Motion estimation and quality enhancement for a single image in dynamic single-pixel imaging

Area of Science:

• Optics and Photonics
• Image Processing
• Computational Imaging

Background:

• Single-pixel imaging (SPI) typically requires numerous illuminations, limiting frame rates for dynamic scenes.
• Fast-moving objects in SPI result in reconstructed images with significant blur and noise.
• Previous research focused on inter-frame motion estimation, neglecting single-frame analysis in dynamic SPI.

Purpose of the Study:

• To develop a novel method for motion estimation and quality enhancement in dynamic single-pixel imaging using a single frame.
• To address the limitations of low frame rates and image degradation in SPI of moving objects.
• To leverage prior knowledge of object motion for improved image reconstruction.

Main Methods:

• Constructing a motion model based on prior knowledge of the target object's movement.
• Optimizing motion parameters within a defined search space.
• Implementing a deblurring technique specifically adapted for the unique motion blur mechanism in SPI.

Main Results:

• The proposed scheme effectively reduces motion blur and noise in dynamic SPI.
• Reconstructed images exhibit significantly improved quality compared to conventional methods.
• The approach demonstrates superior performance in handling motion artifacts in single-pixel imaging.

Conclusions:

• The developed method offers a viable solution for enhancing image quality in dynamic single-pixel imaging.
• Modeling object motion is crucial for overcoming limitations in capturing fast-moving subjects with SPI.
• This work advances the capabilities of SPI for applications involving dynamic scenes.