Game theory-based visual tracking approach focusing on color and texture features

Area of Science:

• Computer Vision
• Artificial Intelligence
• Game Theory

Background:

• Single-feature tracking algorithms struggle with robustness in complex environments.
• Effective feature fusion is crucial for improving tracking performance.
• Existing methods may not optimally leverage multiple features simultaneously.

Purpose of the Study:

• To propose a robust multifeature fusion tracking algorithm.
• To enhance tracking performance in challenging conditions like occlusion and scene variation.
• To utilize game theory for optimal feature contribution balancing.

Main Methods:

• Developed a multifeature fusion tracking algorithm based on game theory.
• Treated color and texture features as competing 'gamers'.
• Employed a mean shift iterative formula to find the Nash equilibrium for tracking.

Main Results:

• The algorithm demonstrated strong robustness in complex environments.
• Achieved superior performance under target occlusion and similar interference.
• Effectively balanced feature contributions for optimal fusion.

Conclusions:

• The proposed game theory-based multifeature fusion tracking algorithm significantly improves robustness.
• This approach offers a promising solution for real-world tracking challenges.
• Optimal feature fusion through game theory enhances tracking accuracy and stability.