Real-time hand posture recognition using hand geometric features and Fisher Vector

作者：

Highlights：

• Angle, distance and curvature features are devised for each hand contour point.

• All descriptors are encoded with a FV to form a compact hand shape representation.

• A technique is proposed to separate hand from forearm obtaining accurate hand mask.

• Extensive experiments demonstrate the effectiveness and efficiency of our method.

摘要

•Angle, distance and curvature features are devised for each hand contour point.•All descriptors are encoded with a FV to form a compact hand shape representation.•A technique is proposed to separate hand from forearm obtaining accurate hand mask.•Extensive experiments demonstrate the effectiveness and efficiency of our method.

论文关键词：Hand posture recognition,Geometric features,Fisher Vector

论文评审过程：Received 21 March 2018, Revised 6 September 2019, Accepted 1 December 2019, Available online 6 December 2019, Version of Record 10 December 2019.

论文官网地址：https://doi.org/10.1016/j.image.2019.115729