1.前言林业生产中的陆地卫星图像数字解译主要基于逐个像元进行的解译方法,如极大似然法。由于TM图像和SPOT图像空间分辨率的提高,使图像中相邻像元光谱向量的变化更加复杂,噪音成分相对增加,如果仍用经典的方法进行图像解译可能会产生混淆不清的结果。为了更有效地处理这些图像,必须对原有方法进行修改。 1. Introduction Digital interpretation of terrestrial satellite images in forestry is based primarily on pixel-by-pixel interpretation, such as maximum likelihood. Due to the increase of spatial resolution of TM image and SPOT image, the change of spectral elements of adjacent pixels in the image is more complicated and the noise component is relatively increased. If the traditional method is still used for image interpretation, confusion may be produced. In order to deal with these images more effectively, the original method must be modified.