| In computer vision and image processing, edge detection concerns the localization of significant variations of the grey level image and the identification of the physical phenomena that originated them. This information is very useful for applications in 3D reconstruction, motion, recognition, image enhancement and restoration, image registration, image compression, and so on. Consequently, edge detection plays a basic and main role in the image engineering. Although an amount of algorithms have been created all over the world, they are not satisfied on denoising and extract information. Therefore, edge detection is always a hot point for researchers in image processing.As the requirement of the accuracy and the intellectual faculties in the digital image process go on developing, more and more mathematical knowledge and concepts are being used to find new methods. At present, applying multi-scale wavelet transforms, fuzzy and general fuzzy, stochastic process and so on to extract edge are novel and efficient thoughts. In this thesis, two novel methods are proposed: 1. Combining the analysis of multi-scale wavelet edge detection principle, different regulation of wavelet transformation to the noise and edge, and general fuzzy operator (GFO), a new method for wavelet edge detection based-on multi-scale general fuzzy image enhancement is proposed which performs well to phantom edge especially. 2. General fuzzy operator (GFO) enhancement is done firstly, and competitive fuzzy edge detection algorithm is run as a second step to extract edge. Consequently, improved competitive fuzzy edge detection algorithm is acquired.In the last part of this paper, some proposals for research of image process thought which are based on stochastic process and fuzzy theory are provided. |
PDF Full Text Request