Feature Enhancement in Visually Impaired Images

Area of Science:

• Computer Vision
• Image Analysis
• Computational Physics

Background:

• Feature detection in visually impaired images remains a significant challenge in computer vision.
• Existing methods struggle with low contrast and limited dynamic range in degraded image data.

Purpose of the Study:

• To introduce and validate the Phase Stretch Transform (PST) as a solution for feature detection in visually impaired images.
• To demonstrate PST's capability for image analysis, edge detection, and resolution enhancement.

Main Methods:

• Derivation of the intrinsic nonlinear transfer function of the Phase Stretch Transform.
• Application of PST to equalize image brightness and enhance dynamic range.
• Combination of PST with conventional techniques for further improvement.

Main Results:

• PST demonstrates superior performance in low-contrast environments.
• The method achieves high dynamic range equalization in visually impaired images.
• PST enables a reconfigurable operator for hyper-dimensional classification.

Conclusions:

• Phase Stretch Transform presents a new paradigm for image analysis and feature detection in challenging visual conditions.
• The technique offers significant improvements in dynamic range and performance at low contrast.
• PST opens new avenues for computing mathematical derivatives using group delay dispersion.