Image sharpness assessment based on local phase coherence.

Area of Science:

• Computer Vision
• Image Processing
• Perceptual Quality

Background:

• Image sharpness is crucial for visual quality assessment.
• Current algorithms often rely on edge width or frequency content.
• The human visual system's mechanism for sharpness detection is not fully understood.

Purpose of the Study:

• To propose a new perspective on image sharpness based on local phase coherence (LPC).
• To develop a flexible framework for LPC computation across arbitrary fractional scales.
• To create a novel sharpness assessment algorithm independent of the original image.

Main Methods:

• Utilizing the complex wavelet transform domain to evaluate image features.
• Implementing a flexible framework for local phase coherence (LPC) computation.
• Developing a no-reference image sharpness assessment algorithm.

Main Results:

• The proposed algorithm demonstrates competitive performance against state-of-the-art methods.
• Evaluation was conducted on four publicly available, subject-rated image databases.
• The method identifies sharpness through strong local phase coherence near distinctive image features.

Conclusions:

• Local phase coherence in the complex wavelet domain offers a robust approach to image sharpness assessment.
• The developed no-reference algorithm provides a flexible and effective solution for image quality evaluation.
• This work advances the understanding of visual perception and image quality metrics.