Color image segmentation without any information loss

Abstract

We can partition the background from foreground and locate the objects of interest using image segmentation techniques. In other words we can say image segmentation is the process of grouping adjacent pixels in to segments. In this research we proposed a model which can differentiate maximum and minimum frequencies for both color and grayscale images without any information loss. After getting the result of both images, we will check which (gray scale image or color image) gives better performance to the image segmentation techniques. So, here we will take the two techniques edge detection and threshold. This research gives better result of segmentation by using the relationship discontinuous and similar pixel values.