CASIA OpenIR  > 模式识别国家重点实验室  > 模式分析与学习
Neural Network Based Over-Segmentation for Scene Text Recognition
He X(贺欣); Wu YC(吴一超); Chen K(陈凯); Yin F(殷飞); Liu CL(刘成林); He X(贺欣)
2015
Conference NameAsian Conference on Pattern Recognition
Source PublicationAsian Conference on Pattern Recognition
Conference Date2015-11
Conference Place吉隆坡
Abstract
Over-segmentation is often used in text recognition to generate candidate characters. In this paper, we propose a neural network-based over-segmentation method for cropped scene text recognition. On binarized text line image, a segmentation window slides over each connected component, and a neural network is used to classify whether the window locates a segmentation point or not. We evaluate several feature representations for window classification and combine sliding window-based segmentation with shape-based splitting. Experimentalresults on two benchmark datasets demonstrate the superiority and effectiveness of our method in respect of segmentation point detection and word recognition.
Keyword场景文字识别 过切分
Indexed ByEI
Document Type会议论文
Identifierhttp://ir.ia.ac.cn/handle/173211/11764
Collection模式识别国家重点实验室_模式分析与学习
Corresponding AuthorHe X(贺欣)
Affiliation中国科学院自动化研究所
Recommended Citation
GB/T 7714
He X,Wu YC,Chen K,et al. Neural Network Based Over-Segmentation for Scene Text Recognition[C],2015.
Files in This Item: Download All
File Name/Size DocType Version Access License
Neural Network Based(444KB)会议论文 开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[He X(贺欣)]'s Articles
[Wu YC(吴一超)]'s Articles
[Chen K(陈凯)]'s Articles
Baidu academic
Similar articles in Baidu academic
[He X(贺欣)]'s Articles
[Wu YC(吴一超)]'s Articles
[Chen K(陈凯)]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[He X(贺欣)]'s Articles
[Wu YC(吴一超)]'s Articles
[Chen K(陈凯)]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: Neural Network Based Over-Segmentation for Scene Text Recognition.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.