Character Region Awareness for Text Detection

TL;DR

This paper introduces a novel scene text detection method that leverages character-level information and affinities to accurately detect arbitrarily shaped and curved texts in natural images, outperforming existing methods.

Contribution

It proposes a new framework that estimates character and affinity regions using both synthetic and real image annotations, improving detection of complex text shapes.

Findings

Significantly outperforms state-of-the-art detectors on six benchmarks.

Effectively detects arbitrarily-oriented and curved texts.

Demonstrates high flexibility in complex scene text detection.

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.