基于图像识别的输电线路状态智能监测研究

Research on Intelligent Monitoring of Transmission Line Status Based on Image Recognition

为输电线路安全运行提供基础,提出基于图像识别的输电线路状态智能监测方法。利用多尺度Retinex算法对输电线路图像增强处理,提升输电线路图像质量,利用Mallat多尺度小波变换方法提取输电线路图像边缘的小波系数特征,选取模糊支持向量机方法依据小波系数特征进行实现输电线路状态智能监测。实验结果表明,该方法处理后输电线路图像信息熵高,可以有效识别输电线路的绝缘子覆冰、输电导线舞动等异常状态,输电线路状态智能监控效果理想。

In order to provide the basis for the safe operation of transmission line,an intelligent monitoring method of transmission line condition based on image recognition is proposed.The multi-scale Retinex algorithm is used to enhance the transmission line image and improve the image quality of the transmission line.The Mallat multi-scale wavelet transform method is used to extract the wavelet coefficient characteristics of the edge of the transmission line image,and the fuzzy support vector machine method is selected to realize the intelligent monitoring of the transmission line state according to the wavelet coefficient characteristics.The experimental results show that the transmission line image information entropy processed by this method is high,which can effectively identify the abnormal states such as insulator icing and transmission conductor galloping,and the effect of intelligent monitoring of transmission line state is ideal.