一种采用航拍图像识别高压输电线路防震锤缺失的检测算法

An algorithm for missing detection of anti-vibration hammer on high voltage transmission line based on aerial image

高压输电线上防震锤长期经受各种恶劣的自然条件影响,容易导致重锤坠落,如不及时更换缺失的防震锤,会使导线疲劳断股的风险增加。为了解决高压输电线上的防震锤巡检问题,提出一种高压输电线路上的防震锤缺失检测方法。首先使用基于HOG特征与SVM分类器对无人机航拍图像进行防震锤目标定位,然后结合防震锤的结构对称特征,将定位到的防震锤旋转变换至水平状态后进行垂直投影,依据投影分布中两侧重锤投影坐标与线夹投影坐标之间的距离差异来判断防震锤是否缺失。结果表明,该方法将二维图像数据转换为一维数据分析,在保证准确率的同时,简化了分析问题的难度,处理速度快,且极大降低了拍摄光照、目标颜色等因素对缺失检测的影响。

The anti-vibration hammer on the high-voltage transmission line is subjected to various harsh natural conditions for a long time,which is easy to lead to the fall of the heavy hammer to fall.If the missing anti-vibration hammer is not replaced in time,the risk of wire fatigue and strand breakage will increase.In order to solve the problem of anti-vibration hammer inspection on high-voltage transmission lines,a detection method for the absence of anti-vibration hammers on high-voltage transmission line is proposed.Firstly,using HOG features and SVM classifiers to locate the anti-vibration hammer target of UAV aerial images,next,combined with the structural symmetry characteristics of the anti-vibration hammer,the positioned anti-vibration hammer is rotated to the horizontal state and then vertically projected.According to the distance difference between the projection coordinates of the heavy hammer on both sides and the projection coordinates of the line clamp in the projection distribution,whether the shock hammer is missing or not can be judged.The result shows that this method converts two-dimensional image data into one-dimensional data analysis,which not only ensures accuracy,but also simplifies the difficulty of analysis,and the processing speed is fast,which greatly reduces the influence of shooting light,target color and other factors on missing detection.