|
Mutations often is one of the important characteristics of the transient signal and image analysis , and they are often located in the edge portions of the important structures . Edge accuracy , noise resistance , and edge detection performance . The traditional differential operator edge detection , such as Roberts, Sobel, Laplacian operator , better real-time , but interference poor . The subsequent emergence of Canny operator , although the performance is better than the other differential operator , but poor positioning . This two edge extraction methods from wavelet theory to start constructing appropriate wavelet modulus maxima method to extract the edges , this method can achieve better results , but the selection of wavelet more strict requirements. Another method is a grayscale morphological point of view , the constructor structural unit , the etching is performed on the image or expansion , highlighting its edge information , and then get its edge . The advantage of this method is the edge of high precision , but the details of the edge information extraction with a structural unit selected inextricably linked . In this paper, on the basis of these two methods , the first wavelet modulus maxima method proposed a simplified algorithm , morphological edge detection operator to be improved , a modified edge detection operator , Finally combining these two methods , starting from the theoretical point of of the wavelet transform , the image in the multi-scale decomposition of the details of the various high-frequency of the image at all scales weighting process , and then starting from the angle of the grayscale Morphology proposed a kinds of image edge extraction algorithm based on wavelet transform and mathematical morphology . Simulation and analysis of results , the experimental results show that the edge of precision , the strength of the edge extraction and noise suppression , the proposed method is effective .
|