Image Retrieval through sketches based on Descriptor with Data Retrieval using Reversibility Method

Abstract

The various analysis of Image retrieval makes the Query by sketches as an important branch of image retrieval. It also acts as supplementary to Content Based Image Retrieval. Whereas Sketch based image retrieval plays a vital role in critical applications like military and crime systems, as accuracy becomes the major factor. Here loss can be any data or information hidden in images. The data hiding technique uses reversibility to extract the data, as well as the real images, which can be extracted from sketches. The message embedding and extracting process uses all 3 color channels and a difference histogram. The proposed work of SBIR adopts different processes like pre-processing for extraction of strong edges, feature extraction by descriptor, and constraint's application for the selection of shaping edges. In the proposed work, the framework will significantly improve the SBIDR performance by exploring the method: Image Retrieval with Data Retrieval simultaneously.