Asian Journal of Engineering and Technology Innovation

A Survey On Comparison of Fast Nonparametric clustering algorithm

Chaitali Choudhury, Nirmala Guptha

School of Computing and Information Technology, REVA University, Bangalore, INDIA.

Abstract

The Gaussian Blurring Mean Shift (GBMS) is a clustering algorithm, a procedure that iteratively sharpens a dataset my moving each data point according to Gaussian Mean Shift (GMS). (1) Here a criterion has been used, that stops the procedure as soon as the clusters are formed which shows reliable image segmentation. (2) The results being compared with GMS which shows that GBMS (connecting spectral clustering) produces better result. (3) Further GBMS is accelerated by interleaving connection component and blurring step.

Keywords:

GBMS, GMS, spectral clustering