论文部分内容阅读
联合GPS、ISN、LiDAR、测距机等,构建超POS信息;计算最小视场分辨率、像元数、焦距等选择相机;将POS采集系统与相机组合成LiDAR多通道光谱图像异常识别系统.采用多通道匹配融合法融合紫、红外、彩色图片,基于Hough变换,通过同族容器归纳法确定疑似故障点.运用Hough变换、免疫遗传Snake、最小二乘法解析椭圆形貌,解决绝缘子异常识别问题.工程实验表明,该系统平均探测精度是82.4%,优于直升机与人工平均值24.05%,是一种高效率的智能电网巡线排查手段.
The system integrates GPS, ISN, LiDAR and range finder to construct super-POS information; calculates the minimum field of view resolution, pixel number, focal length and other selection cameras; combines POS acquisition system and camera into LiDAR multi-channel spectral image abnormality identification system. The multi-channel matching fusion method was used to fuse the purple, infrared and color images, and the suspected fault points were determined by the homology container inductive method based on the Hough transform. The Hough transform, immune genetic Snake and least square method were used to resolve the elliptic topography. Engineering experiments show that the average detection accuracy of the system is 82.4%, better than the helicopter and the artificial average of 24.05%, is a high-efficiency smart grid inspection route means.