摘要
信息社会从互联网时代发展到移动互联网时代，视觉数据尤其是图像，一直以惊人的速度增长，这是因为图像已成为用户表达、事件记录分析中不可或缺的部分。针对这些包含丰富视觉信息的海量图像，如何在浩瀚的图像库中快速、准确地搜索到与用户指定图像的相似图像，成为信息检索领域研究的热点。
另一方面，Web（网页）应用程序是一种可以通过网页访问的应用程序，程序的最大好处是用户很容易访问应用程序，用户只需要有浏览器即可，不需要再安装其他软件。
本课题的目标是使用Web技术开发一款相似图像搜索的演示系统。具体内容为：以已有的相似图像搜索算法SDK（软件开发工具包）为基础，开发一款友好的人机交互界面，通过调用算法SDK，在指定的数据集中搜索相似图像。
关键词 Web 图像搜索 深度学习 计算机视觉 人机交互
Abstract
With the development of information society from the Internet era to the mobile Internet era, visual data, especially images, has been growing at an amazing speed, because images have become an indispensable part of user expression and event record analysis. In view of these massive images which contain rich visual information, how to search the similar image with the user specified image quickly and accurately in the vast image database has become a research hotspot in the field of information retrieval.
On the other hand, web application is a kind of application that can be accessed through web pages. The biggest advantage of the program is that users can easily access the application. Users only need to have a browser, and do not need to install other software.
The goal of this project is to develop a similar image search demonstration system using web technology. The specific content is: Based on the existing similar image search algorithm SDK (software development kit), a friendly human-computer interface is developed. By calling the algorithm SDK, similar images are searched in the specified data set.
Keywords Web Image search Deep learning Computer vision Human computer interaction
