论文部分内容阅读
本文采用了I~2S101型图象信息处理系统对胶州湾幅CCT磁带信息进行了非监督分类,这是对多光谱遥感信息(CCT磁带)首次运用机器进行图象自动识别的一种尝试.但取得了理想的解译成果.根据地物光谱特征的差异性,建立在统计概率基础之上而运用机器进行自动图象识别的非监督分类方法,从分类的结果可知,对胶州湾幅地貌形态特征做了圆满的解译(见图3).经实地考查,对解译成果做了验证.特别是分类结果给出了各类地物的面积及百分比(见附表1、2),这是常规方法所难于实现的.由此可知,非监督分类方法是机器自动进行图象识别的基本方法,从地物信息的提取及经济效益的角度分析,它可以和监督分类方法媲美.特别是在那些缺乏独立信息的地区,此方法独具一格.
In this paper, I ~ 2S101 image information processing system is used to unsupervised classification of CCT tape information in Jiaozhou Bay, which is an attempt to automatically identify the images of multi-spectrum remote sensing information (CCT tapes) And achieved the ideal result of the interpretation.According to the differences of the spectral features of the features, based on the statistical probability and based on the unsupervised classification method of the machine for automatic image recognition, we can see from the classification results that the geomorphological features (See Figure 3) .Through the field test, the interpretation of the results have been verified.Especially the classification results give the area and percentage of all kinds of features (see Annex 1, 2), which is Which can not be achieved by the conventional method. Therefore, the unsupervised classification method is the basic method of automatic image recognition by the machine. It can be compared with the supervised classification method from the perspective of the extraction of geomorphologic information and economic benefits, In areas where there is a lack of independent information, this approach is unique.