Abstract: Image retrieval is a specialized data search used to find images. To search for images, a user may provide query terms such as keyword, image file/link, or click on some image and the system will return images "similar" to the query. The similarity used for search criteria could be meta tags, color distribution in images, region/shape attributes, etc. Recently, region based image retrieval has been improved which measures the similarity of the images without proper consideration of the spatial layouts of the ROIs and thus fails to reflect the intent of the user. To meet the challenges, we propose Image Retrieval Method using Relative Location for Multiple Region of Interest (ROI), a similarity measurement using the relative layouts of the ROIs. In contrast to the earlier CBIR, the multiple ROI is specifically designed for the user to choose multiple region of interest from the image and retrieve it. Features of Multiple ROI are the images are divided into blocks of certain sizes to measure the similarities with the target image. In addition to CBIR using multiple ROIs, we extract the texture and object shape features.
G. Raghuraman, S. Sabena and L. Sairamesh, 2016. Image Retrieval Using Relative Location of Multiple ROIS. Asian Journal of Information Technology, 15: 772-775.