Clustering Binary Signature Applied in Content-Based Image Retrieval

Abstract

Image retrieval is an important problem on the multimedia systems. It is time-consuming to query directly on a large image database. So, the paper approaches to build an image retrieval system CBIR (Content-Based Image Retrieval) based on binary signature to retrieve effectively on the large data of images. Firstly, the paper presents the segmentation method based on low-level visual features including color and texture of image. On the basis of segmented image, the paper creates binary signature to describe location, color and shape of interest objects. In order to match similar images, the paper presents a similarity measure between the images based on binary signatures. On the basis of the similarity measure, the paper proposes the clustering binary signature method to quickly query similar images. At the same time it describes the splitting and group method to apply for clusters of binary signatures. From there, the paper gives the CBIR model to describe the process of similar image retrieval. To demonstrate the proposed method, the paper builds application and assesses experimental results on image databases including Corel, Corel Wang and ImageCLEF.