模糊分类在土地利用/覆盖变更调查中的应用

Application Study of Fuzzy Classification in the Investigation of Land Use

模糊分类是近年来在遥感影像分类中引入的一种新的研究方法,是一种针对不确定性事物的分析方法。它以模糊集合论作为基础,运用数学模型计算它对于所有集合的隶属度,然后根据隶属度的大小,确定归属。笔者把“模糊分类”方法运用于吉林省白山市八道江区土地利用/覆盖变更调查中,从试验结果发现“模糊分类”在复杂地块影像分类中较传统的贝叶斯分类有更好的分类结果。

Remote sensing is more and more used in the land administration. Improving the classification precise of the remotely sensed imagery is an important issue in the application of remote sensing. Fuzzy classification is one of the new research method introduced to the remotely sensing imagery classification. People pay more focus on the new method nowadays. Fuzzy classification is a method based on the fuzzy set to analyze the uncertainty. It is a tool using the subjection degree to decide adscription. The advantage of fuzzy classification in the investigation of land use/cover in Badaojiang,Baishan city in Jilin province was evaluated. From the experimental result we can find that fuzzy classification is better than the traditional Bayes classification in the complication plot classification of remotely sensed imagery.