View planning method for complete 3D digitisation of the scene

Abstract

The problem of where to position the range sensor next for generating a completely seamless integration of a targeted surface from multiple range datasets acquired at different spatial poses is one of the challenging tasks in machine vision. A new view planning method is proposed for complete reconstruction of the real scene from the minimum number of range datasets. The generated surface model covers approximately all major attributes of the scene with minimal a priori knowledge about its geometry. Finally, the experimental results show the effectiveness of the proposed method.

Research on 3D reconstruction technology of large‐scale substation equipment based on NeRF

“…Owing to the in-depth research of deep neural network and the enhancement of GPU graphics processing capability, neural network is increasingly used to reconstruct models and extract features from complex 3D data [13,14]. Neural network has excellent fitting and approximation ability, and has achieved excellent results in tasks such as image registration, model reconstruction, real-time rendering, and point cloud repairing.…”