isroalka/hformer

Hformer: Hybrid CNN-Transformer for Single Wrapped Phase to Fringe Order Prediction Fringe Projection 3D measurement

Xinjun Zhu, Zhiqiang Han, Mengkai Yuan, Qinghua Guo, Hongyi Wang, Limei Song, "Hformer: hybrid convolutional neural network transformer network for fringe order prediction in phase unwrapping of fringe projection," Opt. Eng. 61(9), 093107 (2022), doi: 10.1117/1.OE.61.9.093107.

Note

• The code is tested with python=3.7 and paddlepaddle=2.0.2

Begin to train

• python PaddleSeg/train.py --config PaddleSeg/configs/Hformer/Hformer.yml --do_eval --use_vdl

Begin to test

• python PaddleSeg/predict.py --config PaddleSeg/configs/Hformer/Hformer.yml --model_path output/best_model/model.pdparams --image_path dataset path --save_dir result --aug_pred --is_slide --crop_size 384 384 --stride 256 96

Begin to test with horizontal flip and vertical flip

• python PaddleSeg/predict.py --config PaddleSeg/configs/Hformer/Hformer.yml --model_path output/best_model/model.pdparams --image_path dataset path --save_dir result --aug_pred --is_slide --crop_size 384 384 --stride 256 96 --flip_horizontal --flip_vertical

Requirements

• pip install -r requirements.txt

References of datasets

• [1] Qian, J., Feng, S., Tao, T., Hu, Y., Li, Y., Chen, Q., Zuo, C.: Deep learning-enabled geometric constraints and phase unwrapping for single shot absolute 3d shape measurement. Apl Photonics 5(4), 046105 (2020). https://doi.org/10.1063/5.0003217
• [2] Zuo, C., Qian, J., Feng, S. et al. Deep learning in optical metrology: a review. Light Sci Appl 11, 39 (2022). https://doi.org/10.1038/s41377-022-00714-x

Acknowledgement

• codebase of CAT block from https://github.com/linhezheng19/CAT
• codebase of Patch Expanding from https://github.com/HuCaoFighting/Swin-Unet

Citation

Xinjun Zhu, Zhiqiang Han, Mengkai Yuan, Qinghua Guo, Hongyi Wang, Limei Song, "Hformer: hybrid convolutional neural network transformer network for fringe order prediction in phase unwrapping of fringe projection," Opt. Eng. 61(9), 093107 (2022), doi: 10.1117/1.OE.61.9.093107.