Content Based Sub-Image Retrieval with Relevance Feedback

In this era of computing data’s are represented as images either as compressed documents or as video images itself. Since data is transmitted as images, they have to be processed for retrieval. Image retrieval is since a challenging issue in the area of computer research as systems that are currently evolving around for this purpose are not able to give an accurate level of retrieval of images as per expectations. Hence there is a high demand for such a system that can evaluate the images and fetches images of much accuracy as possible and at the earliest possible time. The aim of the paper is to improve efficiency of image retrieval, a new image retrieval scheme that applies sub-image processing with low level features of image such as color and shape embedded with segmentation and relevance feedback. It also applies local feature descriptor attributes that are computed on regions of the image. So, a combination of hybrid features and techniques are used to form a retrieval system.