Smart Multi-Sensor Wristband for Gesture Classification 14 th Nov - - PowerPoint PPT Presentation

smart multi sensor wristband for gesture classification
SMART_READER_LITE
LIVE PREVIEW

Smart Multi-Sensor Wristband for Gesture Classification 14 th Nov - - PowerPoint PPT Presentation

Smart Multi-Sensor Wristband for Gesture Classification 14 th Nov 2019 Yuchi Liu and Rami Ghannam Microelectronics Lab (meLAB) www.melabresearch.com School of Engineering, University of Glasgow G12 8QQ, Scotland, UK 2109504L@student.gla.ac.uk


slide-1
SLIDE 1

Yuchi Liu and Rami Ghannam Microelectronics Lab (meLAB) www.melabresearch.com School of Engineering, University of Glasgow G12 8QQ, Scotland, UK

2109504L@student.gla.ac.uk rami.ghannam@glasgow.ac.uk

Smart Multi-Sensor Wristband for Gesture Classification

14th Nov 2019

slide-2
SLIDE 2

Hand gesture recognition

  • Liang, X., Ghannam, R., & Heidari, H. (2018). Wrist-Worn Gesture Sensing With Wearable Intelligence. IEEE Sensors Journal, 19(3), 1082-1090.
slide-3
SLIDE 3

Power Harvester

Max=5.64mW Max=46.4mW Vout=3V Vbattery =4.2V

slide-4
SLIDE 4

Fabricate Flexible Piezoelectric Sensor

PDMS+KNLN

Mixing Vacuum Mould High voltage High tempe rature Coating conductive Poling

Flexible piezo sensor

Flexible sensor Copper tape Silver epoxy Solder

slide-5
SLIDE 5

Selected features Time domain Max Min Peak to peak RMS MAD Autoregression coefficient Frequency domain Peak frequency Median frequency Mean frequency Spectral moments ……

Algorithm KNN BDT SVM Accuracy with row data 96.49% 92.86% 98.25% Accuracy with features extraction 73.68% 70.18% 85.95% Time(s) 24.53 4.25 485

Hand Gesture Recognition of Piezo Sensor

slide-6
SLIDE 6

Thanks for Your Listening