Research On Change Detection Technology Of Remote Sensing Image

Change detection is one of key technique in remote sensing technology. It detect local texture or gray value change happened in the same imaging scene, to acquire the surface or surface features change information in shape, position, quantity and other attributes. In the past few decades, a large number of change detection methods have been developed. In this thesis, the current typical change detection techniques are studied. The major innovations of this thesis are listed as follows:1. By discussing direct comparison change detection method and post-classification change detection method, the strength and weakness of the two methods are summarized. Some typical change detection algorithms are studied, the factors, which restricting algorithm performance, are analyzed from the aspect of accuracy, timeliness, etc.2. The reason why classic FCM is sensitive to outliers, neighborhood information underutilize, etc been studied. Based on the reason this paper proposed an improved FCM algorithm. The new algorithm can take full advantage of neighborhood spatial information. The new algorithm is tested by the measured SAR data. Experimental results show that the algorithm is effective and reasonable.3. Based on the complementarity between direct comparison change detection method and post-classification change detection method, a new change detection algorithm framework, which taking the accuracy and efficiency into account, is proposed. Under this framework, the user can combine the different types of change detection algorithms freely, to accomplish a specific change detection tasks. Finally, the algorithm framework is utilized for the detection of floods. According to the experiment result, the feasibility of the new framework is proved.