摘要
为了在视频图像中进行字幕信息的实时提取,提出了一套简捷而有效的方法。首先进行文字事件检测,然后进行边缘检测、阈值计算和边缘尺寸限制,最后依据文字像素密度范围进一步滤去非文字区域的视频字幕,提出的叠加水平和垂直方向边缘的方法,加强了检测到的文字的边缘;对边缘进行尺寸限制过滤掉了不符合文字尺寸的边缘。应用投影法最终确定视频字幕所在区域。最后,利用OCR识别技术对提取出来的文字区域进行识别,完成视频中文字的提取。以上方法的结合保证了提出算法的正确率和鲁棒性。
A simple and effective method is presented for real-time text segmentation and recognition in videos.Firstly,an algorithm is provided which detects text event and gets edges,then makes a size restrict to the edges and eventually wipes off nontext regions according to the textual energy.The overlaying of detected horizontal edges and vertical edges enhances the text edges,the size restrict of edges helps to wipe off the non-text edges.Later, image projection is applied to get text regions.Eventually,the text regions are processed by OCR technology.The combination of these methods guarantees the performance of this algorithm.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第10期180-183,共4页
Computer Engineering and Applications
基金
国家自然科学基金No.60575013
西北工业大学研究生创新实验中心资助项目~~
关键词
光学文字识别
文字事件检测
数字视频
Optical Character Recognition（OCR）
detection of text event
digital video
