作者: Min Zhang , Lequan Min , Xiaoliang Zhang
DOI: 10.1007/978-3-642-54924-3_55
关键词: Artificial intelligence 、 Cellular neural network 、 Image (mathematics) 、 Prewitt operator 、 Image processing 、 Corner detection 、 Algorithm 、 Edge detection 、 Template 、 Sobel operator 、 Computer science
摘要: This paper proposes a robust design theorem for kind of uncoupled cellular neural/nonlinear networks (CNNs) in aim to process images with some specific image properties. A novel algorithm scheme is introduced automatically determine the required parameters given theorem, and implement processing. Using proposed designs convex corner detection, CNN template, edge detection templates. Four processing examples are given. Comparing algorithms “Sobel,” “Prewitt,” “Roberts,” “Canny” our shows that presented has promising results.