一种快速有效的视频镜头边界检测方法

本文在分析现有的镜头边界检测方法的基础上,提出了一种层次化镜头边界检测方法.首先我们利用小波变换对信号奇异点的检测能力和对噪声具有较强的鲁棒性,实现了镜头边界的预检测.然后我们通过对候选镜头边界的虚警分析,有效地降低了闪光灯和快速运动对镜头边界检测产生的影响,并实现了镜头突变检测.在渐变类型分类中,我们分别实现了淡入淡出、扫换和溶解等三种重要镜头边界类型的检测.实验结果表明,本文方法能够快速、有效地实现视频镜头边界类型的检测,对运动和光线变化有较强的鲁棒性.

A Rapid and Effective Method for Shot Boundary Detection

A rapid and effective shot boundary detection algorithm is proposed.First,the video sequence is mapped into a temporal evolution curve(TEC) by using the L_2-norm of HS histogram.Then singularity detection with wavelet is employed to identify the potential shot boundaries in TEC curve.Next then,we utilize intensity of motion intensity and block-based edge direction histogram to distinguish shot boundaries from the false positives caused by rapid motion and complex flashlight scenes.At last,gradual transition is further classified into dissolve,wipe and fade in/fade out.The experimental results compared with other methods achieve satisfying precision and recall of detecting shot boundaries in this algorithm.