Title :
Angle-dependent edge detection for traffic signs recognition
Author :
Sandoval, Hector ; Hattori, Taizoh ; Kitagawa, Sachiko ; Chigusa, Yasutami
Author_Institution :
Dept. of Inf. Technol., Tokyo Univ. of Technol., Japan
Abstract :
This paper describes an alternative method to generate convolution masks for edge detection. Values of the convolution masks depend on the relative position of the pixel being analyzed. As a result, only edges that fulfill a predetermined condition are detected and other edges are filtered. The method is then applied to shape detection of circular traffic signs
Keywords :
computer vision; edge detection; traffic engineering computing; convolution masks; edge detection; shape detection; traffic signs recognition; Brightness; Business; Convolution; Filters; Image edge detection; Information technology; Machine vision; Performance evaluation; Pixel; Shape;
Conference_Titel :
Intelligent Vehicles Symposium, 2000. IV 2000. Proceedings of the IEEE
Conference_Location :
Dearborn, MI
Print_ISBN :
0-7803-6363-9
DOI :
10.1109/IVS.2000.898360
