摘要
为实现管道内表面腐蚀图像的边缘检测,分析了经典的边缘检测方法,针对其存在检测精度低和抗噪声性能差等缺点,研究了一种基于BP神经网络的图像边缘检测算法。利用标准图像和经传统边缘检测算法检测得到的边缘图像作为输入输出数据,并用大量数据进行训练,构建了可实现图像边缘检测的BP神经网络。用训练好的神经网络实现管道内表面腐蚀图像边缘检测,并与传统的边缘检测算法检测结果进行了比对,实验结果表明,该算法可明显提高检测精度及抗噪声能力,具有广泛的适用性。
In order to detect the edge of pipeline inner corroded image, the classic edge detection method is analyzed. And it is found that the detection precision is not high and the anti-noise performance is poor. On this basis, an image edge detection algorithm based on BP neural network is researched. To build the BP neural network, standard image is made as input data, and the edge image of standard image detected by traditional edge detection operator is made as output data. And a large amount of data is used for training. Finally, the experimental result of the edge detection of the corroded image inside the pipeline detected with the BP neural network method is given, and it is compared with the result of traditional edge detection algorithm. The results show that the proposed algorithm can improve the detection precision and anti-noise ability significantly, and it is a kind of algorithm with extensive adaptability.
出处
《电子测量与仪器学报》
CSCD
北大核心
2017年第11期1788-1795,共8页
Journal of Electronic Measurement and Instrumentation
基金
国家自然科学基金(61362023)资助项目
关键词
管道腐蚀
边缘检测
图像处理
BP神经网络
pipeline corrosion
edge detection
image processing
BP neural network
