DETECTING THE ENVIRONMENTAL CHANGES FROM SATELLITE IMAGE

摘要

The goal of this paper is detecting the e nvironmental changes that could affect the globe in the future conditions like pollution, global climate change, natural resource management, urban growth, and much more and trend across large geographic areas from a subset of a Landsat Thematic Mapper (TM) multispec tralimage touse in GIS. In this paper, Environmental Changes are detected from both low and high-resolution satellite images. Some of the detectable features are extracted from blur or noisy prev ious-date satellite images. Firstly, blur images are prepro cessed and restored using complex wavelet packets deconvolution algorithm that show the estimation correctly only if it is made from the original image and gives an approximation of sufficiently good quality needed to com pute to provide useful estimation results if the image is unknown. Both blur or noisy im age and present-date image are processed with color segmentation techniques, thresholding and feature extraction for detecting the environmental changing of regions with image differencing technique and then extracting to compute the percentage of changes an d field seene from satellite images.