基于随机投影的镜头边界检测算法研究

The Shot Boundary Detection Algorithm Based on Random Projection

现有镜头边界检测算法普遍存在检测精度和检测时间难以平衡的问题。视频帧提取特征越精细,算法准确度越高,同时算法复杂度和检测时间也将增加。受到压缩感知思想的启发,文中提出一种镜头边界检测的改进算法,利用随机投影对提取特征进行随机加权求和,以减小特征向量的维数。实验结果表明,所提出的算法在保持检测精度的条件下,降低了算法的复杂度、节约了检测时间。

The existing shot boundary detection algorithm exist the problem that the detection accuracy and the detection time are difficult to balance. The finer the extracted features of the video frames,the finer the accuracy will be while at the cost of increased computational complexity and detection time. Inspired by the compressed sensing theory,an improved shot boundary detection algorithm is proposed in this paper in which the random weighted sums of the extracted features were calculated through the random projection. Experimental results show that the proposed method can preserve the detection accuracy well and reduces the complexity and detection time at the same time.