Skip to content

Latest commit

 

History

History
22 lines (18 loc) · 894 Bytes

README.md

File metadata and controls

22 lines (18 loc) · 894 Bytes
Raw

3D-depth-prediction

3D depth prediction based on neural network and coded Light

基于语义分割模型（本repository以U-Net模型为baseline），实现单帧条纹图到深度图的预测
Single-frame fringe map to depth map prediction based on the semantic segmentation model (this repository uses U-Net model as baseline)

Prerequisites：

PyTorch 1.8.x+cu102
NVIDIA CUDA 11.6

How to implement model training:

  1. Preprocess pics data (512pixel or others) and label data.
    path is './Dataset/image' and './Dataset/seg_image'
  2. Customize the input and output of the network layer.
    Run train.py and predict.py by turn.

author read:

file copy from: PythonSpace/labelme_test
20220905165348
refer: 多阶段深度学习单帧条纹投影三维测量方法