A Framework for Extensive Content-Based Image Retrieval System Incorporating Relevance Feedback and Query Suggestion

Abstract

Over the last decade, there has been a huge increase in the number of images, in which visual information has become increasingly important. If images are to be utilized, they must be organized into databases, from which they can be searched based on different criteria. Content-Based Image Retrieval (CBIR) frameworks have become a mainstream subject of research due to their ability to retrieve images based on actual visual content rather than manually assigned textual descriptions. In the CBIR framework, the analysis and interpretation of image information in large and diverse image databases are evidently complex because there is no prior information about the size or scale of individual structures within the images to be analyzed. In CBIR, retrieval is based on visual image features, which can be extracted automatically from images with the help of human intervention through a technique known as relevance feedback. Nonetheless, efficient methods for differentiating the visual content of images are complex to develop. Therefore, rather than providing a perfect solution, CBIR systems must be able to exploit partial solutions to the problem of image understanding. In this paper, an implementation of a CBIR framework is introduced that not only attempts to efficiently capture user intent based on feedback but also provides query suggestions that help users formulate better queries to retrieve desired results efficiently. The proposed technique is straightforward to implement and scales efficiently to large datasets. Extensive experiments on diverse real-world datasets using image similarity measures have revealed the superiority of the proposed method over existing algorithms.