Generalized likelihood ratio test (GLRT) was widely used in the change detection of synthetic aperture radar (SAR) images and employed as a competitive method in many studies. GLRT is applied to the neighborhood areas corresponding to the same geographical area in multitemporal SAR images. Generally, the difference image is acquired by applying the GLRT on a moving window empirically determined by the user, which is clearly unreasonable because the different pixels in an image may have different optimal windows. To overcome this drawback, we propose an adaptive GLRT (AGLRT) method for change detection in SAR images. AGLRT adaptively selects the homogeneity window for each pixel and can decrease the uncertainty of change detection due to manual selection of the moving window. Experiments are conducted on two data sets to confirm the performance of the proposed method. Experimental results show that AGLRT can both suppress the influence of noise and preserve edge details and improve the accuracy of change detection.

Subsidence coefficient is a key parameter for ground movement and deformation prediction when mining under the building, water, and railway; so how to get exact subsidence coefficient is one of the most important problems in the discipline of mining subsidence. Support vector machine (SVM) is a new algorithm of machine learning based on statistical learning theory. Compared with traditional method, SVM can be established under condition of deficient samples and abnormal observation result can be rejected effectively. Based on comprehensive analysis of effect factors on subsidence coefficient such as mechanical characteristics of upper rock stratum, thickness of alluvium deposit, ratio value of mining deepness to thickness, mining method and roof control method, etc, data from tens of typical observation stations was used as training samples, by means of electing kernel function, insensitive loss function, proper penalty factor, regression relation model of SVM was designed between subsidence coefficient and affecting factors. Finally, testing and analyzing was done, and research results show that the SVM relation model can calculate subsidence coefficient and reliable precision can be got, which can meet the requirement of engineering. Research findings prove that the method to calculate subsidence coefficient based on SVM method is feasible. Besides, multiple effect factors can be comprehensively considered with this method, thus a new approach of efficient and accurate calculation of subsidence coefficient is provided for future research.

Change vector analysis (CVA) and spectral angle mapper (SAM) are usually used to generate difference image in change detection of multispectral images. Although CVA and SAM can describe the difference between multispectral images, they are defined mathematically and lack support of human visual system (HVS) theory. Advanced structural similarity (ASSIM) complies with the pattern that human perceives the changes occurred in an objective scene. Nevertheless, ASSIM was designed for single band images and cannot be used for extracting multiband structural information from multispectral images directly. Therefore, we first propose two strategies to extract multiband structural information from multiband images. Then, we propose the approaches based on multiband structural information for change detection in multispectral images. Experimental results from one semisynthetic data set and two real data sets acquired by Sentinel-2A and QuickBird satellites validate the effectiveness of the proposed approaches.