Linear detection research of remote sensing image based on morphological filter and hough transformation

Abstract

Because remote sensing image is characterized by abundant grayscale change, great amount of information, complicated background noise, some methods can successfully detect binary image, however, detection of these methods in the remote sensing image is difficult to meet actual needs. The paper summed up the key problems of Hough transformation, and a modified algorithm for peak value cluster round phenomena in the peak value extraction of Hough transformation is proposed in this paper, the experiments have shown that peak value extraction algorithm proposed in this paper can greatly increase the accuracy of parameter estimation. Combining the modified Hough transformation and morphological filter in the linear detection of remote sensing image, the paper put forward a linear detection method based on morphological filter and Hough transformation, comparing to the simulation results of classical Hough transformation, the algorithm in this paper can achieve good results in the detection speed and the target location precision. © 2011 Springer-Verlag Berlin Heidelberg.