Convex Shape Retrieval from Edge Maps by the use of an Evolutionary Algorithm

There is a need for a high-throughput approach for extracting biological shapes from images. The approach for automated extraction of convex biological shapes presented in this paper is an Evolutionary Algorithm. As opposed to existing model based segmentation methods this approach is uniform for different images, needs no training set and is initialized automatically. The process of finding the shape is considered an optimization problem and for that reason an Evolutionary Algorithm was a good candidate for a solution. The results show that the proposed Evolutionary Algorithm gives a fast solution for pattern recognition and shape extraction.