Representative and Diverse Image Set Gathering for Geographic Area and its Surrounding Region

Geo-tagging is becoming famous in today’s era of web 2.0 technologies. Geo-tagging is a concept in which we are adding information about location to the image. In proposed approach such geo-tagged images are used for making visual summary. Generally, visual summary of geographic location provides images of dominate location and ignore surrounding areas of that location which are potentially valuable. Multiple images of the same location suppress other places close to the dominant location and can be visited. In this proposed approach automatic visual summary of geographic area and its surrounding location is generated. For this community contributed images and their associated metadata are extracted from popular social websites such as Flickr [11]. Proposed approach generate visual summary which contains representative and diverse image set of given geographic location and its surrounding areas. For automatic generation of visual summary four layer graph of images and its associated metadata is formed. Proposed edge weight calculation method in this system is novel. With the help of random walk with restarts representative and diverse image set is selected which is visual summary of the given geographic location. Such visual summary can motivate a person to visit nearby areas of given geographic location.