Combination of histogram of oriented gradients and hierarchical centroid for sketch-based image retrieval

Norah M. Asiri; Nourah AlHumaidi; Nourah AlOsaim

doi:10.1109/CSCESM.2015.7331884

Back

Conference proceeding

Combination of histogram of oriented gradients and hierarchical centroid for sketch-based image retrieval

Norah M. Asiri, Nourah AlHumaidi and Nourah AlOsaim

2015 Second International Conference on Computer Science, Computer Engineering, and Social Media (CSCESM), pp.149-152

01/09/2015

DOI: https://doi.org/10.1109/CSCESM.2015.7331884

Abstract

hierarchical centroid

HOG

Image edge detection

Multimedia communication

shape properties

sketch-based retrieval

The sketch based image retrieval (SBIR) systems is still little argued and debated. Histogram of oriented gradients (HOG) has shown good performance in sketch based retrieval systems over all other methods. In this paper, we propose an enhancement of HOG by using HOG combined with hierarchical centroid of irregular shape. We also compare the result with original HOG, edge histogram and the shape properties descriptors. The system uses free-hand sketched queries as input and matches it with images from different 12 classes.

Metrics

1 Record Views

Details

Title: Combination of histogram of oriented gradients and hierarchical centroid for sketch-based image retrieval
Creators - without role: Norah M. Asiri - King Saud University
Nourah AlHumaidi - King Saud University
Nourah AlOsaim - King Saud University
Publication Details: 2015 Second International Conference on Computer Science, Computer Engineering, and Social Media (CSCESM), pp.149-152
Publisher: IEEE
Identifiers: 9951999708331
Academic Unit: King Saud University
Language: English
Resource Type: Conference proceeding