Multispectral Image Out-of-Focus Deblurring Using Interchannel Correlation.

Area of Science:

• Multispectral Imaging
• Image Processing
• Computational Optics

Background:

• Out-of-focus blur is a common issue in multispectral imaging.
• Wavelength-dependent focal lengths cause varying blur levels across spectral channels.

Purpose of the Study:

• To propose a novel multispectral image deblurring framework.
• To restore spectral images suffering from out-of-focus blur using interchannel correlation (ICC).

Main Methods:

• Classifying spectral images into focused and out-of-focus sets.
• Utilizing well-focused images as priors for deblurring out-of-focus images.
• Modeling blur as a Gaussian point spread function and employing generalized cross-validation for parameter determination.

Main Results:

• The proposed framework effectively restores out-of-focus spectral images.
• The method demonstrates superior performance compared to existing state-of-the-art techniques.
• No manual parameter tuning is required due to generalized cross-validation.

Conclusions:

• The interchannel correlation (ICC) based framework successfully addresses multispectral image deblurring.
• The method offers an efficient and accurate solution for restoring spectral image quality.
• The approach generalizes well and outperforms current deblurring techniques.