Title: A comparative analysis of SIFT, SURF and ORB on sketch and paint based images

Authors: R. Radha; M. Pushpa

Addresses: Department of Computer Science, Quaid-E-Millath Government College for Women (Autonomous), Chennai, 600 002, Tamilnadu, India ' Department of Computer Science, Quaid-E-Millath Government College for Women (Autonomous), Chennai, 600 002, Tamilnadu, India

Abstract: Image retrieval has been one of the most interesting and emergent research areas in the field of computer vision. Content-based image retrieval (CBIR) systems are used in order to automatically index, search, retrieve and browse images from the databases. Content-based image retrieval system consider colour and texture features of the image, however those features are different in transformed images even though it is the toughest challenge for the CBIR to understand the image. Human perspective that has been based on input is essential for any retrieval system. Hand drawing images and painting images are considered as query image for this retrieval system. This paper has explored few eminent feature extraction techniques like scale invariant feature transform (SIFT), speeded up robust features (SURF) and oriented FAST and rotated BRIEF (ORB) as well as the performances of these techniques for sketch and paint based images. The suitable extraction technique is identified by this examination, the significance of SIFT, SURF and ORB features are listed.

Keywords: CBIR; content-based image retrieval; KNN; K nearest neighbour; ORB; oriented FAST and rotated BRIEF; scale invariant feature transform; SIFT; SURF; speeded up robust features.

DOI: 10.1504/IJFE.2021.118910

International Journal of Forensic Engineering, 2021 Vol.5 No.2, pp.102 - 110

Received: 28 May 2019
Accepted: 15 Feb 2020

Published online: 11 Nov 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article