Image edge detection using adaptive morphology Meyer wavelet-CNN
Young-Hyun Baek, Oh-Sung Byun, Sung-Rung Moon · 2004
The boundary of an object image in a digital image is distorted by noise or other elements of the system that mix with the image during transmission. In this paper, it is proposed that the object boundary be detected and exactly divided for optimal edge detection method. After bringing up the object boundary by applying adaptive morphology on the threshold of the input image, the optimal edge is detected using wavelet-cellular neural network (CNN). The proposed method is compared with the conventional Sobel method used as an edge detection algorithm. The proposed algorithm is confirmed to be superior than the conventional methods.