Design and implementation of web product search engine (part 1)

Abstract

The convenience of online shopping on the Internet encourages online businesses to expand their range of products and to deliver customer-oriented services to serve online shoppers better. This project will study on product search using images and explores the different types of Image Retrieval Engine which will complement with existing Web Product Search Engine.

The typical waterfall with feedback model was adopted to provide a systematic approach in fulfilling the requirements for this project. This model also facilitates maintenance and future expansion for this project.

An experiment to compare between Local Image Feature Detection engine and Global Image Feature Detection engine shows the differences in search result accuracy, search efficiency and performance. A detailed discussion and analysis was conducted and reported in this project.

The Image Retrieval Engine will provide convenience for a better and accurate product search which served as a value added services for online shoppers. The study on such engines also explores the possibility for a better search performance through distributed and paralleled search architecture.