Abstract:Based on the complex environmental characteristics of the power grid, a suitable 3D imaging device is developed to achieve all-round and high-precision imaging of the work scene. This achievement can provide more comprehensive environmental perception and risk monitoring, and provide reliable basic data for three-dimensional prevention and control. Therefore, this article proposes a 3D point cloud reconstruction method based on improved ORBSLAM. Firstly, a depth camera is selected as the experimental equipment to obtain the pixel coordinates of point cloud data. Then, through the transformation from pixel coordinates to image coordinates, it is converted into three-dimensional point cloud data. Finally, a simple loosely coupled improved ORBSLAM algorithm is used to supplement the estimated camera pose and detect errors with large estimation deviations, improving the accuracy of point cloud fusion in three-dimensional reconstruction of work scenes.