradation of agriculture, land features are changing frequently. Featureextraction and Analysis of these landscapes from remote sensing imagery depend largely on the characteristics of SARImages. A recognition method for remote sensing imagery using the hybrid method of Entropy Decomposition andSupport Vector Machine (EDSVM) is proposed to handle the processing of images efficiently to maximum limit. Theabove given classifier demonstrates the advantages of the valuable decomposed parameters and statistical machinelearning theory in performing better results compared with the standalone SVM classifier. All the pixel patterns aregrouped under prominently exposed color patterns such as red, blue, green which are indicated through differentluminance exposure changes in the considered image. Each of these grouped information are cross verified withreference datasets to provide relevance of matching ability to that of the images that are in it. Segregation ofdissimilarities in the image pattern are driven out which defines the proportionality between the actual image patternand the content regarding to the original surface area. Those depictions help us in identifying and acquiring thestatistics about the survey over the land through the images.
