Technical evaluation for mashing up crowdsourcing images

Abstract

The popularity of smart mobile devices with camera enabled results in a growing number of photos on the Internet. Those with semantic and geographic data can be used to enrich the data source of geographic information system, whereas, the widespread sources of these photos make them difficult to retrieve and utilize. We use a third-party image search APIs to retrieve the images dynamically and use both keywords and geographic coordinates as queries to implement the mashup service of searching multi-source POIs and web images. We also designed an efficient schema of storing and managing images uploaded from mobile deuce with rich metadata. We developed a prototype framework of searching and managing crowdsourcing images and verified its feasibility.