A color image contrast enhancement method based on improved PSO.

Area of Science:

• Computer Vision
• Artificial Intelligence
• Image Processing

Background:

• Image contrast enhancement is crucial for extracting information from images.
• Traditional methods often struggle with complex image features and optimization challenges.
• Viewing image enhancement as an optimization problem offers a new perspective.

Purpose of the Study:

• To develop an improved particle swarm optimization (PSO) algorithm for image contrast enhancement.
• To optimize image transformation parameters for superior contrast and brightness.
• To evaluate the proposed algorithm's performance against existing methods.

Main Methods:

• An improved PSO algorithm incorporating individual, local, and global optimization strategies.
• Utilizing topology for particle communication and a sparse penalty term for solution space adjustment.
• Representing RGB color channels using quaternion matrices for parameter optimization.
• Integrating contrast and brightness into the fitness function to guide optimization.

Main Results:

• The improved PSO algorithm demonstrated significant performance gains in benchmark function tests, with average value increases of at least 15x and 1.3x.
• When applied to image contrast enhancement, the proposed algorithm showed at least a 5% improvement in performance indicators and a 15% reduction in running time compared to other evolutionary algorithms.
• Subjective and qualitative evaluations confirmed the superiority of the proposed method.

Conclusions:

• The developed improved PSO algorithm is highly effective for image contrast enhancement.
• The method offers significant advantages in terms of enhancement quality, optimization efficiency, and computational speed.
• This approach provides a robust framework for image enhancement tasks.