Effective Keyframe Extraction from Personal Video by Using Nearest Neighbor Clustering

作者：Chaohui Lv, Yiyang Huang · 发表于：2018 11th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI) · 年份：2018 · DOI：10.1109/cisp-bmei.2018.8633207 · 被引用次数：7 · 研究领域：Computer Science

Keyframe extraction is one of the most important issue for video understanding and retrieval researches, especially with the rapid growth of uploaded personal video numbers. The frame clustering method has been a theoretically mature method in keyframe extraction. However, the process of clustering often takes long time that seems not very practical. In this paper, we propose an improved clustering method based on video feature, and it proves to be effective in keyframe extraction.