摘要
通过IDL语言调用ENVI二次开发函数库,首先基于多波段谱间关系法实现华南山区细小水体的自动提取,然后利用数学形态学膨胀滤波算法对图像进行断线连接和小孔填充,最后通过对比形态学腐蚀算法和Pavli-dis异步细化算法对图像进行细化处理,从而得到比较理想的细小水体。结果表明:该方法能够有效解决水体提取后断线连接、空洞填充及细小水体细化等问题,从而为水体自动、快速提取和分类提供有力帮助。
The IDL language calling the ENVI secondary development library,firstly,this paper achieved the automatic small water-body extraction in the mountain area of south China based on method of multi-band spectral relationship,and then used the mathematical morphological dilation filtering algorithm to broken link and fill holes.Finally,by comparing the thinning result between morphological erosion algorithm and Pavlidis asynchronous thinning algorithm,it acquired an ideal small water-body.The conclusion proved this method can effectively solve the problem of broken link,filling holes and small water-body refining and so on,after picking-up water-body.Thus it provide quick and effective help for automatic picking-up water-body and classification.
出处
《测绘与空间地理信息》
2011年第6期64-66,75,共4页
Geomatics & Spatial Information Technology
基金
中铁第四勘察设计院集团有限公司基金项目(2009D-1)资助
关键词
水体提取
多波段谱间关系法
形态学膨胀滤波
图像细化
IDL
water-body extraction
method of multi-band spectral relationship
morphological dilation filtering
image thinning
IDL
