| A learning-based method to detect and segment text from scene images |
| |
| 作者姓名： | JIANG Ren-jie QI Fei-hu XU Li WU Guo-rong ZHU Kai-hua |
| |
| 作者单位： | Department of Computer Science and Technology，Shanghai Jiao Tong University，Shanghai 200240，China |
| |
| 基金项目： | Project supported by the OMRON and SJTU Collaborative Founda-tion under PVS project (2005.03～2005.10) |
| |
| 摘 要： | This paper proposes a learning-based method for text detection and text segmentation in natural scene images. First, the input image is decomposed into multiple connected-components (CCs) by Niblack clustering algorithm. Then all the CCs including text CCs and non-text CCs are verified on their text features by a 2-stage classification module, where most non-text CCs are discarded by an attentional cascade classifier and remaining CCs are further verified by an SVM. All the accepted CCs are output to result in text only binary image. Experiments with many images in different scenes showed satisfactory performance of our proposed method.
|
| 关 键 词： | 景象图 文字特征 文字检测 文字分割 学习算法 |
| 收稿时间： | 2006-06-01 |
| 修稿时间： | 2006-10-10 |
A learning-based method to detect and segment text from scene images |
| |
| JIANG Ren-jie QI Fei-hu XU Li WU Guo-rong ZHU Kai-hua.A learning-based method to detect and segment text from scene images[J].Journal of Zhejiang University Science,2007,8(4):568-574. |
| |
| Authors: | Jiang Ren-jie Qi Fei-hu Xu Li Wu Guo-rong Zhu Kai-hua |
| |
| Institution: | (1) Department of Computer Science and Technology, Shanghai Jiao Tong University, Shanghai, 200240, China |
| |
| Abstract: | This paper proposes a learning-based method for text detection and text segmentation in natural scene images. First, the input image is decomposed into multiple connected-components (CCs) by Niblack clustering algorithm. Then all the CCs including text CCs and non-text CCs are verified on their text features by a 2-stage classification module, where most non-text CCs are discarded by an attentional cascade classifier and remaining CCs are further verified by an SVM. All the accepted CCs are output to result in text only binary image. Experiments with many images in different scenes showed satisfactory performance of our proposed method. |
| |
| Keywords: | Text detection Text segmentation Text feature Attentional cascade |
| 本文献已被 CNKI 维普 SpringerLink 等数据库收录！ |
