文章基本信息

标题：A Survey on Various Techniques for Image Segmentation
本地全文：下载
作者：Rajiv Bansal ; Sona Kajla
期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2016
卷号：4
期号：4
页码：6715
DOI：10.15680/IJIRCCE.2016.0404061
出版社：S&S Publications
摘要：Image segmentation is the process through which an image is divided into multiple segments. The aim of segmentation is to change the features and representation of an image according to our need or into something that is more meaningful and easy to analys e. Using image segmentation we can classify an image into different group. Till now many research has been done using clustering in the area of image segmentation. There are different methods of clustering e.g. k - means clustering, c - means cluster ing, fuzzy c - means and many more but k - means is one of the most popular methods of clustering. K - means is an unsupervised clustering algorithm. There are many features of an image with which we can improve image and two of them are PSNR (peak signal to noi se ratio) and RMSE ( root mean square error). To improve image, PSNR rate must be increased and RMSE rate must be decreased. When the mean square error is decreased the quality of image will automatically increased. Varieties of clustering algorithm are us ed to improve these features. Different clustering algorithm will give different values for PSNR and RMSE rate. In this paper, we are doing a survey of different clustering algorithm in various application of image segmentation. Mainly image segmentation i s used in medical science to locate disease, tumo u r, cancer and many other problems. Some of their shortcomings are also mentioned for further development
关键词：image segmentation; k-means clustering; PSNR rate; RMSE rate