Realtime multi-scale scene text detection with scale-based region proposal network

作者：

Highlights：

• We propose a novel network named SRPN to realize both text/non-text localization and scale estimation efficiently.

• A two-stage detection scheme based on SRPN is proposed to avoid using multi-scale pyramid input and achieve faster detection speed.

• The proposed method achieves remarkable speedup on ICDAR2015, ICDAR2013 and MSRA-TD500 while keeping competitive performance.

• Ablation experiments are given to prove reasonableness of the proposed method from different aspects.

摘要

•We propose a novel network named SRPN to realize both text/non-text localization and scale estimation efficiently.•A two-stage detection scheme based on SRPN is proposed to avoid using multi-scale pyramid input and achieve faster detection speed.•The proposed method achieves remarkable speedup on ICDAR2015, ICDAR2013 and MSRA-TD500 while keeping competitive performance.•Ablation experiments are given to prove reasonableness of the proposed method from different aspects.

论文关键词：Scene text detection,Multi-scale,Speedup,Scale-based region proposal network

论文评审过程：Received 5 September 2018, Revised 27 June 2019, Accepted 26 August 2019, Available online 3 September 2019, Version of Record 9 September 2019.

论文官网地址：https://doi.org/10.1016/j.patcog.2019.107026