Digital Image Edge Detection Based on LVQ Neural Network

In view of the traditional image edge detection method easily lead to the disadvantage such as burr and discontinuous, use LVQ neural network to detect the digital image edge. By extracting and calculating the median characteristics, directional information characteristics and Krisch operator direction characteristics of each sample image point, generate the neural network training set and train the network. In order to enhance the effect of edge detection and eliminate the effects of the noise signal, it can use smoothing filter to eliminate or minimize noise and improve the quality of the image. The simulation results show that compared with the traditional edge detection operator, the edge image is more clear and denoising effect is more obvious by using the neural network, which also has better effectiveness and robustness.