This toolkit has everything you need to perform imaging with a lensless camera. The sensor in most examples is the Raspberry Pi HQ, camera sensor as it is low cost (around 50 USD) and has a high resolution (12 MP). The lensless encoder/mask used in most examples is either a piece of tape or a low-cost LCD. As modularity is a key feature of this toolkit, we try to support different sensors and/or lensless encoders.
The toolkit includes:
- Training scripts/configuration for various learnable, physics-informed reconstruction approaches, as shown here.
- Camera assembly tutorials (link).
- Measurement scripts (link).
- Dataset preparation and loading tools, with Hugging Face integration (slides on uploading a dataset to Hugging Face with this script).
- Reconstruction algorithms (e.g. FISTA, ADMM, unrolled algorithms, trainable inversion, , multi-Wiener deconvolution network, pre- and post-processors).
- Pre-trained models that can be loaded from Hugging Face, for example in this script.
- Mask design and fabrication tools.
- Simulation tools.
- Evalutions tools (e.g. PSNR, LPIPS, SSIM, visualizations).
- Demo that can be run on Telegram!
Please refer to the documentation for more details, while an overview of example notebooks can be found here.
We've also written a few Medium articles to guide users through the process of building the camera, measuring data with it, and reconstruction. They are all laid out in this post.
The following works have been implemented in the toolkit:
Reconstruction algorithms:
- ADMM with total variation regularization and 3D support (source code, usage). [1]
- Unrolled ADMM (source code, usage). [2]
- Unrolled ADMM with compensation branch (source code, usage). [3]
- Trainable inversion from Flatnet (source code, usage). [4]
- Multi-Wiener deconvolution network (source code, usage). [5]
- SVDeconvNet (for learning multi-PSF deconvolution) from PhoCoLens (source code, usage). [6]
- Incorporating pre-processor (source code). [7]
- Accounting for external illumination(source code 1, source code 2, usage). [8]
Camera / mask design:
- Fresnel zone aperture mask pattern (source code). [9]
- Coded aperture mask pattern (source code). [10]
- Near-field Phase Retrieval for designing a high-contrast phase mask (source code). [11]
- LCD-based camera, i.e. DigiCam (source code). [7]
Datasets (hosted on Hugging Face and downloaded via their API):
- DiffuserCam Lensless MIR Flickr dataset (copy on Hugging Face). [2]
- TapeCam MIR Flickr (Hugging Face). [7]
- DigiCam MIR Flickr (Hugging Face). [7]
- DigiCam MIR Flickr with multiple mask patterns (Hugging Face). [7]
- DigiCam CelebA (Hugging Face). [7]
- MultiFocal mask MIR Flickr under external illumination (Hugging Face). [8] Mask fabricated by [12]
If you are just interested in using the reconstruction algorithms and
plotting / evaluation tools you can install the package via pip:
pip install lenslessFor plotting, you may also need to install Tk.
For performing measurements, the expected workflow is to have a local computer which interfaces remotely with a Raspberry Pi equipped with the HQ camera sensor (or V2 sensor). Instructions on building the camera can be found here.
The software from this repository has to be installed on both your local machine and the Raspberry Pi. Note that we recommend using Python 3.11, as some Python library versions may not be available with earlier versions of Python. Moreover, its end-of-life is Oct 2027.
Below are commands that worked for our configuration (Ubuntu 22.04.5 LTS), but there are certainly other ways to download a repository and install the library locally.
Note that (lensless) is a convention to indicate that the virtual
environment is activated. After activating your virtual environment, you only
have to copy the command after (lensless).
# download from GitHub
git clone [email protected]:LCAV/LenslessPiCam.git
cd LenslessPiCam
# create virtual environment (as of Oct 4 2023, rawpy is not compatible with Python 3.12)
# -- using conda
conda create -n lensless python=3.11
conda activate lensless
# -- OR venv
python3.11 -m venv lensless_env
source lensless_env/bin/activate
# install package
(lensless) pip install -e .
# extra dependencies for local machine for plotting/reconstruction
(lensless) pip install -r recon_requirements.txt
# pre-commit hooks for code formatting
(lensless) pip install pre-commit black
(lensless) pre-commit install
# (optional) try reconstruction on local machine
(lensless) python scripts/recon/admm.py
# (optional) try reconstruction on local machine with GPU
(lensless) python scripts/recon/admm.py -cn pytorchNote (25-04-2023): for using the :py:class:`~lensless.recon.apgd.APGD` reconstruction method based on Pycsou (now Pyxu), a specific commit has to be installed (as there was no release at the time of implementation):
pip install git+https://github.com/matthieumeo/pycsou.git@38e9929c29509d350a7ff12c514e2880fdc99d6eIf PyTorch is installed, you will need to be sure to have PyTorch 2.0 or higher, as Pycsou is not compatible with earlier versions of PyTorch. Moreover, Pycsou requires Python within [3.9, 3.11).
Moreover, numba (requirement for Pycsou V2) may require an older version of NumPy:
pip install numpy==1.23.5After flashing your Raspberry Pi with SSH enabled, you need to set it up for passwordless access. Do not set a password for your SSH key pair, as this will not work with the provided scripts.
On the Raspberry Pi, you can then run the following commands (from the home
directory):
# dependencies
sudo apt-get install -y libimage-exiftool-perl libatlas-base-dev \
python3-numpy python3-scipy python3-opencv
sudo pip3 install -U virtualenv
# download from GitHub
git clone [email protected]:LCAV/LenslessPiCam.git
# install in virtual environment
cd LenslessPiCam
virtualenv --system-site-packages -p python3 lensless_env
source lensless_env/bin/activate
pip install --no-deps -e .
pip install -r rpi_requirements.txt
# test on-device camera capture (after setting up the camera)
(lensless_env) python scripts/measure/on_device_capture.pyYou may still need to manually install numpy and/or scipy with pip in case libraries (e.g. libopenblas.so.0) cannot be detected.
The idea of building a lensless camera from a Raspberry Pi and a piece of tape comes from Prof. Laura Waller's group at UC Berkeley. So a huge kudos to them for the idea and making tools/code/data available! Below is some of the work that has inspired this toolkit:
A few students at EPFL have also contributed to this project:
- Julien Sahli: support and extension of algorithms for 3D.
- Yohann Perron: unrolled algorithms for reconstruction.
- Aaron Fargeon: mask designs.
- Rein Bentdal and David Karoubi: mask fabrication with 3D printing.
- Stefan Peters: imaging under external illumination.
We also thank the Swiss National Science Foundation for funding this project through the Open Research Data (ORD) program.
If you use this toolkit in your own research, please cite the following:
@article{Bezzam2023,
doi = {10.21105/joss.04747},
url = {https://doi.org/10.21105/joss.04747},
year = {2023},
publisher = {The Open Journal},
volume = {8},
number = {86},
pages = {4747},
author = {Eric Bezzam and Sepand Kashani and Martin Vetterli and Matthieu Simeoni},
title = {LenslessPiCam: A Hardware and Software Platform for Lensless Computational Imaging with a Raspberry Pi},
journal = {Journal of Open Source Software}
}
The following papers have contributed new approaches to the field of lensless imaging:
- Introducing pre-processor component as part of modular reconstruction (IEEE Transactions on Computational Imaging and IEEE International Conference on Image Processing (ICIP) 2024):
@ARTICLE{10908470,
author={Bezzam, Eric and Perron, Yohann and Vetterli, Martin},
journal={IEEE Transactions on Computational Imaging},
title={Towards Robust and Generalizable Lensless Imaging With Modular Learned Reconstruction},
year={2025},
volume={11},
number={},
pages={213-227},
keywords={Training;Wiener filters;Computational modeling;Transfer learning;Computer architecture;Cameras;Transformers;Software;Software measurement;Image reconstruction;Lensless imaging;modularity;robustness;generalizability;programmable mask;transfer learning},
doi={10.1109/TCI.2025.3539448}
}
@INPROCEEDINGS{10647433,
author={Perron, Yohann and Bezzam, Eric and Vetterli, Martin},
booktitle={2024 IEEE International Conference on Image Processing (ICIP)},
title={A Modular and Robust Physics-Based Approach for Lensless Image Reconstruction},
year={2024},
volume={},
number={},
pages={3979-3985},
keywords={Training;Multiplexing;Pipelines;Noise;Cameras;Robustness;Reproducibility of results;Lensless imaging;modular reconstruction;end-to-end optimization},
doi={10.1109/ICIP51287.2024.10647433}
}
- Lensless imaging under external illumination (IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2025):
@INPROCEEDINGS{10888030,
author={Bezzam, Eric and Peters, Stefan and Vetterli, Martin},
booktitle={ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
title={Let There Be Light: Robust Lensless Imaging Under External Illumination With Deep Learning},
year={2025},
volume={},
number={},
pages={1-5},
keywords={Source separation;Noise;Lighting;Interference;Reconstruction algorithms;Cameras;Optics;Speech processing;Image reconstruction;Standards;lensless imaging;ambient lighting;external illumination;background subtraction;learned reconstruction},
doi={10.1109/ICASSP49660.2025.10888030}
}
| [1] | Antipa, N., Kuo, G., Heckel, R., Mildenhall, B., Bostan, E., Ng, R., & Waller, L. (2017). DiffuserCam: lensless single-exposure 3D imaging. Optica, 5(1), 1-9. |
| [2] | (1, 2, 3) Monakhova, K., Yurtsever, J., Kuo, G., Antipa, N., Yanny, K., & Waller, L. (2019). Learned reconstructions for practical mask-based lensless imaging. Optics express, 27(20), 28075-28090. |
| [3] | Zeng, T., & Lam, E. Y. (2021). Robust reconstruction with deep learning to handle model mismatch in lensless imaging. IEEE Transactions on Computational Imaging, 7, 1080-1092. |
| [4] | Khan, S. S., Sundar, V., Boominathan, V., Veeraraghavan, A., & Mitra, K. (2020). Flatnet: Towards photorealistic scene reconstruction from lensless measurements. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(4), 1934-1948. |
| [5] | Li, Y., Li, Z., Chen, K., Guo, Y., & Rao, C. (2023). MWDNs: reconstruction in multi-scale feature spaces for lensless imaging. Optics Express, 31(23), 39088-39101. |
| [6] | Cai, X., You, Z., Zhang, H., Gu, J., Liu, W., & Xue, T. (2024). Phocolens: Photorealistic and consistent reconstruction in lensless imaging. Advances in Neural Information Processing Systems, 37, 12219-12242. |
| [7] | (1, 2, 3, 4, 5, 6) Bezzam, E., Perron, Y., & Vetterli, M. (2025). Towards Robust and Generalizable Lensless Imaging with Modular Learned Reconstruction. IEEE Transactions on Computational Imaging. |
| [8] | (1, 2) Bezzam, E., Peters, S., & Vetterli, M. (2024). Let there be light: Robust lensless imaging under external illumination with deep learning. IEEE International Conference on Acoustics, Speech and Signal Processing. |
| [9] | Wu, J., Zhang, H., Zhang, W., Jin, G., Cao, L., & Barbastathis, G. (2020). Single-shot lensless imaging with fresnel zone aperture and incoherent illumination. Light: Science & Applications, 9(1), 53. |
| [10] | Asif, M. S., Ayremlou, A., Sankaranarayanan, A., Veeraraghavan, A., & Baraniuk, R. G. (2016). Flatcam: Thin, lensless cameras using coded aperture and computation. IEEE Transactions on Computational Imaging, 3(3), 384-397. |
| [11] | Boominathan, V., Adams, J. K., Robinson, J. T., & Veeraraghavan, A. (2020). Phlatcam: Designed phase-mask based thin lensless camera. IEEE transactions on pattern analysis and machine intelligence, 42(7), 1618-1629. |
| [12] | Lee, K. C., Bae, J., Baek, N., Jung, J., Park, W., & Lee, S. A. (2023). Design and single-shot fabrication of lensless cameras with arbitrary point spread functions. Optica, 10(1), 72-80. |
