Visible light multi-view image three-dimensional reconstruction method based on deep learning

The invention provides a visible light multi-view image three-dimensional reconstruction method based on deep learning based on improvement of an MVSNet network. A batch normalization layer and a nonlinear activation function layer in the network are replaced by a fused Inplace-ABN layer, so that the occupation amount of a video memory is reduced. And a weighted mean measurement method based on grouping similarity is designed to reduce the feature dimension of the cost body, so that a more lightweight cost body is obtained, network parameters are compressed, and the calculation amount and the video memory consumption are reduced. In order to solve the problem that the resolution of a depth map is lower than that of an input image due to the fact that an MVSNet network uses a low-scale feature map, a feature pyramid module is used for extracting a multi-scale feature map, and staged and multi-scale iterative optimization depth estimation is designed. On the premise of ensuring the precision, the average number