Research And Implementation Of Image Edge Detection Algorithm

In recent years, with the rapid development of computer technology, image edgedetection has become one of the most important field in digital image processing, andplays a significant role in image analysis, image segmentation, feature extraction andimage recognition. With the extension of application field, its implementationdifficulty increases continuously. The main contents are as follows.First of all, briefly introduced the development, concept and application fields ofdigital image processing.At the meantime, introduced the background, history andbasic concept of the study of image edge detection.Secondly, introduced the classical edge detection operators respectively,including Roberts operators, Sobel operators, Laplacian operator, etc. and comparedtheir advantages and range of application through theoretical analysis and simulation.Then introduced some common edge detection methods, such as, Canny operator,SUSAN feature detection, edge detection based on wavelet transformation modulusmaxima, specific angle edge detection based on2D finite impulse response filter andThe morphology gradient edge detection, etc.Finally, this article introduced Hilbert transform in frequency domain and timedomain respectively, and adopts Hilbert transform in time domain to realize crackedge detection. Then made a comparison between theoretical analysis and the result ofthe edge detection method mentioned above, and drew a conclusion that the effect ofHilbert transform in frequency domain is far better than that of Hilbert transform intime domain.