A new image segmentation method based on particle swarm optimization
The paper presents a method for colour image segmentation based on a region growing procedure. The algorithm works without a priori knowledge about the number of regions in the image. The similarity of colour pixels is modelled by distance in different colour spaces (RGB, YUV and IHS). Additional median filtering and small region elimination procedures are investigated. The operation of algorithm for other variants of the RGB/IHS transformation is checked. The presented method of colour image segmentation is tested on real world images.