An adaptive image enhancement technique by combining cuckoo search and particle swarm optimization algorithm.

Area of Science:

• Image Processing and Computer Vision
• Artificial Intelligence
• Optimization Algorithms

Background:

• Image enhancement is crucial for image analysis and interpretation.
• Low contrast images present challenges in visual perception and automated analysis.
• Existing enhancement methods may lack adaptivity or optimal performance.

Purpose of the Study:

• To develop a novel adaptive image enhancement technique for low contrast images.
• To improve image contrast and preserve details using an optimized transformation.
• To evaluate the proposed method against established and evolutionary-based techniques.

Main Methods:

• A modified measure and blending of cuckoo search and particle swarm optimization (CS-PSO) were employed.
• An incomplete Beta function was used as the global transformation function.
• A novel image quality criterion incorporating threshold, entropy, and gray-level probability density was developed.

Main Results:

• The proposed CS-PSO method adaptively enhanced image contrast and details.
• The method demonstrated robustness and superior performance compared to linear contrast stretching, histogram equalization, and other evolutionary algorithms.
• Quantitative and qualitative evaluations confirmed the effectiveness of the proposed technique.

Conclusions:

• The developed CS-PSO technique offers an effective and adaptive solution for enhancing low contrast images.
• The novel image quality criterion and optimization approach contribute to improved image enhancement.
• This method shows significant potential for various image processing and analysis applications.