Mobile imaging helps realize "any time, anywhere" visual communication by allowing consumers to capture, review, share, and print pictures via mobile devices while "on the go". However, major technical challenges exist in bandwidth, power consumption, display and other areas. As a result, the limited resources available in mobile imaging need to be utilized in an intelligent and effective fashion. To this end, we propose to use emphasis image selection (EIS), which automatically selects the most important image, i.e., the one that should receive the most attention or special treatment, given a set of photographic images that typically belong to the same event. Images from a film photofinishing order or a digital camera memory card are first clustered into groups related to different events based on metadata and image content. Emphasis images are selected to provide a summary of the group content and are sent to mobile devices for preview, sharing, browsing, and pri...
Jiebo Luo, Amit Singhal, Andreas E. Savakis