DOI QR코드

DOI QR Code

Study on Hand Gestures Recognition Algorithm of Millimeter Wave

밀리미터파의 손동작 인식 알고리즘에 관한 연구

  • Received : 2020.04.16
  • Accepted : 2020.07.03
  • Published : 2020.07.31

Abstract

In this study, an algorithm that recognizes numbers from 0 to 9 was developed using the data obtained after tracking hand movements using the echo signal of a millimeter-wave radar sensor at 77 GHz. The echo signals obtained from the radar sensor by detecting the motion of a hand gesture revealed a cluster of irregular dots due to the difference in scattering cross-sectional area. A valid center point was obtained from them by applying a K-Means algorithm using 3D coordinate values. In addition, the obtained center points were connected to produce a numeric image. The recognition rate was compared by inputting the obtained image and an image similar to human handwriting by applying the smoothing technique to a CNN (Convolutional Neural Network) model trained with MNIST (Modified National Institute of Standards and Technology database). The experiment was conducted in two ways. First, in the recognition experiments using images with and without smoothing, average recognition rates of 77.0% and 81.0% were obtained, respectively. In the experiment of the CNN model with augmentation of learning data, a recognition rate of 97.5% and 99.0% on average was obtained in the recognition experiment using the image with and without smoothing technique, respectively. This study can be applied to various non-contact recognition technologies using radar sensors.

본 논문에서는 77GHz를 사용하는 밀리미터파 레이더 센서의 반향 신호를 이용하여 손동작의 움직임을 추적한 후 얻어진 데이터로 0부터 9까지의 숫자들을 인식하는 알고리즘을 개발하였다. 손동작을 감지하여 레이더 센서로부터 얻어진 반향 신호들은 산란 단면적의 차이 등에 의해 불규칙한 점들의 군집형태를 보인다. 이들로부터 유효한 중심점을 얻기 위해 3차원 좌푯값들을 이용해 K-Means 알고리즘을 적용하였다. 그리고 얻어진 중심점들을 연결하여 숫자 형태의 이미지를 생성하였다. 얻어진 이미지와 스무딩 기법을 적용해 사람의 손글씨 형태와 유사하게 만든 이미지를 MNIST(Modified National Institute of Standards and Technology database)로 훈련된 CNN(Convolutional Neural Network) 모델에 입력하여 인식률을 비교하였다. 실험은 두 가지 방법으로 진행되었다. 먼저 스무딩 기법을 적용한 이미지와 적용하지 않은 이미지를 사용한 인식 실험에서는 각각 평균 77.0%와 81.0%의 인식률을 얻었다. 그리고 학습데이터를 확장(augmentation)한 CNN 모델의 실험에서는 스무딩 기법을 적용한 이미지와 적용하지 않은 이미지를 사용한 인식 실험에서 각각 평균 97.5%와 평균 99.0%의 인식률을 얻었다. 본 연구는 레이더 센서를 이용한 다양한 비접촉 인식기술에 응용이 가능할 것으로 판단된다.

Keywords

References

  1. S. Mitra and T. Acharya, "Gesture Recognition: A Survey", in IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), vol. 37, no. 3, pp. 311-324, May 2007. DOI: http://dx.doi.org/10.1109/TSMCC.2007.893280
  2. H. C. Yoon and J. S. Cho, "Hand Feature Extraction Algorithm Using Curvature Analysis For Recognition of Various Hand Gestures", Journal of The Korea Society of Computer and Information, Vol.20, No.5, pp.13-20, May 2015. DOI : http://dx.doi.org/10.9708/jksci.2015.20.5.013
  3. Y. S. Lee, "Study on the Hand Gesture Recognition System and Algorithm based on Millimeter Wave Radar", Journal of Korea Institute of Information, Electronics, and Communication Technology, Vol.12, No.3, pp.251-256, Dec. 2019. DOI : http://dx.doi.org/10.17661/jkiiect.2019.12.3.251
  4. P. Molchanov, S. Gupta, K. Kim and K. Pulli, "Short-range FMCW monopulse radar for hand-gesture sensing", 2015 IEEE Radar Conference (RadarCon), Arlington, VA, pp. 1491-1496, Jun. 2015. DOI: http://dx.doi.org/10.1109/RADAR.2015.7131232
  5. http://www.ti.com/lit/ds/symlink/iwr1443.pdf (accessed Dec. 01, 2019)
  6. http://www.ti.com/lit/an/swra553a/swra553a.pdf (accessed Dec. 01, 2019)
  7. http://yann.lecun.com/exdb/mnist/ (accessed March 10, 2020)
  8. https://www.tensorflow.org/ (accessed March 10, 2020)
  9. Jurgen Schmidhuber, "Deep learning in neural networks: An overview", Neural Networks, Vol.61, pp.85-117, Jan. 2015. DOI: http://dx.doi.org/10.1016/j.neunet.2014.09.003
  10. M. J. Kang. "Comparison of Gradient Descent for Deep Learning" Journal of the Korea Academia-Industrial cooperation Society, Vol.21, No.2, pp.189-194, Feb. 2020. DOI: http://dx.doi.org/10.5762/KAIS.2020.21.2.189
  11. Fritsch, F. N. and R. E. Carlson. "Monotone Piecewise Cubic Interpolation", SIAM Journal on Numerical Analysis. Vol. 17, pp.238-246, 1980. DOI: http://dx.doi.org/10.1137/0717021
  12. https://github.com/franneck94/MNIST-Data-Augmentation (accessed March 10, 2020)