AUTOMATIC ANNOTATION OF LOCATION INFORMATION FOR WWW IMAGES (FriPmOR2)

Author(s) : Zhigang Hua (Chinese Academy of Sciences, China) Chuang Wang (Huazhong Univ. of Science and Technology, China) Xing Xie (Microsoft Research Asia, China) Hanqing Lu (Chinese Academy of Sciences, China) Wei-Ying Ma (Microsoft Research Asia, China)

Abstract : Currently, a crucial challenge is raised on how to manage a large amount of images on the Web. Due to a real synergy between location information and images, we propose an automatic solution to annotate contextual location information for WWW images. We construct an image importance model to acquire a set of images that comprise contextual texts in a web page. For each image in this set, we develop an effective algorithm to extract location from its contextual text. We apply our approach to the home pages of about 1,000 websites for image location annotation. The results demonstrate that more than 30% WWW images are related with location, and our solution can achieve a precision of about 60% for location extraction. Finally, we present some potential applications with location information.

Menu