Compressed Imaging Reconstruction Based on Block Compressed Sensing with Conjugate Gradient Smoothed l₀ Norm

Area of Science:

• Optics and Imaging
• Signal Processing
• Computer Vision

Background:

• Compressed imaging enables high-resolution image reconstruction from limited observations.
• Reconstruction algorithm accuracy is critical in compressed imaging systems.
• Block compressed sensing (BCS) is applied to traditional optical imaging.

Purpose of the Study:

• To design a novel reconstruction algorithm for compressed imaging.
• To improve reconstruction accuracy and reduce block effects.
• To enhance the efficiency of image reconstruction.

Main Methods:

• Developed a conjugate gradient smoothed l0 norm (BCS-CGSL0) algorithm based on block compressed sensing.
• The CGSL0 component optimizes the SL0 algorithm using a novel inverse triangular function and conjugate gradient method.
• The BCS-SPL method is integrated to mitigate block effects within the BCS framework.

Main Results:

• The BCS-CGSL0 algorithm effectively reduces block effects in reconstructed images.
• Demonstrated significant improvements in reconstruction accuracy compared to existing methods.
• Showcased enhanced reconstruction efficiency.

Conclusions:

• The BCS-CGSL0 algorithm offers superior performance in compressed imaging reconstruction.
• This method provides a robust solution for achieving high-resolution images with fewer observations.
• The study validates the advantages of BCS-CGSL0 in accuracy and efficiency through simulations.