Feature Database And Fuzzy Technology-based Remote Sensing Image Classification

As a part of Remote Sensing Quantified Application software, The thesis has made a deep research on the techniques and algorithms about remote sensing image classification. Besides that, the thesis has accomplished a remote sensing classification system which has integrated all the algorithms and techniques mentioned, and this system is a part of RSQA now.The main research work is as follows: (1)Feature extraction's application on remote sensing classification. In this part, mainly introduce the algorithm of K-L transformation and analyze the common image features used in remote sensing image classification. (2)Discuss the traditional methods of image classification, introduce the fuzzy theory and its application on pattern classification. As a key point, analyze the fuzzy C-Means algorithm and giving a improved FCM algorithm. (3)Propose an image classification method based on feature database, mainly introduce the logical structure of the feature database and its application on remote sensing image classification.(3)Accomplish a remote sensing image classification system by the algorithms above.