高压输电线路巡线机器人目标自动识别跟踪方法

Target Automatic Recognition and Tracking Method for Line Patrol Robots in High-voltage Transmission Lines

在巡线机器人目标自动识别跟踪方法中,转换图像信号时易产生噪声,影响识别跟踪结果的准确率。为此,设计了一种新的高压输电线路巡线机器人目标自动识别跟踪方法。首先采用卡尔曼滤波技术对图像进行滤波处理,然后根据Haar-Like特征将图像转换为便于机器人识别的灰度图像。基于此,采取背景差分法排除灰度图像中非识别目标的背景,再根据亮度特性和色彩特性自动识别跟踪目标。以某地区的一段高压输电线路作为测试对象,实验结果显示,本文方法的识别图像受噪声影响较小,且自动识别跟踪的有效性更高。

In the current automatic target recognition and tracking methods for line patrol robots,noise is prone to occur when converting image signals,which affects the accuracy of recognition and tracking results.Therefore,this study designed a new method for automatic target recognition and tracking for line patrol robots in high-voltage transmission lines.Firstly,Kalman filtering technology was used to filter the image,and then the image was converted into a grayscale image that is easy for robots to recognize based on Haar Like features.Based on this,the background difference method was adopted to exclude the background of non recognized targets in grayscale images,and then the tracking targets were automatically recognized based on the brightness and color characteristics.Taking a high-voltage transmission line in a certain region as the test object,the experimental results show that the recognition image of this method is less affected by noise,and the effectiveness of automatic recognition and tracking is higher.