[1]贾冬勤,王洪元,程起才,等.基于自适应阈值Canny算子的视频文本定位方法[J].常州大学学报(自然科学版),2014,(01):47-51.[doi:10.3969/j.issn.2095-0411.2014.01.011]
　JIA Dong-qin,WANG Hong-yuan,CHENG Qi-cai.Video Text Localization Method Based on Adaptive Threshold Canny Operator[J].Journal of Changzhou University(Natural Science Edition),2014,(01):47-51.[doi:10.3969/j.issn.2095-0411.2014.01.011]
点击复制

基于自适应阈值Canny算子的视频文本定位方法()
分享到：

常州大学学报(自然科学版)[ISSN:2095-0411/CN:32-1822/N]

卷:
期数:
2014年01期
页码:
47-51
栏目:
出版日期:
2014-02-28

文章信息/Info

Title:
Video Text Localization Method Based on Adaptive Threshold Canny Operator
作者:
贾冬勤；王洪元；程起才；
常州大学 信息科学与工程学院,江苏 常州 213164
Author(s):
JIA Dong-qin; WANG Hong-yuan; CHENG Qi-cai
School of Information Science and Engineering,Changzhou University,Changzhou 213164,China
关键词:
文本定位;  自适应阈值;  Canny算子
Keywords:
text localization;  adaptive threshold;  Canny algorithm
分类号:
TP319
DOI:
10.3969/j.issn.2095-0411.2014.01.011
文献标志码:
A
摘要:
为实现视频图像中人工文本的快速定位,提出了一种有效的字幕定位方法。该方法首先对视频帧进行灰度变换,去除冗余颜色信息,利用自适应阈值Canny算子对文本区域进行边缘检测,再将得到的子图像在水平方向进行投影行定位,最后对行定位结果进行文字区域精确化。结果表明:该方法能避免人为设置阈值不当使得在不同背景情况下产生伪边缘或边缘漏检,能够有效定位文本,不受文字字体、大小等因素的影响。
Abstract:
The paper proposes an effective method for location of theartificial text in video images.This algorithm transmits the grey level of videoframe first,and wipes off the information of redundancy colors.Then makes use of adaptive threshold Canny algorithm to detect the boundary of text area.Row location will be in progress in horizontal direction of the subimage.And accuratethe text area of the horizontal direction results.The consequence shows that this method can avoid faults,which will lead to unsatisfactory results of Canny boundary detection,in setting threshold by human.This algorithm can locate the text effectively,ignoring the impact of the typeface and size of the text.

参考文献/References:

[1]叶静.基于内容的新闻视频检索[D].上海:上海大学,2001.
[2]Zhang Dong-qing,Chang Shih-fu.Learning to detect scene text using a higher-order MRF with belief propagation[C]∥Computer Vision and Pattern Recognition Workshops.Washington D C:IEEE Computer Society,2004.
[3]LI Chuang,Ding Xiao-qing,Wu You-shou.Automatic text location in natural scene images[C]∥Proc of ICDAR.Seattle:IEEE Computer Society,2001.
[4]蔡波,周洞汝,胡宏斌.数字视频中字幕检测及提取的研究和实现[J].计算机辅助设计与图形学学报,2003, 15(7):898-903.
[5]Wu V,Manmatha R,Riseman E M.Text finder:An automatic system to detect and recognize text in images[J].IEEE Trans on Pattern Analysis and Machine Intelligence,1999,20(11):1224-1229.
[6]Jung K.Neural network-based text location in color images[J].Pattern Recognition Letters,2001,22(14):1503-1515.
[7]Chen Da-tong,Odobez Jean-Marc,Bourlard H.Text detect and recognition inimages and video frames[J].Pattern Recognition,2004(37):595-608.
[8]冈萨雷斯.数字图像处理[M].2版.北京:电子工业出版社,2007:463-491.
[9]Otsu N.A threshold selection method from gray-level histogram[J].IEEE Transactions on system Man Cybernetics(S1083-4419),1979,9(1):62-66.
[10]李牧,闫继红,李戈,等.自适应Canny算子边缘检测技术[J].哈尔滨工程技术学报,2007,28(9):1002-1007.
[11]唐露露,张启灿,胡松.一种自适应阈值的Canny边缘检测算法[J].光电工程,2011.38(05):127-132.

备注/Memo

备注/Memo:
基金项目:国家自然科学基金项目资助(61070121) 作者简介:贾冬勤(1988-),女,江苏南通人,硕士生; 通讯联系人:王洪元。
更新日期/Last Update: 2014-02-20