自适应无监督聚类算法的运动图像关键帧跟踪

Key Frame Tracking of Moving Image Based on Adaptive Unsupervised Clustering Algorithm

针对运动图像关键帧跟踪错误大,速度慢等缺陷,为了获得更高精度的运动图像关键帧跟踪结果,设计了自适应无监督聚类算法的运动图像关键帧跟踪方法。首先对运动图像关键帧跟踪流程进行分析,找到影响运动图像关键帧跟踪效果的因素,然后采集运动图像序列,对其进行分帧处理,提取运动图像关键帧特征,采用聚类算法对特征运动图像关键帧进行处理,最后引入自适应无监督聚类算法进行运动图像关键帧跟踪,并进行运动图像关键帧跟踪仿真实验,实验结果表明,文中方法获得了理想的运动图像关键帧跟踪效果,不但运动图像关键帧跟踪误差要小于其它方法,而且运动图像关键帧跟踪速度更快,具有十分显著的优越性。

Aiming at the defects of large error and slow speed in key frame tracking of moving image,in order to obtain more accurate key frame tracking results of moving image,an adaptive unsupervised clustering algorithm for key frame tracking of moving image is designed.Firstly,the key frame tracking process of moving image is analyzed to find out the factors of the key frame tracking effect of moving image,then the moving image sequence is collected and processed by frame,the key frame feature of moving image is extracted,and the key frame of moving image is processed by clustering algorithm according to the feature moving image key frame,finally,the adaptive unsupervised clustering algorithm is introduced to track the key frame of moving image.The experimental results show that the proposed method can achieve an ideal key frame tracking effect.The key frame tracking error of the moving image is smaller than that of other methods,and the key frame tracking speed of the moving image is faster,which has significant advantages.