High order statistics based blind deconvolution of bi-level images with unknown intensity values

Area of Science:

• Image processing
• Computational imaging
• Signal processing

Background:

• Bi-level image deconvolution is crucial for image restoration.
• Existing methods like minimum entropy deconvolution and least squares minimization have limitations.
• These limitations include unrealistic statistical assumptions and the need for parameter tuning.

Purpose of the Study:

• To propose a novel linear blind deconvolution method for bi-level images.
• To develop a method that avoids common limitations of existing techniques.
• To enhance the accuracy and robustness of bi-level image deconvolution.

Main Methods:

• The proposed method employs a high-order statistics-based objective function.
• It optimizes for the point spread function and two additional parameters simultaneously.
• This approach avoids assumptions about pixel value distributions and regularization parameter tuning.

Main Results:

• The method successfully performs linear blind deconvolution on bi-level images.
• Simulations and experimental results validate the effectiveness of the proposed approach.
• Performance is demonstrated without relying on independent and identically distributed assumptions.

Conclusions:

• The novel method offers a robust alternative for bi-level image deconvolution.
• It overcomes key limitations of current deconvolution techniques.
• The approach shows promise for practical applications in image restoration.