中文摘要：

视频镜头类型能够弥补低级特征和高级语义之间的鸿沟,有利于精彩事件的检测.文中在现有镜头分类算法的基础上,提出了一种将足球视频中的近镜头进一步分成特写镜头和场外观众镜头的分类方法.该方法选取近镜头关键帧图像在RGB颜色空间中颜色分量最少的单色图像,并对其进行尺寸归一化,然后采用各向同性Sobel算子和闭运算处理,获得二值边缘图像,最后通过计算二值边缘图像中黑色分块的个数实现两类镜头的分类判别.实验结果表明,文中方法是有效且实用的.

英文摘要：

　Video shot type can bridge the gap between the lowlevel features and the high-level semantics and helps to detect highlights.In this paper,a new method for the classification of the close short shot in soccer video is proposed,which classifies the shot into close-up shot or out-of-field audiences＇ shot.In the proposed method,a component image with the minimal summation in all pixel values is selected from the key frame of a close short shot in RGB color space,and the size of the image is then normalized.After that,the isotropic Sobel operator and the closed operation are adopted to obtain the corresponding binarization edge image,and the classification decision is made by calculating the amount of black blocks in the binarization edge image.Experimental results finally reveal the validity and practicality of the proposed method.