$f$ -measure is over 76%, much better than the state-of-the-art performance of 71%. Experiments on multilingual, street view, multi-orientation and even born-digital databases also demonstrate the effectiveness of the proposed method. Finally, an online demo of our proposed scene text detection system has been set up at http://prir.ustb.edu.cn/TexStar/scene-text-detection/." /> $f$ -measure is over 76%, much better than the state-of-the-art performance of 71%. Experiments on multilingual, street view, multi-orientation and even born-digital databases also demonstrate the effectiveness of the proposed method. Finally, an online demo of our proposed scene text detection system has been set up at http://prir.ustb.edu.cn/TexStar/scene-text-detection/." /> $f$ -measure is over 76%, much better than the state-of-the-art performance of 71%. Experiments on multilingual, street view, multi-orientation and even born-digital databases also demonstrate the effectiveness of the proposed method. Finally, an online demo of our proposed scene text detection system has been set up at http://prir.ustb.edu.cn/TexStar/scene-text-detection/." /> Robust Text Detection in Natural Scene Images
The Community for Technology Leaders
RSS Icon
Subscribe
Issue No.05 - May (2014 vol.36)
pp: 970-983
Xu-Cheng Yin , Dept. of Comput. Sci. & Technol., Univ. of Sci. & Technol. Beijing, Beijing, China
Xuwang Yin , Dept. of Comput. Sci. & Technol., Univ. of Sci. & Technol. Beijing, Beijing, China
Kaizhu Huang , Dept. of Electr. & Electron. Eng., Xi'an Jiaotong-Liverpool Univ., Suzhou, China
Hong-Wei Hao , Inst. of Autom., Beijing, China
ABSTRACT
Text detection in natural scene images is an important prerequisite for many content-based image analysis tasks. In this paper, we propose an accurate and robust method for detecting texts in natural scene images. A fast and effective pruning algorithm is designed to extract Maximally Stable Extremal Regions (MSERs) as character candidates using the strategy of minimizing regularized variations. Character candidates are grouped into text candidates by the single-link clustering algorithm, where distance weights and clustering threshold are learned automatically by a novel self-training distance metric learning algorithm. The posterior probabilities of text candidates corresponding to non-text are estimated with a character classifier; text candidates with high non-text probabilities are eliminated and texts are identified with a text classifier. The proposed system is evaluated on the ICDAR 2011 Robust Reading Competition database; the f-measure is over 76%, much better than the state-of-the-art performance of 71%. Experiments on multilingual, street view, multi-orientation and even born-digital databases also demonstrate the effectiveness of the proposed method.
INDEX TERMS
probability, learning (artificial intelligence), object detection, pattern clustering,born-digital databases, robust text detection, natural scene images, content-based image analysis, pruning algorithm, MSER, maximally stable extremal regions, minimizing regularized variations strategy, single-link clustering algorithm, distance weights, clustering threshold, self-training distance metric learning algorithm, posterior probabilities, text classifier, f-measure, multilingual database, street view database, multiorientation database,Clustering algorithms, Algorithm design and analysis, Measurement, Vegetation, Robustness, Databases, Educational institutions,Computing Methodologies, Text processing, Scene Analysis, Image Processing and Computer Vision,distance metric learning, Scene text detection, maximally stable extremal regions, single-link clustering
CITATION
Xu-Cheng Yin, Xuwang Yin, Kaizhu Huang, Hong-Wei Hao, "Robust Text Detection in Natural Scene Images", IEEE Transactions on Pattern Analysis & Machine Intelligence, vol.36, no. 5, pp. 970-983, May 2014, doi:10.1109/TPAMI.2013.182
112 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool