A Local Feature Extraction Mechanism for Effective Content Based Image Retrieval
Sumit Kumar, Manya Singhal, R Revthi · 2022 2nd International Conference on Emerging Frontiers in Electrical and Electronic Technologies (ICEFEET) · 2022
A CBIR system works on the primitive visual features extracted from the images of a concerned database. A CBIR system should provide better results and need to have a close resemblance with human perception. So, to keep these points in mind, in this paper, we have proposed a CBIR system that works on local image features to provide better efficiency. In contrast, the HSV color model has been incorporated as it is closest to human perception. Various statistical parameters are evaluated from H and S components for the color features extraction. Now, for texture and shape features BDIP and BVLC are deployed. This work has been validated through various benchmark databases, and overlying results are capable enough to demonstrate its superiority over contemporaries.