Title :
A new edge detection algorithm: Fast and localizing to a single pixel
Author :
Yu, Shunjia ; Liu, Rueywen
Author_Institution :
Dept. of Electr. Eng., Notre Dame Univ., IN, USA
Abstract :
Commonly used methods of edge detection are unable to localize the edges to single pixel, and hence yield thick lines. Edge detection by Haar transform is investigated. The method can localize the edge to single pixels, and its computational complexity is lower than the Sobel operator method by a factor of more than two. The Haar transform method is as efficient as other methods for curved edges, and visibly better for straight edges
Keywords :
computational complexity; edge detection; transforms; Haar transform; computational complexity; curved edges; edge detection algorithm; edge localization; single pixel; straight edges; Computational complexity; Computer vision; Discrete transforms; Equations; Image edge detection; Image segmentation; Image sequence analysis; Laboratories; Pixel; Signal analysis;
Conference_Titel :
Circuits and Systems, 1993., ISCAS '93, 1993 IEEE International Symposium on
Conference_Location :
Chicago, IL
Print_ISBN :
0-7803-1281-3
DOI :
10.1109/ISCAS.1993.393777
