基于PCNN与形态学的坑洞图像边缘提取

Potholes Image Edge Extraction Based on PCNN and Morphology

对坑洞图像边缘提取进行了研究,改进了脉冲耦合神经网络模型,提出了一种PCNN和形态学相结合的边缘提取方法。对基本PCNN模型进行优化,简化了原模型参数,并改进了原模型的线性输入项和脉冲输出计算方法。在图像边缘提取过程中,先对图像进行增强,在一定程度上消除坑洞周围环境对坑洞边缘的影响,再利用改进的PCNN模型和形态学的膨胀腐蚀特性对其进行边缘提取。实验结果表明:该方法对路面坑洞图像的边缘提取比传统边缘提取算法更为有效,抗干扰能力强,能有效地抑制路面环境对坑洞边缘的影响,所提取到的边缘更加清晰、可用。

The potholes image edge extraction was studied,the pulse coupled neural network model was improved,and an image edge extraction method based on PCNN and morphology was proposed. The basic PCNN model was optimized and the original model parameters were simplified. Furthermore,the linear input and the calculation method of the output pulse of the original model were improved. First of all,the image was enhanced to eliminate the influence of potholes surrounding environment on the potholes edges to some extent in the process of image edge extraction. And then,the improved PCNN model as well as the dilation and erosion characteristics of morphology was used to carry out the edge extraction. The experimental results show that： the proposed method is more effective than the traditional edge extraction method in the road potholes image edge extraction and has stronger anti-interference ability,which can effectively restrain the influence of the road surrounding environment on the potholes edge. And the extracted edges are clearer and more available.