Abstract

Scene text detection methods based on neural networks have emerged recently and have shown promising results. Previous methods trained with rigid word-level bounding boxes exhibit limitations in representing the text region in an arbitrary shape. In this paper, we propose a new scene text detection method to effectively detect text area by exploring each character and affinity between characters. To overcome the lack of individual character level annotations, our proposed framework exploits both the given character-level annotations for synthetic images and the estimated character-level ground-truths for real images acquired by the learned interim model. In order to estimate affinity between characters, the network is trained with the newly proposed representation for affinity. Extensive experiments on six benchmarks, including the TotalText and CTW-1500 datasets which contain highly curved texts in natural images, demonstrate that our character-level text detection significantly outperforms the state-of-the-art detectors. According to the results, our proposed method guarantees high flexibility in detecting complicated scene text images, such as arbitrarily-oriented, curved, or deformed texts.

Keywords

Character (mathematics)Computer scienceBounding overwatchArtificial intelligenceRepresentation (politics)Pattern recognition (psychology)Word (group theory)Image (mathematics)Flexibility (engineering)Text detectionNatural language processingComputer visionMathematics

Affiliated Institutions

Related Publications

Publication Info

Year
2019
Type
article
Pages
9357-9366
Citations
981
Access
Closed

External Links

Social Impact

Social media, news, blog, policy document mentions

Citation Metrics

981
OpenAlex

Cite This

Youngmin Baek, Bado Lee, Dongyoon Han et al. (2019). Character Region Awareness for Text Detection. , 9357-9366. https://doi.org/10.1109/cvpr.2019.00959

Identifiers

DOI
10.1109/cvpr.2019.00959