An image-to-class dynamic time warping approach for both 3D static and trajectory hand gesture recognition

作者：

Highlights：

• A technique to compute the Image-to-Class DTW distance is proposed.

• A fingerlets’ model for static gesture representation is proposed.

• A strokelets’ model for trajectory gesture representation is proposed.

• A 3D hand trajectory gesture dataset (UESTC-HTG) is built by Kinect.

摘要

Highlights•A technique to compute the Image-to-Class DTW distance is proposed.•A fingerlets’ model for static gesture representation is proposed.•A strokelets’ model for trajectory gesture representation is proposed.•A 3D hand trajectory gesture dataset (UESTC-HTG) is built by Kinect.

论文关键词：Image-to-class distance,Fingerlets,Strokelets,Dynamic time warping,3D hand gesture recognition,Human computer interaction

论文评审过程：Received 21 July 2015, Revised 7 January 2016, Accepted 13 January 2016, Available online 3 February 2016, Version of Record 21 March 2016.

论文官网地址：https://doi.org/10.1016/j.patcog.2016.01.011