Hybrid Adaptive Crayfish Optimization with Differential Evolution for Color Multi-Threshold Image Segmentation

Area of Science:

• Computer Vision
• Artificial Intelligence
• Image Processing

Background:

• Multi-threshold image segmentation is crucial for image analysis but faces challenges with algorithm convergence.
• Existing optimization algorithms, like the basic crayfish optimization algorithm, struggle to find optimal solutions in later stages.

Purpose of the Study:

• To propose a novel hybrid algorithm, adaptive crayfish optimization with differential evolution (ACOADE), for color multi-threshold image segmentation.
• To enhance the convergence and exploration capabilities of the crayfish optimization algorithm for improved segmentation results.

Main Methods:

• ACOADE optimizes the foraging parameter 'p' and incorporates an adaptive strategy to boost randomness.
• The differential evolution (DE) algorithm's core formula is integrated to balance exploration and exploitation.
• Performance validated using IEEE CEC2020 test functions and compared against eight other algorithms.
• Effectiveness for image segmentation assessed using Kapur entropy and Otsu methods, with comparisons to eight other algorithms.

Main Results:

• ACOADE demonstrated superior performance in objective function values compared to benchmark algorithms.
• Image segmentation quality metrics, including Peak Signal-to-Noise Ratio (PSNR), Feature Similarity Index Measure (FSIM), and Structural Similarity Index Measure (SSIM), showed significant improvements.
• The Wilcoxon signed-rank test confirmed ACOADE's statistical superiority.

Conclusions:

• ACOADE offers significant advantages in objective function value, image quality, convergence speed, and robustness for multi-threshold image segmentation.
• The hybrid approach effectively addresses the limitations of traditional crayfish optimization algorithms.
• ACOADE represents a promising advancement for color image segmentation tasks.