Related Experiment Video
Updated: Jul 9, 2025
11:34
High-resolution, High-speed, Three-dimensional Video Imaging with Digital Fringe Projection Techniques
Published on: December 3, 2013
15.7K
Projecting colorful images through scattering media via deep learning
Optics Express
|November 29, 2023
Summary
This study introduces a projector network capable of displaying colorful images through scattering media like fog or turbid water. This breakthrough overcomes limitations of previous methods, enabling full-color image projection in challenging environments.
Area of Science:
- Optics
- Image Processing
- Computational Imaging
Background:
- Light scattering by particles obstructs image projection through media like fog, water, and clouds.
- Previous wavefront shaping and neural network techniques primarily projected grayscale images using monochromatic light due to wavelength control complexity.
Purpose of the Study:
- To develop a projector network for projecting full-color images through scattering media.
- To overcome the limitations of monochromatic and grayscale image projection in scattering environments.
Main Methods:
- Development of a novel projector network architecture.
- Experimental validation using the MINST dataset and two stacked diffusers to simulate scattering.
- Quantitative evaluation using the averaged intensity Pearson's correlation coefficient.
Main Results:
- Successful projection of colorful images through scattering media.
- Achieved an averaged intensity Pearson's correlation coefficient of approximately 90.6% for 1,000 test images.
- Demonstrated the superiority of the developed network over prior methods.
Conclusions:
- The developed projector network effectively enables full-color image projection through scattering media.
- This technology has significant potential for display applications in environments with scattering.
- The method addresses the challenge of simultaneous multi-wavelength control in scattering compensation.
