Significant scene extraction method using situation importance

Sports videos for mobile terminals and PCs and other multimedia information services have recently spread. The ability to efficiently extract significant scenes is important with these services. We proposed the personal digest making scheme (PDMS), which calculates the significance of events in a sports game based on video meta data and extracts significant scenes to compose a digest. PDMS calculations focus on additional-point event significance, so the system cannot analyze low-score games correctly. To address the problem, we explain a method of calculating situation importance before events happen, which takes into account when extracting significant scenes. Situation importance affords a degree of chance. Using situation importance makes it possible to extract significant scenes more correctly.