摘要
在高分辨率全色遥感影像中,阴影的存在既给目标识别带来了困难,又包含了地物的层次和高度信息。本文运用灰度统计直方图的势函数对全色遥感影像的灰度统计直方图函数进行拟合,使用一阶差分找到直方图势函数的第一个谷点,以该谷点对应的灰度值对原始影像做阈值分割完成阴影检测。实验表明,在各种复杂地物环境下,本文方法都能完成灰度分割阈值的自动选取,其阴影检测结果是快速有效的。
Shadow brings difficulties to object recognition in the high-resolution panchromatic remote sensing images. On the other hand, it contains much information about the hierarchy and height of the objects on the ground. We fit the gray histogram of the panchromatic remote sensing image by its potential function, and find the first vale point by the first difference. Shadow detection is then performed by threshold segmentation using the gray value which corresponds to the vale point. The experimental results show that this method is capable of selecting the segmentation threshold rapidly and auto- matically in complex images,and the shadows are detected correctly.
出处
《计算机工程与科学》
CSCD
2008年第12期43-44,93,共3页
Computer Engineering & Science
关键词
遥感影像
阴影检测
直方图
势函数
图像分割
remote sensing
shadow detection
histogram
potential function
image segmentation
