Edge Detection Using Modified Directional Coefficient Mask in AWGN

Image segmentation is required for the analysis of images and edge is one of the essential elements of image segmentation. Edge contains image information and it is applied in various fields of image processing. Typical methods of edge detection include Sobel, Prewitt and Roberts method and such methods have the advantage of simple realization and fast processing speed as they process images with mask in spatial area. However, when images are degraded by the addition of AWGN, an error of detecting edge in noise areas occur. Therefore, in this paper a new edge detection algorithm with excellent edge detection characteristics which effectively removes AWGN is proposed.

[1]  Zhou-Ping Yin,et al.  The Fast Multilevel Fuzzy Edge Detection of Blurry Images , 2007, IEEE Signal Processing Letters.

[2]  Reinhard Klette,et al.  Tracking of 2D or 3D Irregular Movement by a Family of Unscented Kalman Filters , 2012, J. Inform. and Commun. Convergence Engineering.

[3]  S. Agaian,et al.  Progressive Edge Detection on multi-bit images using polynomial-based binarization , 2008, 2008 International Conference on Machine Learning and Cybernetics.

[4]  Nam-Ho Kim,et al.  A Study on Wavelet-based Image Denoising Using a Modified Adaptive Thresholding Method , 2012, J. Inform. and Commun. Convergence Engineering.

[5]  Yi-Xin Zhao,et al.  Crack edge detection of coal CT images based on LS-SVM , 2009, 2009 International Conference on Machine Learning and Cybernetics.

[6]  Shun-feng Ma,et al.  Directional multiscale edge detection using the contourlet transform , 2010, 2010 2nd International Conference on Advanced Computer Control.

[7]  Tae-Yeon Kim,et al.  A Novel Approach to General Linearly Constrained Adaptive Arrays , 2012, J. Inform. and Commun. Convergence Engineering.