A Journey from basic Image Features to Lofty Human Intelligence in Content-based Image Retrieval: Motivation, Applications and Future Trends
Shikha Bhardwaj1, Gitanjali Pandove2, Pawan Kumar Dahiya3
1Shikha Bhardwaj*, ECE Department, DCRUST, Murthal and UIET, Kurukshetra University, Haryana, India.
2Gitanjali Pandove, ECE Department, DCRUST, Murthal, Sonepat, Haryana, India.
3Pawan Kumar Dahiya, ECE Department, DCRUST, Murthal, Sonepat, Haryana, India.
Manuscript received on May 25, 2020. | Revised Manuscript received on June 29, 2020. | Manuscript published on July 30, 2020. | PP: 990-998 | Volume-9 Issue-2, July 2020. | Retrieval Number: B4011079220/2020©BEIESP | DOI: 10.35940/ijrte.B4011.079220
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Due to a remarkable increase in the complexity of the multimedia content, there is a cumulative enhancement of digital images both online and offline. For the purpose of retrieving images from a vast storehouse of images, there is an urgent requirement of an effectual image retrieval system and the most effective system in this domain is denoted as content-based image retrieval (CBIR) system. CBIR system is generally based on the extraction of basic image attributes like texture, color, shape, spatial information, etc. from an image. But, there exists a semantic gap between the basic image features and high-level human perception and to reduce this gap various techniques can be used. This paper presents a detailed study about the various basic techniques with an emphasis on different intelligent techniques like, the usage of machine learning, deep learning, relevance feedback, etc., which can be used to achieve a high level semantic information in CBIR systems. In addition, a detailed outline regarding the framework of a basic CBIR system, various benchmark datasets, similarity measures, evaluation metrics have been also discussed. Finally, solution to some research issues and future trends have also been given in this paper.
Keywords: Deep Learning, Feature extraction, Image retrieval, Relevance Feedback, Similarity Matching.