Abstract: A method of road boundary and feature point extraction for remote sensing image was proposed based on contrast enhancement and morphology. Firstly, it carries on contrast enhancement of remote sensing image. Piecewise linear enhancement method with larger discriminative power was selected to do image enhancement after comparative analysis difference between histogram equalization and contrast piecewise linear enhancement. In order to realize the separation of road and other information effectively, image segmentation was done using mathematical morphological method. Then, using edge detection method by Krisch operator to extract road edge information, and feature points extracted by improved Harris operator based on edge feature. Last, fitted the extracted feature points and the function model was used to describe the road information of the image and applied to vector road information in the later period of drawing.

Key words: contrast enhancement, mathematical morphological method, edge detection, feature point extraction, curve fitting