Researches On Land Use Change Detection Based On Change Vector Analysis Method

Land use change detection is an important aspect of global change research, and it plays an important role in resource sustainable utilization. Land use change detection by Remote Sensing technology is the most economical and efficient way to obtain land use change information. Therefore, researches on land use change detection methods based on Remote Sensing technology are important aspects of Remote Sensing applicable research.The purpose of the research presented on this dissertation is to explore modified methodologies for extracting land use change information based on change vector analysis method. Multifarious knowledges, such as Remote Sensing technology, Geographical Information System, digital image processing technology, pattern recognition , artificial intelligence, numerical analysis, and statistics are integraed in this research. High spatial resolution data, such as TM, ETM+, and ASTER data, are utilized in our research. The main original reseach results are as follows:First, an algorithm of relative radiometric nornalization technique based on feature corner was developed. With a new feature corner detection adaptive algorithm based on gray gradient and a new control point matching algorithm based on unequidistant dynamic template being developed, The algorithm could select stable ground points by extracting feature corners and matching homonymous ground points automatically. The first virtue of the algorithm is that it can select stable ground points widely instead of the limitation of bright object or dark object. The second virtue of...