基于图像识别的输电线路设备缺陷识别系统设计

Design of transmission line equipment defect identification system based on image recognition

因输电线路所处位置大部分为山区,受环境、气候等条件的限制,无法及时识别输电线路设备缺陷。为了能够保证输电线路正常运行,预防意外事故发生,提出基于图像识别的输电线路设备缺陷识别系统设计方案。通过调用图像识别程序判断线路的周围状况,将图像数据压缩发送到后台服务器,设置存储逻辑对画面分类和存档。将图像直接输入卷积神经网络,利用激活函数生成特征图,获得向量表达抽象图像,优化损失函数。分析输电线路点云数据分布的投射点,提取图像初值匹配,剔除离群特征点,实现线路设备缺陷识别。试验结果表明,所设计系统可以精确地识别出输电线路设备缺陷位置,误差较小,能保证输电线路运行安全。

Due to the fact that most of the transmission lines are located in mountainous areas,the defects of transmission line equipment cannot be identified in time due to the limitations of environment,climate and other conditions.In order to ensure the normal operation of transmission lines and prevent accidents,an application system design scheme of transmission line equipment defect recognition based on image recognition is proposed.Judge the surrounding conditions of the line by calling the image recognition program,compress and send the image data to the background server,and set the storage logic to classify and archive the images.The image is directly input into the convolutional neural network,and the feature map is generated using the activation function to obtain the vector expression abstract image and optimize the loss function.Analyze the projection points of transmission line point cloud data distribution,extract image initial value matching,eliminate outliers,and realize line equipment defect recognition.The experimental results show that the designed system can accurately identify the defect location of transmission line equipment,with small error,and can ensure the safe operation of transmission lines.