Beyond OCR: Multi-faceted understanding of handwritten document characteristics

作者:

Highlights:

• Joint feature distribution principle is proposed to design new powerful features.

• Seventeen features have been evaluated for document understanding beyond OCR.

• Co-occurrence features provide promising results on writer and script identification, historical document dating and localization.

摘要

Highlights•Joint feature distribution principle is proposed to design new powerful features.•Seventeen features have been evaluated for document understanding beyond OCR.•Co-occurrence features provide promising results on writer and script identification, historical document dating and localization.

论文关键词:Handwritten document understanding,Joint feature distribution principle,Writer identification,Script identification,Historical manuscript dating and localization

论文评审过程:Received 10 May 2016, Revised 19 September 2016, Accepted 20 September 2016, Available online 23 September 2016, Version of Record 27 October 2016.

论文官网地址:https://doi.org/10.1016/j.patcog.2016.09.017