基于改进K-means算法的关键帧提取

为克服传统聚类算法在关键帧提取过程中对初始参数较为敏感的问题，提出一种基于改进K-means算法的关键帧提取算法。在人工鱼群算法中，依据人工鱼群体相似度对提取的特征向量进行自组织聚类，采用进步最大原则使人工鱼聚集在几个极值点位置，以每个极值点群体相似度最高的人工鱼为初始聚类中心，执行K-means算法，得到聚类结果，并提取关键帧。实验结果表明，该算法的准确率较高，能较好地表达视频的主要内容。

Key Frame Extraction Based on Improved K-means Algorithm

In order to overcome the problems that the traditional clustering algorithm is sensitive to the initial parameter in the key frame extraction process, an efficient algorithm for key frame extraction based on improved K-means algorithm is proposed in this paper. In the implementation of the Artificial Fish Swarm Algorithm(AFSA) clustering algorithm, the artificial fish implement self-organization clustering under the guidance of group similarity and ultimately the artificial fish gathered in several extreme points, according to the greatest progress principle. The artificial fish with the biggest group similarity in each extreme point is set as the initialized cluster center. This paper implements K-means algorithm to obtain the final clustering result and extracts key frame. Experimental result shows that the accuracy of this algorithm is high, and can well express the main content of the video.