DOI QR코드

DOI QR Code

정밀도로지도 제작을 위한 모바일매핑시스템 기반 딥러닝 학습데이터의 자동 구축

Automatic Construction of Deep Learning Training Data for High-Definition Road Maps Using Mobile Mapping System

  • Choi, In Ha (Dept. of Spatial Information Engineering, Namseoul University) ;
  • Kim, Eui Myoung (Dept. of Drone.GIS Engineering, Namseoul University)
  • 투고 : 2021.04.07
  • 심사 : 2021.05.03
  • 발행 : 2021.06.30

초록

현재 정밀도로지도 구축 공정은 수작업의 비율이 높아 구축 시간과 비용의 한계가 따른다. 인공지능을 이용하여 정밀도로지도 제작을 자동화하기 위한 연구가 활발하게 진행되고 있으나 정밀도로지도 제작을 위한 학습데이터의 구축 또한 수동으로 이루어지고 있어 학습데이터를 자동으로 구축할 필요성이 있다. 이에 본 연구에서는 모바일매핑시스템으로 취득한 포인트 클라우드를 이용하여 영상으로 변환한 후, 임계치를 이용한 영상 재분류와 중첩 분석 등을 통해 도로 노면표시 영역을 추출하고 추출한 영역의 다각형 유형 분류를 통해 정밀도로지도 제작을 위한 딥러닝 학습데이터를 자동으로 구축하는 방법론을 제안하였다. 제안한 방법론을 통해 구축한 2,764개의 차선 데이터를 딥러닝 기반의 PointNet 모델에 학습한 결과 학습 정확도는 99.977%로 나타났으며, 학습된 모델을 이용하여 3가지 색상 유형의 차선을 예측한 결과 정확도는 99.566%로 나타났다. 따라서, 본 연구에서 제안한 방법론으로 정밀도로지도 구축을 위한 학습데이터를 효율적으로 제작할 수 있는 것을 알 수 있었으며, 도로 노면표시의 정밀도로지도 제작과정 또한 자동화할 수 있을 것으로 사료된다.

Currently, the process of constructing a high-definition road map has a high proportion of manual labor, so there are limitations in construction time and cost. Research to automate map production with high-definition road maps using artificial intelligence is being actively conducted, but since the construction of training data for the map construction is also done manually, there is a need to automatically build training data. Therefore, in this study, after converting to images using point clouds acquired by a mobile mapping system, the road marking areas were extracted through image reclassification and overlap analysis using thresholds. Then, a methodology was proposed to automatically construct training data for deep learning data for the high-definition road map through the classification of the polygon types in the extracted regions. As a result of training 2,764 lane data constructed through the proposed methodology on a deep learning-based PointNet model, the training accuracy was 99.977%, and as a result of predicting the lanes of three color types using the trained model, the accuracy was 99.566%. Therefore, it was found that the methodology proposed in this study can efficiently produce training data for high-definition road maps, and it is believed that the map production process of road markings can also be automated.

키워드

과제정보

이 논문은 2020년도 정부(국토교통부)의 재원으로 공간정보 융복합 핵심인재 양성사업의 지원을 받아 수행된 연구임 (2020-02-01).

참고문헌

  1. Axelsson, P. (2000), DEM generation from laser scanner data using adaptive TIN model, International Archives of Photogrammetry and Remote Sensing, Vol. 33, pp. 110-117.
  2. Bleau, A. and Leon, L.J. (2000), Watershed-based segmentation and region merging, Computer Vision and Image Understanding, Vol. 77, No. 3, pp. 317-370. https://doi.org/10.1006/cviu.1999.0822
  3. Bodansky, E., Gribov, A., and Pilouk, M. (2002), Smoothing and compression of lines obtained by raster-to-vector conversion, Lecture Notes in Computer Science 2390, Springer, pp. 256-265.
  4. Douglas, D.H. and Peucker, T.K. (1973), Algorithms for the reduction of the number of points required to represent a digitised line or its caricature, The Canadian Cartographer, Vol. 10, No. 2, pp. 112-122. https://doi.org/10.3138/FM57-6770-U75U-7727
  5. Hong, S.P. and Kim, E.M. (2018), Classification of 3D road objects using machine learning, The Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography, Vol. 36, No. 6, pp. 535-544. (in Korean with English abstract) https://doi.org/10.7848/KSGPC.2018.36.6.535
  6. Jenson, S.K. and Domingue, J.O. (1988), Extracting topographic structure from digital elevation data for geographic information system analysis, Photogrammetric Engineering and Remote Sensing, Vol. 54, No. 11, pp. 1593-1600.
  7. Lee, S.J., Kim, J.S., Yoon, J.S., Shin, S.H., Bailo, O., Kim, N.I., Lee, T.H., Hong, H.S., Han, S.H., and Kweon, I.S. (2017), VPGNet: vanishing point guided network for lane and road marking detection and recognition, International Conference on Computer Vision (ICCV), pp. 1947-1955.
  8. Park, Y.K., Kang, W.P., Choi, J.E., and Kim, B.J. (2019), A study on the evaluation of real-time map update technology for automated driving, Journal of the Korean Association of Geographic Information Studies, Vol. 22. No. 3, pp. 146-154. (in Korean with English abstract)
  9. Qi, C.R, Su, H., Mo, K., and Gulibas, L.J (2017), PointNet: deep learning on point sets for 3D classification and segmentation, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 652-660.
  10. So, J.H. and Moon, Y.J. (2018), Plan for autonomous cooperation driving safety and infrastructure implementation, The Journal of The Korean Institute of Communication Sciences, Vol. 35, No. 5, pp. 37-43.
  11. Soilan, M., Riveiro, B., Martinez-Sanchez, J., and Arias, P. (2017). Segmentation and classification of road markings using MLS data. ISPRS Journal of Photogrammetry and Remote Sensing. Vol. 123, pp. 94-103. https://doi.org/10.1016/j.isprsjprs.2016.11.011
  12. Wang, Z., Ren, W., and Qiu, Q. (2018). LaneNet: real-time lane detection networks for autonomous driving. arXiv:1807.01726.
  13. Wen, C., Sun, X., Li, J., Wang, C., Guo. Y., and Habib, A. (2019). A deep learning framework for road marking extraction, classification and completion from mobile laser scanning point clouds, ISPRS Journal of Photogrammetry and Remote Sensing, Vol. 147, pp. 178-192. https://doi.org/10.1016/j.isprsjprs.2018.10.007
  14. Wold, S., Esbensen, K., and Geladi, P. (1987), Principal component analysis. Chemometrics and Intelligent Laboratory Systems, Vol. 2, pp. 37-52. https://doi.org/10.1016/0169-7439(87)80084-9