LVQ神经网络在图像边缘检测中的应用OA

Addressing the issues of noise sensitivity and edge blurriness present in traditional edge de-tection methods,an enhanced approach called LVQCA is proposed,which integrates the Canny operator with a learning vector quantization(LVQ)neural network.By using the Canny operator as a guidance signal for the LVQ neural network,the training process effectively captures edge information within images while enhancing the completeness and connectivity of the detected edges.Experimental results demonstrate that,compared to traditional operators like Roberts and Log,the LVQCA neural network exhibits higher accura-cy and robustness in edge detection of traffic images under adverse weather conditions,providing more pre-cise critical road information for autonomous driving.