A parallel approach for region-growing segmentation

Title

A parallel approach for region-growing segmentation

Creator

Baby A.S.; Balachandran K.

Description

Image Segmentations play a heavy role in areas such as computer vision and image processing due to its broad usage and immense applications. Because of the large importance of image segmentation a number of algorithms have been proposed and different approaches have been adopted. In this theme I tried to parallelize the image segmentation using a region growing algorithm. The primary goal behind this theme is to enhance performance or speed up the image segmentation on large volume image data sets, i.e. Very high resolution images (VHR). In parliamentary law to get the full advantage of GPU computing, equally spread the workload among the available threads. Threads assigned to individual pixels iteratively merge with adjacent segments and always ensuring the standards that the heterogeneity of image objects should be belittled. An experimental analysis upon different orbital sensor images has made out in order to assess the quality of results. 2015 IEEE.

Source

Conference Proceeding - 2015 International Conference on Advances in Computer Engineering and Applications, ICACEA 2015, pp. 196-200.

Date

2015-01-01

Publisher

Institute of Electrical and Electronics Engineers Inc.

Subject

Graphical Processing Unit; Image Segmentation; Parallel Processing; Region Growing

Coverage

Baby A.S., Dept. of CSE, Christ University, Bengaluru, India; Balachandran K., Dept. of CSE, Christ University, Bengaluru, India

Rights

Restricted Access

Relation

ISBN: 978-146736911-4

Format

Online

Language

English

Type

Conference paper

Identifier

https://doi.org/10.1109/ICACEA.2015.7164694

https://www.scopus.com/inward/record.uri?eid=2-s2.0-84947996743&doi=10.1109%2fICACEA.2015.7164694&partnerID=40&md5=abc1e9b0378914dc8109ac581482a7cb