A Novel Objects of Interest Extraction Approach Using Attention-Driven Model for Content-Based Image Retrieval

Visual attention plays a vital role for humans to understand a scene by intuitively emphasizing some focused objects, and recent work in the computational model of visual attention has demonstrated that a purely bottom-up approach to identify salient regions within an image can be successfully applied to diverse and practical problems. Being aware of this, we propose a new approach of extracting objects of interest (OOIs) using attention-driven model. In this approach, images are coarsely segmented into regions using EM (Expectation-Maximization) algorithm, and then we use the modified Itti-Koch model (M-Itti-Koch) of visual attention to find salient peaks, if these peaks overlap with regions generated by EM algorithm, we proceed to extract attentive object around those points. Experiment results demonstrate that the proposed approach gives good performance, as compared with the current peer method in the literature.