Text detection in scene images based on exhaustive segmentation

作者：

Highlights：

• A parallel structure to generate character candidate regions with the exhaustive segmentation of scene images is presented.

• A well-designed two-layer filtering method is used to filter out non-character candidate regions.

• We argue that the text line grouping problem can be posed as the edge cut of the fully connected graph.

• The proposed method is robust to the blurred, low-resolution and small-sized texts.

摘要

Highlights•A parallel structure to generate character candidate regions with the exhaustive segmentation of scene images is presented.•A well-designed two-layer filtering method is used to filter out non-character candidate regions.•We argue that the text line grouping problem can be posed as the edge cut of the fully connected graph.•The proposed method is robust to the blurred, low-resolution and small-sized texts.

论文关键词：Segmentation,Text detection,Scene image,Support vector machine

论文评审过程：Received 17 June 2016, Revised 22 September 2016, Accepted 18 October 2016, Available online 19 October 2016, Version of Record 5 November 2016.

论文官网地址：https://doi.org/10.1016/j.image.2016.10.003