Automated — Seeded average approach of color segmentation for still color image and video

In segmentation we basically partition the image into various regions so that we can have particular region of interest. Color segmentation is basically used in many computer vision applications such as face recognition, object recognition etc Here we are providing the method for doing color segmentation of still and moving images over RGB (Red, Green, Blue) color space, selecting the seed automatically comparing it with the neighborhood pixels until the threshold wish we have consider a numeric non-negative value. Aim is to provide better color segmentation for the image having multiple colors, images with light colors, images with multiple objects, images with irregular shaped objects. To convert an image into a portrait by the term portrait we mean that providing realistic painting view to color image which identify and segment the image into various regions of different colors. Identifying various colors in the image and providing better color segmentation to a still and moving objects. To remove unwanted noise, better edge detection and convert a movie clip into an animation by working over RGB color space and giving the artistic effect to an image.