Open AccessDetection of quasi-circular vegetation patches using GF-2 image with tasseled cap and watershed transformations
Author(s) -
Qingsheng Liu
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/768/6/062053
Subject(s) - watershed , hue , artificial intelligence , remote sensing , segmentation , image segmentation , computer science , decision tree , change detection , ground truth , transformation (genetics) , vegetation (pathology) , environmental science , pattern recognition (psychology) , computer vision , geography , medicine , pathology , biochemistry , chemistry , gene
It is a key to detect the quasi-circular vegetation patches (QVPs) for studying the establishment and encroachment mechanisms of the QVPs in the Yellow River Delta, China. A variety of spatial resolution remote sensing data have been used to map the QVPs. However, the adhesion between the QVPs with the QVPs or the vegetations of other shape makes the detection accuracy of the QVPs unsatisfactory. This study applied the decision tree classifier to map the QVPs using the brightness and greenness components of the modified intensity-hue-saturation pansharpened Gaofen 2 imagery. Then, the watershed transformation was used to segment the classification result. The final result was obtained using the thresholds of statistical features of the QVPs. It indicated that the method of this work could well detect the QVPs. In the future, more effective image segmentation algorithms should be used to deal with the over-segmentation in order to further improve the detection accuracy.