【摘要】 针对无人机高空输电线实时巡线识别速度慢、精度低等问题,提出基于图像特征检测识别EDLines算法,以去除背景干扰提高输电线识别精度。通过改进锚点判定的方式搜索锚节点,搜索邻域梯度最大值作为绘制边缘的路径,利用随机抽样一致性(RANSAC)算法拟合直线段,以亥姆霍兹原理对虚假线段进行滤波处理,采用K均值聚类筛选出输电线特征直线,设计方向模板二次滤波定位输电线的位置。实验结果表明:本文算法识别输电线准确率提高,抗干扰能力增强,运行效率高,具有工程实践意义。更多还原

【Abstract】 Aiming at the problems of low speed and low accuracy of high-altitude transmission line inspection by UAV in reatime, EDLines algorithm based on image feature detection and recognition is proposed to remove background interference and improve the accuracy of power line recognition.The algorithm searches for anchor nodes by improving the way of anchor point determination, searches for the maximum value of the neighborhood gradient as the path for drawing the edge, uses the random sample consensus(RANSAC)algorithm to fit the straight line segment, filters the false line segment based on the Helmholtz principle, and uses K-means aggregation Classes are used to filter out the characteristic lines of the transmission line, and the design direction template is used for secondary filtering to locate the transmission line.Experimental results show that the algorithm in this paper improves the accuracy of power transmission line recognition, enhances anti-interference ability, and has high operating efficiency, which has practical engineering significance.更多还原