一种基于自适应阈值Canny算法的输变电设备图像检测方法研究

Research on image detection of power transmission and transformation equipment based on adaptive threshold Canny algorithm

图像边缘检测是输变电设备故障检测或获取关键信息的重要方法。传统的Canny边缘检测算法存在对噪声敏感、滤除噪声时容易丢失弱边缘信息、其固定参数适应性差的问题,本文提出了一种基于传统Canny算法的改进算法,引入了引力场强度算子来计算图像梯度,并针对较少边缘信息和丰富边缘信息两种典型图像,分别提出了基于图像梯度幅度和标准差均值的自适应阈值选取方法。在输变电设备图像边缘检测的几组对比试验表明,改进的Canny算法对噪声更具鲁棒性,可以保留更多有价值的边缘信息,能够有效提高输变电巡检图像识别准确率,具有很好的工程应用价值。

The traditional Canny edge detection algorithm is sensitive to noise,easy to lose weak edge information when filtering noise,and its fixed parameter is poorly adaptable. This paper proposes an improved algorithm based on the traditional Canny algorithm,which introduces the gravitational field intensity operator to calculate Image gradient,and for two typical images with less edge information and rich edge information,an adaptive threshold selection method based on image gradient amplitude and standard deviation mean were put forward. Comparative experiments on image edge detection of power transmission and transformation equipment show that the improved Canny algorithm is simple and easy, can retain more useful edge information,and is more robust to noise,effectively improving the accuracy of image recognition for power transmission and transformation inspection,and has a good engineering application value.