DOI QR코드

DOI QR Code

Deep learning based symbol recognition for the visually impaired

시각장애인을 위한 딥러닝기반 심볼인식

  • Park, Sangheon (Electronics and Telecommunications Research Institute) ;
  • Jeon, Taejae (School of Electrical and Electronic Engineering, Yonsei University) ;
  • Kim, Sanghyuk (School of Electrical and Electronic Engineering, Yonsei University) ;
  • Lee, Sangyoun (School of Electrical and Electronic Engineering, Yonsei University) ;
  • Kim, Juwan (Electronics and Telecommunications Research Institute)
  • Received : 2016.04.04
  • Accepted : 2016.05.10
  • Published : 2016.06.30

Abstract

Recently, a number of techniques to ensure the free walking for the visually impaired and transportation vulnerable have been studied. As a device for free walking, there are such as a smart cane and smart glasses to use the computer vision, ultrasonic sensor, acceleration sensor technology. In a typical technique, such as techniques for finds object and detect obstacles and walking area and recognizes the symbol information for notice environment information. In this paper, we studied recognization algorithm of the selected symbols that are required to visually impaired, with the deep learning algorithm. As a results, Use CNN(Convolutional Nueral Network) technique used in the field of deep-learning image processing, and analyzed by comparing through experimentation with various deep learning architectures.

최근 시각장애인 및 교통약자의 자유로운 보행을 보장하기 위한 많은 기술들이 연구되고 있다. 자유로운 보행을 위한 장치로는 영상카메라, 초음파센서 및 가속도 센서 등을 이용하는 스마트 지팡이와 스마트 안경 관련 기술이 있다. 대표적인 기술로는 물체를 식별하여 장애물을 검출하고 보행 가능 영역을 추출하는 기술, 랜드마크 심볼 정보를 인식하여 주위 환경 정보를 주는 기술 등 여러 가지 기술이 개발되고 있다. 본 논문에서는 시각장애인에게 필요한 시설의 대표 심볼을 선정하여 착용한 영상 장치의 정보로부터 심볼을 인식하는 알고리즘을 딥러닝 기술을 이용하여 연구하였다. 그 결과로 딥러닝 영상처리 분야에서 사용되는 CNN(Convolutional Neural Network)기법을 사용하여 서로 다른 딥러닝 구조를 실험을 통하여 비교하고 분석하였다.

Keywords

References

  1. J. Choi, G. Jeong, "Development of Walking Assist Smartphone Case for Blind People", The Journal of Korea Institute of Information, Electronics, and Communication Technology, Vol.8, No.3, pp.239-242, 2015. https://doi.org/10.17661/jkiiect.2015.8.3.239
  2. C. Kang, H. Jo, B. Kim, "A Machine-to-machine based Intelligent Walking Assistance System for Visually Impaired Person", The Journal of The Korean Institute of Communication Sciences, Vol.36, No.3, pp.287-296, 2011.
  3. P. Sermanet, K. Kavukcuoglu, S. Chintala and Y. LeCun, "Pedestrian detection with unsupervised multi-stage feature learning", Proc. IEEE Conference on Computer Vision Pattern Recognition (CVPR), pp.3626-3633, 2013.
  4. D. Dajun and C. Lee, "Fast algorithm for Traffic Sign Recognition", Journal of IKEEE, Vol.16, No.4, pp.356-363, December 2012. https://doi.org/10.7471/ikeee.2012.16.4.356
  5. W. W. Zhu, et al, "Searching for Pulsars Using Image Pattern Recognition", The Astrophysical Journal, Vol.781, No.2, pp.117-128, 2014. https://doi.org/10.1088/0004-637X/781/2/117
  6. A. Krizhevsky, I. Sutskever, and G. E. Hinton. Imagenet classification with deep convolutional neural networks. In Advances in Neural Information Processing Systems 25, pages 1106-1114, 2012.
  7. N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever and R. Salakhutdinov, "Dropout: A simple way to prevent neural networks from overfitting", The Journal of Machine Learning Research, Vol.15, Issue.1, pp.1929-1958, 2014.
  8. Y. Jia, et al. "Caffe: Convolutional architecture for fast feature embedding", Proceedings of the 22nd ACM International Conference on Multimedia, pp. 675-678, 2014.

Cited by

  1. Improving Indentification Performance by Integrating Evidence From Evidence vol.9, pp.6, 2016, https://doi.org/10.17661/jkiiect.2016.9.6.546
  2. Fire Detection System Using Arduino Sensor vol.9, pp.6, 2016, https://doi.org/10.17661/jkiiect.2016.9.6.624
  3. A Fast Algorithm for Region-Oriented Texture Coding vol.9, pp.6, 2016, https://doi.org/10.17661/jkiiect.2016.9.6.519
  4. 무인기 탑재 다중 센서 기반 국지 산불 감시 및 상황 대응 플랫폼 설계 및 구현 vol.10, pp.6, 2016, https://doi.org/10.17661/jkiiect.2017.10.6.626
  5. 합성곱 신경망을 이용한 딥러닝 기반의 프레임 동기 기법 vol.24, pp.4, 2016, https://doi.org/10.6109/jkiice.2020.24.4.501