2015International Conference on Computing for Sustainable Global DevelopmentRequires access

Study on efficacy of relevance feedback for Content Based Image Retrieval

Jasvinder A. Singh, Navin Rajpal

Open publisher page 1 citations

Abstract

This paper focuses on a very powerful and efficient technique using which Content Based Image Retrieval (CBIR) is being performed. If one drills down to the origin of CBIR, one would come across low level features like shape, color and texture as parameters which have been used to represent images. These low level features allow one to locate images which are generally visually similar. Although, the image retrieval is performed but visually similar images are not mapped to nearby locations. Hence the need arose to create a powerful technique for image retrieval. Thus came into existence relevance feedback. We have performed an exhaustive study on the efficacy of relevance feedback for CBIR from its commencement till present day and have tried to highlight the advancements done so far in this promising technique.

About this research paper

What this paper is about

This paper focuses on a very powerful and efficient technique using which Content Based Image Retrieval (CBIR) is being performed. If one drills down to the origin of CBIR, one would come across low level features like shape, color and texture as parameters which have been used to represent images. These low level features allow one to locate images which are generally visually similar. Although, the image retrieval is performed but visually similar images are not mapped to nearby locations. Hence the need arose to create a powerful technique for image retrieval. Thus came into existence relevance feedback. We have performed an exhaustive study on the efficacy of relevance feedback for CBIR from its commencement till present day and have tried to highlight the advancements done so far in this promising technique.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

This paper focuses on a very powerful and efficient technique using which Content Based Image Retrieval (CBIR) is being performed. If one drills down to the origin of CBIR, one would come across low level features like shape, color and texture as parameters which have been used to represent images. These low level features allow one to locate images which are generally visually similar. Although, the image retrieval is performed but visually similar images are not mapped to nearby locations. Hence the need arose to create a powerful technique for image retrieval. Thus came into existence relevance feedback. We have performed an exhaustive study on the efficacy of relevance feedback for CBIR from its commencement till present day and have tried to highlight the advancements done so far in this promising technique.

Key concepts: Relevance feedback, Image retrieval, Content-based image retrieval, Computer science, Relevance (law), Information retrieval, Visual Word, Image (mathematics)

Related papers

Back to paper searchBrowse research topicsOriginal source
Study on efficacy of relevance feedback for Content Based Image Retrieval — Research Paper | ScholarLens