Urban Vegetation Change Detection based on Image Information Enhancement

Abstract

In this paper, a kind of urban vegetation change detection method based on image information enhancement is proposed for the problem of error detection caused by the haze and shadow in vegetation change detection of urban high-resolution remote sensing images. Firstly, conduct haze removal on images through HSV transformation; then, carry on vegetation shadow compensation by PESR method; next, perform the vegetation information post-processing by using SNDS method; finally, obtain the information on vegetation change by using difference method. The vegetation change detection on WorldView-2 image and Ikonos image of Xi'an City shows that the average precision reaches 80.81%. The experimental results indicate that the proposed method in this paper has brought about a certain improvement of vegetation identification rate in the high-resolution image and a significant decrease in missing extraction of vegetation information in shadow area, thus effectively improving the accuracy of vegetation change detection.