Novel Image Segmentation through Agile Expanse Merging

Abstract

This paper gives the explanation of novel image segmentation by agile expanse merging. Initially, the image is segmented into many expanses and the expanses of the same colour are continuously merged by the proposed algorithm. The proposed algorithm for expanse merging process will seek for homogeneity among the expanses and merges it iteratively by Sequential Probability Ratio Test (SPRT) and minimal cost condition. This merging algorithm combines the expanses in an orderly manner and stops the merging according to the criteria which is set by SPRT TEST. The merging done here goes on the agile programming technique. The final segmentation is done based on the studied or perceived image after merging process. This algorithm acquires the global standards especially the speed in which the merging is carried out by using Nearest Neighbor Graph Method (NNG).