Efficient Edge Detection Using Mean Shift Smoothing and Gradient-Based Noise Removal

Haitao Sang, Zhen Zhou
2016 Innovative Computing Information and Control Express Letters, Part B: Applications  
For the low-contrast edge and slow calculation speed of conventional algorithms for edge detection, this paper proposes an efficient method of edge detection using mean shift smoothing and gradient-based noise removal. The mean shift smoothing could efficiently protect the edge details and amplify the contrast along image edges to give more clear contours in edge detection. Then an improved method based on the eight neighborhood gradient magnitude is designed to overcome the disadvantages of
more » ... ng sensitive to noise in the calculation of the traditional operators gradient. Finally, a comparative study of different edge detection techniques is tested and analyzed, which proves the effectiveness of our method such as continuity, smoothness, noise removal and execution time.
doi:10.24507/icicelb.07.03.519 fatcat:ydmkuvoaafelxcr2fqotto2nki