基于三目视觉的输电导线弧垂三维重建方法研究

Research on 3D Reconstruction Method of Power Wire Arc Sag Based on Triocular Vision

输电线弧垂的三维重建对及时发现输电线路潜在危险源和电网的安全稳定运行具有重要意义。为了保证重建的准确率,提出了一种基于三目视觉的输电线弧垂三维重建方法。首先单独标定各个摄像机,再基于特征点使用深度特征两两标定摄像机,并采用光束平差参数优化方法标定三目摄像机;然后,通过视点平面扫描技术为每个摄像机计算空间概率图,使用半全局代价空间优化方法优化空间概率图,获取3个深度点云。最后,通过融合和过滤深度点云,最终得到重建后的深度图。实验证明本文的方法可以自动实现高精度的输电线弧垂的三维重建,有利于保证输电线路的安全。

The three-dimensional reconstruction of power line sag is of great significance in avoiding dangerous factors and realizing the safe and stable operation of power grid.To ensure the accuracy of reconstruction,a three-dimensional reconstruction method of power line sag based on trinocular vision is proposed.After the separate calibration of each camera,and the binocular calibration based on deep features,this paper employs the beam adjustment parameter optimization to calibrate the three-lens cameras.Then,the spatial probability map of each camera is calculated by viewpoint plane scanning technology,and the spatial probability map is further optimized by semi-global cost space optimization method,hence three depth point clouds are obtained.Finally,the reconstructed depth map is obtained by fusing and filtering the depth point cloud.The real experiments show that our method can automatically realize the three-dimensional reconstruction for the high-precision power line sag,which is beneficial to ensuring the safety of the power transmission line.