Related Experiment Video
Updated: Nov 20, 2025
03:31
Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
807
Underwater Object Detection and Reconstruction Based on Active Single-Pixel Imaging and Super-Resolution
Mengdi Li1,2, Anumol Mathai2, Stephen L H Lau2
1College of Optoelectronic Engineering, Changchun University of Science and Technology, Changchun 130022, China.
Sensors (Basel, Switzerland)
|January 20, 2021
Summary
This study introduces a novel single-pixel imaging (SPI) system using compressive sensing super-resolution convolutional neural network (CS-SRCNN) for clearer underwater object inspection. The CS-SRCNN method significantly improves image quality, offering a better alternative for underwater optical imaging.
Area of Science:
- Optical Engineering
- Image Processing
- Underwater Technology
Background:
- Underwater imaging is challenging due to light scattering and absorption.
- Single-pixel imaging (SPI) offers a potential solution for low-light conditions.
Purpose of the Study:
- To develop an improved single-pixel imaging system for underwater object inspection.
- To enhance image reconstruction quality using a novel algorithm.
Main Methods:
- Implementation of a single-pixel object inspection system for underwater environments.
- Application of a compressive sensing super-resolution convolutional neural network (CS-SRCNN) algorithm.
- Investigation of compression ratios and analysis of peak signal to noise ratio (PSNR) and structural similarity index (SSIM).
Main Results:
- Successful image reconstruction using only 30% of total pixels with CS-SRCNN.
- Significant improvements in PSNR (35.44%) and SSIM (73.07%) compared to existing methods.
- Demonstrated efficiency of the proposed method over standard SPI and SRCNN.
Conclusions:
- The CS-SRCNN based SPI system provides a high-quality imaging solution for underwater objects.
- This research offers new insights into SPI applications in challenging underwater environments.
- The proposed method presents a superior alternative for underwater optical object imaging.
