Improved algorithm of non-line-of-sight imaging based on the Bayesian statistics

Area of Science:

• Optics and Photonics
• Computational Imaging
• Signal Processing

Background:

• Recovering obscured objects is a key challenge in imaging.
• Existing methods often struggle with data errors and object variability.

Purpose of the Study:

• To develop a robust non-line-of-sight (NLOS) reconstruction algorithm.
• To improve imaging quality for occluded objects using Bayesian statistics.

Main Methods:

• A novel algorithm based on Bayesian statistics was developed.
• The method integrates temporal, spatial, and intensity information from signals.
• An adjustable compensation mechanism addresses diverse object reflectivity.

Main Results:

• The algorithm demonstrates superior performance over conventional back-projection methods.
• It effectively handles random data errors, enhancing image quality.
• Efficiency was validated using both simulated and experimental data.

Conclusions:

• The proposed Bayesian NLOS algorithm offers a robust solution for imaging obscured objects.
• It provides higher quality reconstructions and adaptability to different object properties.
• Further improvements and advantages over existing techniques are discussed.