Erscheint 4 Ausgaben pro Jahr
ISSN Druckformat: 1065-3090
ISSN Online: 1940-4336
Indexed in
CELLULAR NEURAL NETWORK IMAGE EDGE DETECTION BASED ON HYPERBOLIC TANGENT FUNCTION − FROM PHOTOGRAPHIC IMAGE TO FLOW VISUALIZATION
ABSTRAKT
This paper proposes the architecture of cellular neural network (CNN) for image edge detection based on a hyperbolic tangent function. This network gives better results compared to other methods and is capable of application in real time due to the parallel processing nature of the CNN. The paper first discusses the equilibrium point of the hyperbolic tangent function, and thereafter the values of a template suitable for cellular neural network are determined. The method effectively defines the template of CNN and uses MATLAB simulation to verify the effectiveness of the edge detection of wood defect images and fluid images. The experimental results show that the improved CNN model can significantly reduce the number of iterations in the edge extraction, and can more accurately determine the edge.