Image Retrieval System Based on Feature Extra...

Image Retrieval System Based on Feature Extraction

Author Name : Ms M. Pallavi, T. Sandhya, S. Venkata Sowjanya, P. Madhuri, M. Aruna

ABSTRACT

Nowadays the use of multimedia data is massive, because of the difference in the type of electronic devices used. Humans cannot have grip on things like capturing, storing, indexing, retrieving, analysing, and summarizing the data used in these devices. The images have an irreplaceable role for all multimedia applications. It is highly imperative for the user to have a perfect image mining system. Earlier, this image retrieval system was termed as text based image retrieval. Unfortunately, this system has many drawbacks. Textural description is not capable of capturing visual content. The Image content can be expressed in different ways and images are beyond the description of words. In order to overcome these issues, a new retrieval technique based on content for image mining is being widely used and practiced. In this process only two images are taken for comparison. The whole database may be searched to find the closest matching image this process is a genuine example of this content based image searching process. This content based image searching system contains many components such as feature extraction and representation, similarity measurement, databases of pre analysed image collections. The images are retrieving from a large collection of image data base. The retrieving images having some features such as colour, texture and shape. The features can be automatically extracted from the images data base.

Keywords: Deep Learning, Convolutional Neural Network,VGG-16.