| 基于提升树的自然场景中文文本定位算法研究 |
| |
| 引用本文： | 徐琼,干宗良,刘峰,陈昌红.基于提升树的自然场景中文文本定位算法研究[J].南京邮电大学学报(自然科学版),2013(6):76-82. |
| |
| 作者姓名： | 徐琼 干宗良 刘峰 陈昌红 |
| |
| 作者单位： | [1]南京邮电大学图像处理与图像通信江苏省重点实验室,江苏南京210003 [2]南京邮电大学宽带无线通信与传感网技术教育部重点实验室,江苏南京210003 |
| |
| 基金项目： | 国家自然科学基金（61071166,60802021,61172118）、江苏省普通高校研究生科研创新计划（CXLX12_0476）、江苏省省属高校自然科学研究（11KJB510012）、江苏省自然科学基金（BK2010523）和广东省科技项目（2011A011305005）资助项目 |
| |
| 摘 要： | 提出了一种新的基于提升树算法的自然场景中文文本定位技术.首先利用边缘特征进行文本区域的检测,即对下采样后彩色图像首先进行边缘提取、二值化处理,然后通过形态学运算以及连通区域分析去除大量的非字符连通域,获得候选的文本区域,最后,提取候选文本区域的PHOG-Gabor特征,通过提升树算法进一步确认是否为字符连通域.通过实验验证,该算法具有很高的召回率和准确率,综合性能较高.
|
| 关 键 词： | 文本定位 自然场景图像 提升树算法 PHOG-Gabor特征 |
Chinese Text Localization Method Based on Boosting Tree in Natural Images |
| |
| XU Qiong,GAN Zong-liang,LIU Feng,CHEN Chang-hong.Chinese Text Localization Method Based on Boosting Tree in Natural Images[J].Journal of Nanjing University of Posts and Telecommunications,2013(6):76-82. |
| |
| Authors: | XU Qiong GAN Zong-liang LIU Feng CHEN Chang-hong |
| |
| Institution: | 1.Jiangsu Key Laboratory of Image Processing ＆ Image Communication, Nanjing University of Posts and Telecommunications,Nanjing 210003, China; 2.Jiangsu Key Laboratory of Image Processing ＆ Image Communication, Nanjing University of Posts and Telecommunications,Nanjing 210003, China;Key Lab of Broadband Wireless Communication and Sensor Network Technology, Ministry of Education,Nanjing University of Posts and Telecommunications, Nanjing 210003, China;) |
| |
| Abstract: | A novel Chinese text localization method based on boosting tree in natural images is proposed.The edge feature is used to detect text regions,where color image is processed by down-sampling,edge extraction and binarization; then the morphological operation and connected component analysis are used to remove non-text regions and get candidate text regions.Finally,PHOG-Gabor features of candidate text regions are extracted,and boosting tree is employed to determine text regions.Experiments show that the proposed Chinese text localization method has high recall rate,precision rate and combination property. |
| |
| Keywords: | text localization natural images boosting tree PHOG-Gabor features |
| 本文献已被 维普 等数据库收录！ |
|
