Robust Text Detection in Natural Scene Images

TL;DR

This paper presents a robust and accurate method for detecting text in natural scene images, utilizing MSERs, clustering, and machine learning techniques to outperform previous state-of-the-art methods.

Contribution

The paper introduces a novel self-training distance metric learning algorithm and an integrated system that significantly improves text detection accuracy in natural scenes.

Findings

F measure over 76% on ICDAR 2011 dataset, surpassing previous 71%.

Outperforms other methods with over 9% increase in F measure on multilingual datasets.

System is available online for demonstration and further testing.

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 ingle-link clustering algorithm, where distance weights and threshold of the clustering algorithm 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 an character classifier; text candidates with high probabilities are then eliminated and finally texts are identified with a text classifier. The proposed system is evaluated on the ICDAR 2011 Robust Reading Competition dataset; the f measure is over 76% and is significantly better than the state-of-the-art performance of 71%. Experimental results on a publicly available multilingual dataset also show that our proposed method can outperform the other competitive method with the f measure increase of over 9 percent. Finally, we have setup an online demo of our proposed scene text detection system at http://kems.ustb.edu.cn/learning/yin/dtext.