Academic Journals Database
Disseminating quality controlled scientific knowledge

Semi-supervised Method of Multiple Object Segmentation with a Region Labeling and Flood Fill

ADD TO MY LIST
 
Author(s): Uday Pratap Singh | Kanak Saxena | Sanjeev Jain

Journal: Signal & Image Processing
ISSN 2229-3922

Volume: 2;
Issue: 3;
Start page: 175;
Date: 2011;
VIEW PDF   PDF DOWNLOAD PDF   Download PDF Original page

Keywords: oversegmentation | similar regions | Bhattacharyya distance | region merging | mean shift | flood fill.

ABSTRACT
Efficient and efficient multiple object segmentation is an important task in computer vision and objectrecognition. In this work; we address a method to effectively discover a user’s concept when multipleobjects of interest are involved in content based image retrieval. The proposed method incorporate aframework for multiple object retrieval using semi-supervised method of similar region merging and floodfill which models the spatial and appearance relations among image pixels. To improve the effectiveness ofsimilarity based region merging we propose a new similarity based object retrieval. The users only need toroughly indicate the after which steps desired objects contour is obtained during the automatic merging ofsimilar regions. A novel similarity based region merging mechanism is proposed to guide the mergingprocess with the help of mean shift technique and objects detection using region labeling and flood fill. Aregion R is merged with its adjacent regions Q if Q has highest similarity with Q (using Bhattacharyyadescriptor) among all Q’s adjacent regions. The proposed method automatically merges the regions thatare initially segmented through mean shift technique, and then effectively extracts the object contour bymerging all similar regions. Extensive experiments are performed on 12 object classes (224 images total)show promising results.
Affiliate Program      Why do you need a reservation system?