PROJECT TITLE :

Imagehawk Search Engine: Content Based Image Retrieval System - 2017

ABSTRACT:

This paper describes massive-scale content based image retrieval system, Image Hawk search engine. ImageHawk search engine uses twenty three.4 million pictures in its gallery. Users have two completely different methods to form their search: Product Quantization (PQ) and Transductive Support Vector Machine primarily based Hashing using Binary Hierarchical Trees (TSVMH-BHT). Pictures are 1st represented with 20480-dimensional Fisher vectors and then binary codes are extracted from Fisher vectors by using these 2 methods. 256-bit binary codes are used for PQ and 512-bit binary codes are used for TSVMH-BHT. When a question image is given to the search engine, the system returns the most similar a hundred images in thirty-forty seconds based on the dimensions of the query image. Additionally we tend to additionally describe our new image retrieval dataset created by using ImageCLEF 2013 and report the accuracies of some well-liked image retrieval methods on this dataset.