A LAND COVER CHANGE DETECTION METHOD BASED ON CHANGE DIFFERENCE MAP FUSION
Keywords: Change Detection, Remote Sensing Image, Change Difference Map Fusion, Expectation Maximization
Abstract. Direct radiometric comparison-based change detection methods have been widely used for detecting land cover change areas. However, the traditional methods are usually developed by using a single change index, which may cause omission or commission of land cover change. To address this challenge, we propose a change difference map fusion-based land cover change detection approach. First, operators of change vector analysis (CVA) and spectral gradient difference (SGD) are used for constructing change difference map, respectively. Second, image product weighted fusion method is introduced to construct a comprehensive change difference map. Finally, the expectation maximization algorithm and Bayesian rule with minimum error rate are applied to get the change/unchanged area. Experimental results of Landsat5 TM in 2010 and Landsat8 OLI in 2015 in two test areas are conducted. The experimental results show that accuracy of the proposed method is superior to the traditional change detection methods.