A Review of Neural Radiation Field Based 3D Reconstruction Methods for Spatial Targets
摘要
With the development of space technology, optical images of space targets have become an indispensable source of information for gaining insight into the space situation and understanding the space environment. In recent years, accurate 3D reconstruction of space targets using these high-resolution optical images has become a research hotspot at the intersection of computer vision and space technology. Meanwhile, Neural Radiance Field (NeRF) utilizes micro-renderable technology to represent the scene implicitly, which has gained wide attention for its realistic visual effects. The purpose of this paper is to comprehensively review the recent progress and research results of NeRF technology in the field of 3D reconstruction of space targets. The article firstly describes the basic theory of neural radiance field, then systematically combs through and summarizes the 3D reconstruction methods of spatial targets based on this technique, then introduces the relevant image datasets, and finally looks forward to the future development and research trends.