THE ENVIRONMENT MEANING OF FACTOR ANALYSIS OF REMOTE SENSING IMAGES
遥感图像因子分析的环境意义

In this paper, experiments and theory analysis were made. The remote sensing images of 6 geographical environment were processed with factor analysis algorithm, the results were stated in table 1. Results introduced the relationship of image factors and the bases of Tasseled Cap space Generally, the first factor vector most similar to SB and second factor vector most similar to GV, as for plain agricultural environments (environment 5 in the table), inverse re-sults appeared.To analyze the experimental results, we developed a kind of theory to explore these results with physical geographical theory.In first work, the unequivalentity of the response to geographical landscapes of the sensor has been discovered, with using the supposing that ground geographical feature is 2-di-mension zero mean stationary stochastic processes and the effect of IFOV's convolution integral. It shows that the signal of primary environment landscape factor would been increased and secondary environment landscape factor would relatively been decreased under the observation of the sensor. The gain factor is sinwx sinwy/wxwy, where wx and wy are angular frequency of space, and X, Y are defined as the sweeping direction of the sensor and the moving direction of the sensor, respectively.In second work, a concept of resolution of a environment landscape, was developed dependent on the discovery. A method that equivalent to SVD transformation of image processing has been developed, it would used to process the environment information of remote sensing images.