Efficient character segmentation approach for machine-typed documents

作者：

Highlights：

• Efficient character segmentation algorithm for machine typed documents is presented.

• The efficient algorithm supports general pipeline from grayscale conversion to segmentation.

• The mathematical background of the novel idea is also presented in detail.

• The algorithm core part is shown in pseudo-code proved with a large set of empiric results.

摘要

•Efficient character segmentation algorithm for machine typed documents is presented.•The efficient algorithm supports general pipeline from grayscale conversion to segmentation.•The mathematical background of the novel idea is also presented in detail.•The algorithm core part is shown in pseudo-code proved with a large set of empiric results.

论文关键词：Character segmentation,Character recognition,Machine-typed documents,Machine-printed documents

论文评审过程：Received 27 November 2016, Revised 11 March 2017, Accepted 12 March 2017, Available online 16 March 2017, Version of Record 23 March 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.03.027