遥感图像线性构造信息微机机助提取

Use of Microcomputer in Automated Lineament Detection from Remotely-Sensed Imagery

本文采用图像平滑、边缘跟踪、霍夫变换、逆霍夫变换等综合图像处理技术，从卫星ＴＭ图像中提取线性体信息。在霍夫变换后的累加器阵列中采用局部最大值选择，逆霍夫变换后采用剖面分析等人一机交互处理方法，结合遥感地质学家的图像判读经验，提高了线性体信息自动提取的精度。

This paper has demonstrated the possibility of developing algorithms toextract lineament information from Landsat TM imagery.We made use of imagesmoothing,edge tracking,Hough transformation,inverse Hough transformation etc.in automated lineament detection.After,Hough transformation,local maxima areselected in the accumulator array and after inverse,Hough transformation,the pro-file analysis for line segment identification and human-machine interactive methodsare used. Some of the rules used by geologists in their image interpretation are app-lied in automated lineament extraction from digital imagery for improving precision.