Investigation of an Efficient Integrated Semantic Interactive Algorithm for Image Retrieval

In this research, a novel integrated semantic interactive algorithm for image retrieval is proposed to retrieve set of relevant images for a given query. The main challenge in image retrieval is to retrieve relevant images with high precision and to diminish the semantic gap. Ranking relevance feedback, semantic feature template, and unsupervised cluster learning techniques are the three approaches investigated to meet the challenge in image retrieval. In this paper, integrated ranking relevance feedback–semantic feature template and integrated ranking relevance feedback–unsupervised cluster learning is proposed to increase the retrieval rate. The experimental results are explored with Corel-1K image database and are evaluated using precision, recall, error rate and execution time to demonstrate the effectiveness of the proposed system. The simulation results reveal that the integrated ranking relevance feedback–semantic feature template algorithm yields high precision and outperforms the other existing methods with the maximum precision