Gray-level grouping (GLG): an automatic method for optimized image contrast enhancement--Part II: the variations

Area of Science:

• Computer Vision
• Image Processing
• Digital Signal Processing

Background:

• Conventional contrast enhancement techniques have limitations.
• The basic Gray-Level Grouping (GLG) method struggles with noisy images and specific enhancement needs.
• There is a need for advanced automatic contrast enhancement methods.

Purpose of the Study:

• To introduce Selective Gray-Level Grouping (SGLG) as an extension of GLG.
• To address limitations of the basic GLG algorithm.
• To enhance a wider variety of low-contrast images, including noisy ones.

Main Methods:

• Developed Selective Gray-Level Grouping (SGLG) with varied criteria for different histogram segments.
• Introduced two new preprocessing methods for background noise elimination in low-contrast images.
• Explored the extension of (S)GLG to color image enhancement.

Main Results:

• SGLG enhances different histogram parts to varying extents, improving upon basic GLG.
• Preprocessing methods effectively eliminate background noise, enabling better (S)GLG enhancement.
• SGLG and its variations demonstrate superior performance over conventional methods for diverse low-contrast images.

Conclusions:

• SGLG significantly extends the capability of GLG for contrast enhancement.
• The enhanced methods are automatic, versatile, and outperform existing techniques.
• SGLG offers a powerful solution for various low-contrast image enhancement challenges.