基于卷积神经网络的Canny算法优化

Canny Algorithm Optimization Based on Convolutional Neural Network

Canny算法常用于图像边缘特征的有效提取,但是Canny边缘检测需要人工预先设定双阈值参数。针对这一问题,提出一种基于卷积神经网络(CNN)的双阈值参数优化方法,建立Canny双阈值与传播因子的系统模型,通过梯度下降算法快速求解双阈值。实验结果表明,在道賂标识牌边缘特征提取场景中,优化后的Canny算法可有效提取边缘信息,在低错误率、高定位性、最小响应时间三个方面明显优于传统Canny算法。

Canny algorithm is often used to extract image edge features effectively,but Canny edge detection requires manual pre-setting of double threshold parameters.In order to solve this problem,a double threshold parameter optimization method based on convolutional neural network(CNN)is proposed.The system model of Canny double threshold and propagation factor is established,and the double threshold is solved quickly by gradient descent algorithm.The experimental results show that the optimized Canny algorithm can extract the edge information effectively in the scene of road sign edge feature extraction,which is superior to the traditional Canny algorithm in three aspects:low error rate,high positioning and minimum response time.