一种无人机图像的铁塔上鸟巢检测方法

发布时间：2018-01-29 13:34

本文关键词： 无人机巡检鸟巢检测铁塔识别颜色与纹理融合　出处：《计算机工程与应用》2017年06期 　论文类型：期刊论文

【摘要】：由于鸟巢造成的输电线路跳闸事件频频发生,已严重威胁到国家电网的安全运行,为了降低复杂背景的影响,提出了一种自动检测铁塔上鸟巢的方法,首先识别巡检图像上铁塔所在区域,考虑到铁塔是由不同方向的线材构成的空间图像,将巡检图像分块并分析不同方向的线段密度,判决是否属于铁塔区域,将检测的分块铁塔区域聚类,进而识别铁塔区域。在铁塔区域内,搜索符合鸟巢样本的HSV颜色特征量的连通区域,作为候选的鸟巢区域,分析候选鸟巢区域的形状特征参数,描述鸟巢粗糙度的灰度方差特征量,描述鸟巢纹理的惯性矩特征量,通过对无人机巡检采集的输电线路图像的测试,验证了这种方法能有效排除背景的干扰,有效检测出铁塔上的鸟巢。
[Abstract]:Due to the frequent occurrence of transmission line tripping events caused by Bird's Nest, the safe operation of State Grid has been seriously threatened. In order to reduce the influence of complex background, an automatic detection method of bird's nest on iron tower is proposed. Firstly, the area where the iron tower is located in the inspection image is identified. Considering that the tower is a spatial image composed of wire in different directions, the inspection image is divided into blocks and the line segment density in different directions is analyzed to determine whether the tower belongs to the area of the iron tower. The detected blocks are clustered to identify the tower area. In the tower region, the connected area that accords with the HSV color feature of the bird's nest sample is searched as the candidate bird's nest area. The shape characteristic parameters of candidate bird nest area are analyzed, the grayscale variance characteristic quantity of bird nest roughness is described, and the moment of inertia characteristic quantity of bird nest texture is described. The transmission line image collected by UAV patrol is tested. It is proved that this method can effectively eliminate the interference of background and detect the bird's nest on the iron tower.
【作者单位】：国网浙江省电力公司检修分公司;上海大学通信与信息工程学院;
【基金】：2014年国家电网发展项目
【分类号】：TP391.41;TM754
【正文快照】： 随着我国高压输电架空线路的不断增多和生态环境的逐步改善,鸟害事故明显上升,鸟害造成的损失也明显增大,对电力系统的安全运行构成了严重威胁。文献[1]对国家电网公司所属输电线路鸟害故障调查统计进行分类分析,对引起线路鸟害跳闸的原因进行分析研究。鸟类活动引起的线路故

【相似文献】