基于多特征融合的玻璃绝缘子识别及自爆缺陷的诊断

在无人机检测输电线路缺陷研究中,为提高识别绝缘子的正确率,有效降低背景纹理及光线的影响,提出了一种融合绝缘子形状、颜色与纹理进行识别绝缘子的方法。针对玻璃绝缘子的掉片缺陷,研究了一种感知绝缘子片重心间距离的缺陷检测方法。该方法对绝缘子正确识别率高于90%,误识别率低于10%。通过无人机巡检采集的大量输电线路图像,实验结果验证这种方法在各种复杂背景条件下能有效地识别出绝缘子,并能检测玻璃绝缘子的掉片缺陷。

The Identification and Diagnosis of Self-Blast Defects of Glass Insulators Based on Multi-Feature Fusion

In order to improve the recognition accuracy of insulators in UAV inspection and effectively reduce the influence of the background texture and illumination, a new insulator recognition method is proposed, which integrates the shape, color and texture of insulators. Aimed at the off-chip defects of glass insulators, a defect-detecting method is presented, which can sense the distance between gravity centers of insulator chips, and has an recognition accuracy of insulators higher than 90%. Based on testing with numerous UVA inspection images of transmission lines, it is proved that the proposed method can effectively recognize the insulators under various complicated background conditions, and detect the off-chip defects of glass insulators.