Research on Distance Perception Engineering Quantity Accounting Technology of Transmission Line Grounding Device Based on Binocular Stereo Vision AnyNet Network

This paper studies the distance perception engineering quantity accounting technology of transmission line grounding device based on binocular stereo vision AnyNet network. This paper introduces a binocular stereo matching algorithm using deep learning to estimate the depth of the scene. By combining lightweight binocular stereo matching technology and deep neural network, high-precision grounding grid depth length recognition is realized. This method can quickly generate high-precision disparity map through multi-stage prediction and disparity image element correction, improve the accuracy and efficiency of engineering quantity accounting, reduce human error, and ensure the objectivity and reliability of accounting results.