A Novel Approach for Shot Boundary Detection and Key Frames Extraction

This paper presents novel algorithms for shot boundary detection and key frames extraction. The algorithm differs from conventional methods mainly in the use of image segmentation and attention model. Matching difference between two consecutive frames is computed with different weights. Shot boundaries are detected with automatic threshold. Key frame is extracted by using reference frame-based approach. Experimental results show improved performance of shot boundary detection by using the proposed algorithms, and key frames represent satisfactorily shot content.