Block wise local binary count for off-Line text-independent writer identification

作者：

Highlights：

• An effective system for offline text independent writer identification is proposed.

• A reliable writing style representation model for feature extraction is proposed.

• Proposed system evaluated on four challenging databases (two English and two Arabic).

• High identification rates on AHTID\MW, IFN/ENIT and CVL databases.

• Identification rates are recorded over random subdivisions.

摘要

•An effective system for offline text independent writer identification is proposed.•A reliable writing style representation model for feature extraction is proposed.•Proposed system evaluated on four challenging databases (two English and two Arabic).•High identification rates on AHTID\MW, IFN/ENIT and CVL databases.•Identification rates are recorded over random subdivisions.

论文关键词：Off-line writer identification,Handwritten connected components,Text independent,Handwritten documents,Feature extraction,1-NN classifier,Hamming distance

论文评审过程：Received 24 June 2017, Revised 11 September 2017, Accepted 3 October 2017, Available online 4 October 2017, Version of Record 12 October 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.10.010