Handwritten word image matching based on Heat Kernel Signature

作者：

Highlights：

• Handwritten word image matching based on Heat Kernel Signature (HKS) is proposed.

• HKS can capture local characteristics of detected keypoints.

• The triangular mesh structure is used to present global characteristics.

• The use of HKS can tolerate large variations in handwritten images.

• Our proposed method does not require pre-processing.

摘要

Highlights•Handwritten word image matching based on Heat Kernel Signature (HKS) is proposed.•HKS can capture local characteristics of detected keypoints.•The triangular mesh structure is used to present global characteristics.•The use of HKS can tolerate large variations in handwritten images.•Our proposed method does not require pre-processing.

论文关键词：Handwritten documents,Word image matching,Heat Kernel Signature,Triangular mesh,Matching score

论文评审过程：Received 1 February 2014, Revised 16 October 2014, Accepted 21 October 2014, Available online 28 October 2014, Version of Record 16 July 2015.

论文官网地址：https://doi.org/10.1016/j.patcog.2014.10.028