摘要
新闻视频的标题字幕中包含有丰富的语义信息,是实现自动化视频检索、分析和理解的重要信息源之一。通过对新闻字幕的特点进行分析,文章提出了一种基于Prewitt算子的新闻字幕检测的方法。算法首先对图像进行灰度变换,然后利用Prewitt算子进行边缘检测,最后进行字幕区域的探测与合并,检测出字幕。通过对不同的新闻视频帧进行实验,获得了较高的探测准确率。试验证明,文章提出的方法能够较好的完成新闻字幕检测的任务。
Caption in news video is one of the important sources for the realization of automatic retrieval, analysis and comprehension of videos since it implies lots of semantics inherently. On the basis of analyzing the characteristics of news captions, this paper proposes a news caption detection method based on Prewitt operator. The algorithm first makes gray-scale transformation of images and carries out edge detection based on Prewitt operator, finally, captions are extracted through regional exploration and captions merger. On test with different news video frames in experiment the accuracy rate 96.8% is obtained. The experimental results show that the approach proposed can well accomplish the task of detection news caption.
出处
《电子技术（上海）》
2009年第11期24-24,23,共2页
Electronic Technology
基金
国家自然科学基金项目No.60434010和No.60874065
