A STUDY ON THE ANALYSIS OF TM IMAGE INFORMATION CONTENT AND THE EXTRACTION OF CHARACTERISTIC INFORMATION
TM 图像的信息量分析及特征信息提取的研究

Analysis of image information content is the basis of image treatment. The image information content of three areas of different vegefation cover types namely multi-forest region (more than 40% of forest cover), hilly region of average forest land distribution (about 10-30% of forest cover), and hilly and plain region mainly with farmland was studied, the image of the same region in winter and summer were analyzed. On the basis of image information content analysis, methods of treatment of different images were studied, and their practicality was analyzed and explored as well.The analysis of information feature shows the general law that the information content of infrared band is higher than that of visible light band, in which the maximum of information content is band TM5, and the minimum, band TM2, Meanwhile, in lifferent conditions, the statistical feature value of correlation between bands, mean value and standard deviation was analyzed, thus making image treatment establish on the basis of the analysis of information content and statistical feature. Accordingly, the enhencement of image, the method of information feature extraction (such as the analysis of principal composition, KT transformation and ratio method) and the combination of bands were researched in a systematic way, and the conditions of their practicality were also explored and evaluated