Study on Automation Change Detection Based on Object Similarity for High Resolution Imagery
Beijing Normal University
Email: [email protected]
T Sasagawa, P Gong, Peng P J Shi
Change detection is more and more used in the application of urban remote sensing. Current change detection are mainly developed for relatively coarse resolution remote sensing data. Pixel by pixel comparison and post-classification comparison are major techniques. The method of change detection based on integration of change vector analysis and object similarity calibration is presented for high spatial resolution data, which integrates the features of pixel-based comparison and object-based comparison. The result will be compatible to complexity and fuzzy change of objects in high spatial imagery in different time, which is distinguishing to the result using conventional change detection.