期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2009
卷号：XXXVIII-7/C4
页码：199-202
出版社：Copernicus Publications
摘要：With the rapid development of RS and more mature and the precision of number of field surveys on the basis of Quick-Bird remote sensing data through visual interpretation of urban public green vegetation diversity survey the exits of mixed. Pixel mixed is one important factor of influence accuracy of Remote sensing image classification. In this article, combine genetic algorithm with BP propagation to a new analysis module. The results show that the algorithm is effective. Remote sensing image processing detect and obtain features information by every pixel. If can identify the ratio of variety typical features compositing mixed pixel by certain means, you can solve the problem of mixed pixel and improve qualitative and quantitative accuracy of remote sensing survey
关键词：Images; mixed-pixel; neural-network; Public green; genetic BP propagation algorithm
