Modelling Human Body Motion using Lagrange Interpolation

Halici E., BOSTANCI G. E.

24th Signal Processing and Communication Application Conference (SIU), Zonguldak, Türkiye, 16 - 19 Mayıs 2016, ss.493-496 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu.2016.7495785
  • Basıldığı Şehir: Zonguldak
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.493-496
  • Anahtar Kelimeler: Lagrange interpolation, human motion modelling, dynamic movement primitives
  • Ankara Üniversitesi Adresli: Evet

Özet

This study presents research on modelling the movement of human limbs using dynamic movement primitives and recovering the motion from these models. The displacements data for the body joints were used to create minimum error Lagrange polynomials for the time input parameter. Considering the computational load when the complete dataset was used, interval and group parameters were required. These parameters were chosen as the values resulting in minimum mean square and mean absolute errors. Results show that positional information for a given time can be efficiently retrieved for a specific joint by choosing the polynomial group belonging to the joint.