Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography (한국측량학회지)

Korean Society of Surveying, Geodesy, Photogrammetry and Cartography (한국측량학회)
권호 표지
DOI QR코드

DOI QR Code

Performance Analysis of Vision-based Positioning Assistance Algorithm

비전 기반 측위 보조 알고리즘의 성능 분석

  • Park, Jong Soo (Dept. of Geoinformatics, The University of Seoul) ;
  • Lee, Yong (Dept. of Geoinformatics, The University of Seoul) ;
  • Kwon, Jay Hyoun (Dept. of Geoinformatics, The University of Seoul)
  • Received : 2018.12.17
  • Accepted : 2019.04.29
  • Published : 2019.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

Due to recent improvements in computer processing speed and image processing technology, researches are being actively carried out to combine information from camera with existing GNSS (Global Navigation Satellite System) and dead reckoning. In this study, developed a vision-based positioning assistant algorithm to estimate the distance to the object from stereo images. In addition, GNSS/on-board vehicle sensor/vision based positioning algorithm is developed by combining vision based positioning algorithm with existing positioning algorithm. For the performance analysis, the velocity calculated from the actual driving test was used for the navigation solution correction, simulation tests were performed to analyse the effects of velocity precision. As a result of analysis, it is confirmed that about 4% of position accuracy is improved when vision information is added compared to existing GNSS/on-board based positioning algorithm.

최근 컴퓨터 처리 속도의 향상과 영상 처리 기술의 발달로 인해 카메라에서 획득하는 정보를 기존의 GNSS(Global Navigation Satellite System), 추측 항법 기반의 측위 기술과 결합하여 안정적인 위치를 결정하기 위한 연구가 활발히 진행 중이다. 기존 연구에서는 단안 카메라를 이용한 연구가 주로 수행되었으나 이 경우 관심 객체의 절대좌표가 구축이 되어 있어야 한다는 한계점이 있다. 이러한 한계를 극복하기 위해 본 연구에서는 스테레오 영상으로부터 삼각측량법을 적용하여 카메라와 관심 객체간 거리를 추정하는 비전 기반 측위 보조 알고리즘을 개발하고 성능 분석을 수행하였다. 또한, 추정된 거리와 카메라 영상 획득 간격을 이용해 상대적인 속도를 계산하고 이를 기존에 개발된 GNSS/이동체 내부 센서 기반 측위 알고리즘과 결합하여 통합 측위 알고리즘을 구현하였다. 실제 주행 자료를 기반으로 통합측위 알고리즘에 대한 성능을 분석한 결과 기존에 개발된 GNSS/이동체 내부 센서 기반 측위 알고리즘에 비해 속도 정보를 항법해 보정에 활용하였을 때 약 4%의 미미한 위치 정확도 향상 효과를 확인하였다. 이는 영상으로부터 추정된 속도 정보의 정밀도가 낮고, 터널 등을 지날 때는 영상으로부터 적절한 정보를 추출할 수 없다는 한계가 있어 이를 보완한 추가 연구가 필요하다고 판단된다.

Keywords

GCRHBD_2019_v37n3_101_f0001.png 이미지

Fig. 1. Distance estimation process

GCRHBD_2019_v37n3_101_f0002.png 이미지

Fig. 2. Image rectification

GCRHBD_2019_v37n3_101_f0003.png 이미지

Fig. 3. Stereo vision-based depth estimation (Savarese, 2011)

GCRHBD_2019_v37n3_101_f0004.png 이미지

Fig. 4. Process of several weak classifiers (Viola and Jones, 2004)

GCRHBD_2019_v37n3_101_f0005.png 이미지

Fig. 5. Sample image of H target

GCRHBD_2019_v37n3_101_f0006.png 이미지

Fig. 6. Calibration image position

GCRHBD_2019_v37n3_101_f0007.png 이미지

Fig. 7. Two distance combination (left : Day 1, right : Day 2)

GCRHBD_2019_v37n3_101_f0008.png 이미지

Fig. 8. Distance error with respect to the target distance

GCRHBD_2019_v37n3_101_f0009.png 이미지

Fig. 9. Diagram of GNSS/In-vehicle sensor/Vision based positioning algorithms

GCRHBD_2019_v37n3_101_f0010.png 이미지

Fig. 10. Sample image of road sign

GCRHBD_2019_v37n3_101_f0011.png 이미지

Fig. 11. Image-based velocity update

Table 1. Calibration image combination

GCRHBD_2019_v37n3_101_t0001.png 이미지

Table 2. Distance error with respect to the distance

GCRHBD_2019_v37n3_101_t0002.png 이미지

Table 3. Mean of the distance error by the application of scale

GCRHBD_2019_v37n3_101_t0003.png 이미지

Table 4. Compared statistics of position error

GCRHBD_2019_v37n3_101_t0004.png 이미지

Table 5. Compared statistics of position error in velocity correction

GCRHBD_2019_v37n3_101_t0005.png 이미지

References

  1. Dalal, N. and Triggs, B. (2005), Histograms of oriented gradients for human detection, In Computer Vision and Pattern Recognition, 2005. Proceedings of the CVPR 2005, IEEE, 20-25 June, San Diego, California, Vol. 1, pp. 886-893.
  2. Eric, J. and Randy H. (2007), Computer Vision Class Project, University of Utah, Utah, http://www.eng.utah.edu/-hamburge/ (last date accessed: September 18 2017).
  3. Godha, S. and Cannon, M.E. (2007), GPS/MEMS INS intergrated system for navigation in urban areas, GPS Solutions, No. 11, Vol. 3, pp. 193-203. https://doi.org/10.1007/s10291-006-0050-8
  4. Habib, A., Asmamaw, A., Kelley, D., and May, M. (2000), Linear features in photogrammetry, Ohio State University, Division of Geodetic Science.
  5. Han, J.H. (2016), Performance Analysis of Positioning through the Integration of GNSS and On-board Vehicle Sensors, Ph.D. dissertation, University of Seoul, Seoul, Korea, 162p.
  6. Jung, H.G. and Suhr, J.K. (2015), Sensor Fusion-based Precise Localization System for Automotive Applications, Korean Society Automobile Engineering, Vol. 37, No. 6, pp. 29-34.
  7. Jung, J. (2018), Development of Single Image-based Assistant Algorithm for Vehicle Positioning by Detecting Road Facilities, Master's thesis, University of Seoul, Seoul, Korea, 78p.
  8. Kim, N.H. and Park, C.H. (2014), Horizontal Error Analysis of Stereo Vision Positioning System, HCI KOREA 2014, 10-12 December, Seoul, Korea, pp. 24-26.
  9. Nister, D., Naroditsky, O., and Bergen, J. (2006), Visual odometry for ground vehicle applications, Journal of Field Robotics, Vol. 23, No. 1, pp. 3-20. https://doi.org/10.1002/rob.20103
  10. Savarese, S. (2011), Computer Vision EECS 442 Lecture note, University of Michigan, http://vhosts.eecs.umich.edu/vision//teaching/EECS442_2011/lectures/lecture7.pdf(last date accessed: 19 November 2018).
  11. Theodoridis, S. and Koutroumbas, K. (2008), Pattern recognition 4th edition, Academic Press.
  12. Viola, P. and Jones, M.J. (2004), Robust real-time face detection, International Journal of Computer Vision, Vol. 57, No. 2, pp. 137-154. https://doi.org/10.1023/B:VISI.0000013087.49260.fb