DOI QR코드

DOI QR Code

드론과 이미지 분석기법을 활용한 구조물 외관점검 기술 연구

Study on Structure Visual Inspection Technology using Drones and Image Analysis Techniques

  • 투고 : 2017.07.17
  • 심사 : 2017.11.22
  • 발행 : 2017.12.20

초록

이 연구는 사회 기반 구조물의 노후화에 대한 안전점검 기술분야에서 구조물 외관점검 기술의 효율적 대안에 관한 연구이다. 기존 육안점검 및 조사를 대신하여 산업용 드론과 딥 러닝기반의 이미지 분석 기법을 접목함으로써 막대한 인력과 시간소요 및 비용을 절감하고 높은 구역 및 돔 구조물의 접근 한계를 극복하고자 하였다. 구조물의 0.3mm 이상의 균열 손상을 검지할 수 있는 고 해상도 카메라와 라이다 센서, 임베디드 이미지 프로세서 모듈로 구성된 탑재체를 제작하여 산업용 드론에 탑재하였다. 이를 현장 시험에 적용하여 자동비행항법을 통해 시편의 손상 이미지를 촬영하였다. 또한 균열경을 이용하여 기존 육안 점검 방법으로 백태, 박리박락과 같은 면적형 손상과 선형 손상인 균열의 폭과 길이를 측정하여 최종 이미지 분석 검출 결과와 비교하고자 하였다. 촬영된 이미지 중 80장의 샘플을 골라 이미지 분석 기법을 적용하여 사전처리작업(pre-processing)-분리작업(segmentation)-특징점 추출작업(feature extraction)-분류 작업(Classification)-지도학습작업(supervised learning) 등의 과정을 거쳐 손상을 분리하고, 이를 딥러닝 기반 플랫폼으로 지도학습하여 분석 파라미터를 추출하였다. 지도학습을 수행하지 않은 임의의 이미지 샘플 60장을 신규로 추가하여 추출된 파라미터를 기반으로 이미지 분석을 수행한 결과, 손상 검출율의 90.5%로 나타났다.

The study is about the efficient alternative to concrete surface in the field of visual inspection technology for deteriorated infrastructure. By combining industrial drones and deep learning based image analysis techniques with traditional visual inspection and research, we tried to reduce manpowers, time requirements and costs, and to overcome the height and dome structures. On board device mounted on drones is consisting of a high resolution camera for detecting cracks of more than 0.3 mm, a lidar sensor and a embeded image processor module. It was mounted on an industrial drones, took sample images of damage from the site specimen through automatic flight navigation. In addition, the damege parts of the site specimen was used to measure not only the width and length of cracks but white rust also, and tried up compare them with the final image analysis detected results. Using the image analysis techniques, the damages of 54ea sample images were analyzed by the segmentation - feature extraction - decision making process, and extracted the analysis parameters using supervised mode of the deep learning platform. The image analysis of newly added non-supervised 60ea image samples was performed based on the extracted parameters. The result presented in 90.5 % of the damage detection rate.

키워드

참고문헌

  1. Jung HJ. Current status and global forecast of commercial drones market. Korea aerospace research institute. Daejeon (Korea): Future strategy head office of KARI. 2016 Jul 5-10 p. Report No.: Aviation Issue No.7.
  2. Lee GJ. Epartment of mechanical and aerospace engineering. Seoul national university aerospace vehicle design laboratory. Seoul(Korea): gongsang; 2017. 52 p. Korean.
  3. Lovelace B. Unmanned aerial vehicle bridge inspection demonstrati on project. Minnesota(USA): Minnesota department of transportation(USA); 2015 Jul. 214 p. Report No.: 2015-40.
  4. Chan B. Towards UAV-based bridge inspection systems: a review and an application perspective. Structural monitoring and maintenance. 2015 Feb 3;2(3): 283-300. https://doi.org/10.12989/smm.2015.2.3.283
  5. Lu M. Research on auto-tracking algorithm for power line inspection based on unmanned aerial vehicle. Power and energy engineering conference(APPEEC); 2012 Mar 27-29; Shanghai, China. Shanghai(China): Power and energy engineering IEEE org; 2012 p. 1-2.
  6. Siegel M. Remote enhanced visual inspection of aircraft by a mobile robot. 1998 IEEE workshop on emerging technologies; 1998 May 15-16; Minnesota, USA. Minnesota (USA): The robotics institute school of computer science carnegie mellon university; 1998. p. 10-1.
  7. Nieniewski M. Morphological detection and feature-based classification of cracked regions in ferrites. Machine graphics & vision. 1999 Jan;8(4):699-712.
  8. Salembier P. Comparison of some morphological segmentation algorithms based on contrast enhancement. - application to automatic defect detection. In: European signal processing conference; 1990 Sep 18-21; Barcelona, Spain. Barcelona (Spain): Elsevier; 1990. p. 833-6.
  9. Moselhi O, Shehab E. Prediction of sewer pipe main condition using the linear regression approach. Journal of geoscience and environment protection. 2016 May 19;4(5):97-104.
  10. Lee HB. Development of automatic crack detection system for concrete structure using image processing method. Journal of the korea institute for structural maintenance inspection. 2012 Jan;16(1):64-77. https://doi.org/10.11112/jksmi.2012.16.1.064
  11. Gonzalez RC, Wintz P. Digital image processing. 2nd ed. Boston: Addison-Wesley; 1987. 793 p.
  12. Pratt WK. Digital image processing. 3rd ed. New York: Wiley; 2001. 793 p.
  13. Gonzalez RC, Woods RE, Eddins SL. Digital image processing using MATLAB. 2nd ed. New York: Mcgraw hill; 2011. 738 p.
  14. Brown LG. A Survey of image Registration Techniques. ACM computing surveys. 1992 Dec;24(4):325-76. https://doi.org/10.1145/146370.146374
  15. Salembier P. Comparison of some morphological segmentation algorithms based on contrast enhancement. - application to automatic defect detection. In: European signal processing conference; 1990 Sep 18-21; Barcelona, Spain. Barcelona (Spain): Elsevier; 1990. p. 833-6.
  16. Nobuyuki O. A threshold selection method from gray-level histograms. IEEE Transactions on systems man and cybernetics. 1979 Jan;9(1):62-6. https://doi.org/10.1109/TSMC.1979.4310076
  17. Sinha SK, Fieguth PW. Morphological segmentation and classification of underground pipe images. Machine vision and applications. 2006 Apr;17(1):21-31. https://doi.org/10.1007/s00138-005-0012-0
  18. Zitova B, Flusser J. Image registration methods: a survey. Image and vision computing. 2003 Oct;21(11):977-1000. https://doi.org/10.1016/S0262-8856(03)00137-9
  19. Welling M. Fisher linear discriminant analysis. University of toronto. Toronto: Department of computer science; 2005. 5 p.
  20. Harris C, Stephens M. A combined corner and edge detector. In: Alvey vision conference; 1988 Sep 1-2; Manchester, UK. Romsey (UK): The Plessey Company pic; 1988. p. 147-52.
  21. Ho SK, White RM, Lucas J. A vision system for automated crack detection in welds. measurement science and technology. 1990 Mar;1(3):287-94. https://doi.org/10.1088/0957-0233/1/3/015