Particle Swarm Contour Search Algorithm.

Area of Science:

• Computer Vision
• Artificial Intelligence
• Optimization Algorithms

Background:

• Traditional contour-finding algorithms often rely on image processing and complete search space knowledge, limiting real-world applicability.
• Many applications lack complete environmental information, posing challenges for existing contour detection methods.

Purpose of the Study:

• Introduce a novel algorithm, Particle Swarm Contour Search (PSCS), inspired by Particle Swarm Optimisation.
• Develop a contour-finding method that operates effectively using only local information, removing the need for a complete search space overview.

Main Methods:

• The Particle Swarm Contour Search (PSCS) algorithm utilizes a swarm of particles to explore and identify object contours in 2D environments.
• Particles leverage local information regarding their positions relative to the object (inside/outside) to search for and traverse contours.

Main Results:

• The PSCS algorithm demonstrates the ability to accurately identify object contours.
• Experimental results indicate that PSCS achieves performance comparable to state-of-the-art contour-finding techniques.

Conclusions:

• The PSCS algorithm offers a viable alternative for contour detection, particularly in scenarios where complete search space information is unavailable.
• This approach enhances the feasibility of object contour detection in real-world applications by relying on local data.