OBJECT CONTOUR COMPLETION BY COMBINING OBJECT RECOGNITION AND LOCAL EDGE CUES

Abstract

We developed a top-down and bottom-up segmentation ofobjects using shape contours through a two-stage procedure. First, the object was identified using an edge-based contour feature and then the object contour was obtained using a constraint optimization procedure based on the results from the earlier identified contours. The initial object detection provides object category specific information for the contour completion to be effected. We argue that top-down bottom-up interaction architecture has plausible neurological correlates. This method has an advantage in that it does not require learning boundaries with large datasets.  

Contributors

Kar Seng Loke: Malaysia

Journal

Title: Journal of Information and Communication Technology

Publisher: UUM Press

Publication date (Electronic): November 06 2017

Volume: 16

Pages: 224-242

Affiliations

[1 ]School of Information Technology Swinburne University of Technology Sarawak Campus, Malaysia

Article

Internal ID: 8230

DOI: 10.32890/jict2017.16.2.8230

License:

All content is freely available without charge to users or their institutions. Users are allowed to read, download, copy, distribute, print, search, or link to the full texts of the articles in this journal without asking prior permission of the publisher or the author. Articles published in the journal are distributed under a http://creativecommons.org/licenses/by/4.0/.

Product

Self URI (journal-page): http://e-journal.uum.edu.my/index.php/jict/about

Self URI (article-page): http://e-journal.uum.edu.my/index.php/jict/article/view/8230