Remote Sensing Image Change Detection And Classification Techniques

From the actuality and development of remote sensing, the paPer focuseson tWo aspects of remOte sensing technoloQ; that is the sensor technology andremote sensing information processing techno1ogy. Based on informationacquiration of interestS objectS, the paPer does corresponding research work inthe fields of objectS diversification detechon and classification recognition.According to the basic principles of diversification derection, the paPer makes asystematic research into three mathematic models fOr diversification deteCtionand analyZe aPplication area of each scheme. Further more the paPer makesconclusion and analyZaion about charaCer image processing technology, givesa review of remote sensing classification recognition technology, summarizesand analyrts the developing trend of classification aIgorithIns. MeanwhiIe, acorresponding exPeriment system is founded up aimed at the research ofdiversification detection and images classification. The main work of the paPerare:l. The analyZaion of the aCtUality and development trend of remote sensing.2. Combined with the latest theory of diversificaton detection technology,several feasible schemes for diversification detection are proposed.3. In combination with character images several feasible methods for imageprecision and veCtor data extraction are proposed.4. MaIdng analyZaion and conclusion about chnder election, charactereXtraction-the critica1 tache in remote sensing classification. Giving asummariZation about basic classification algorithIns proposed up in therecam years and concluding the development trCnd of classificationalgorithms.