تحديد الحافات باستخدام خوارزمية مستعمرة النمل و تطبيقها على الصور الطبية

الملخص EN

Ant Colony Optimization (ACO) is a method of heuristic search using in general artificial intelligence (swarm intelligence) to simulate the behavior of the aggregate food for ants to find new solutions to the combinatorial optimization problems.

Artificial ant's behavior depends on the trails of real ant with additional capabilities to make it more effective such as a memory to save the past events.

Every ant build solutions to the problem, and uses the information grouped about the features and performance of the private problem, to change the look to the ant problem.

In this work, an edge detection technique based on Ant Colony Optimization is used by selecting pheromone matrix which represents the information about edges in each pixel based on the guidelines set up by the ant on the image.

Multiple values for different sizes of neighbor pixels are applied and a heuristic information function to test results is proposed.

The results show high accuracy in edge detection of different biomedical images with different neighbors, the proposed algorithm is implemented in C Sharp 2008 language which provides high-efficiency software visible language and speed.

A comparative study is also given illustrating the superiority of the proposed algorithm.