Full-waveform LiDAR echo decomposition based on dense and residual neural networks

Paper and Details:

https://doi.org/10.1364/AO.444910

Architecture:

Dataset and Pre-trained Models:

Google Drive: https://drive.google.com/file/d/1jI9kVC4dHZewRxjOZIttKfcPJo1Dgd53/view?usp=sharing

or Google Drive2:https://drive.google.com/file/d/1xV7iCuQxGroAsB3Cxdwi-IqgbHVC2aXI/view?usp=sharing

New Link:

https://drive.google.com/file/d/1vo25UKvMrFQOoyk5m9w9wpJUEMSUcl_W/view?usp=drive_link

Training Details:

Usage:

Train: <python train.py --snr 30/24/18/12 --model_select 0/1 ### 0: FDCN 1: FDRN >
Test: <python test.py >

Some Result:

FDCN:

FDRN:

Real World Experiment:

Cite

if you find this work useful, please cite : @article{liu2022full, title={Full-waveform LiDAR echo decomposition based on dense and residual neural networks}, author={Liu, Gangping and Ke, Jun}, journal={Applied Optics}, volume={61}, number={9}, pages={F15--F24}, year={2022}, publisher={Optical Society of America} }

Thanks a lot!