DOI QR코드

DOI QR Code

Development of a CNN-based Cross Point Detection Algorithm for an Air Duct Cleaning Robot

CNN 기반 공조 덕트 청소 로봇의 교차점 검출 알고리듬 개발

  • Yi, Sarang (Department of Mechanical Engineering, Kongju National University) ;
  • Noh, Eunsol (Department of Mechanical Convergence Engineering, Kongju National University) ;
  • Hong, Seokmoo (Department of Mechanical & Automotive Engineering, Kongju National University)
  • 이사랑 (공주대학교 기계공학과) ;
  • 노은솔 (공주대학교 융합기계공학과) ;
  • 홍석무 (공주대학교 기계자동차공학부)
  • Received : 2020.05.20
  • Accepted : 2020.08.07
  • Published : 2020.08.31

Abstract

Air ducts installed for ventilation inside buildings accumulate contaminants during their service life. Robots are installed to clean the air duct at low cost, but they are still not fully automated and depend on manpower. In this study, an intersection detection algorithm for autonomous driving was applied to an air duct cleaning robot. Autonomous driving of the robot was achieved by calculating the distance and angle between the extracted point and the center point through the intersection detection algorithm from the camera image mounted on the robot. The training data consisted of CAD images of the duct interior as well as the cross-point coordinates and angles between the two boundary lines. The deep learning-based CNN model was applied as a detection algorithm. For training, the cross-point coordinates were obtained from CAD images. The accuracy was determined based on the differences in the actual and predicted areas and distances. A cleaning robot prototype was designed, consisting of a frame, a Raspberry Pi computer, a control unit and a drive unit. The algorithm was validated by video imagery of the robot in operation. The algorithm can be applied to vehicles operating in similar environments.

건물 내부 공기 순환을 위한 공조 덕트는 장기간 사용 시 오염물질이 내부에 쌓여 인력 또는 로봇이 투입되어 청소가 주기적으로 수행된다. 청소는 작업시간과 인건비 문제를 해결하기 위해 최근 원격 조정으로 로봇을 작동시키는 방법이 사용되고 있다. 하지만 완전 자동화가 아니라 인력 의존적이며 청소 시간 단축에도 한계가 있다. 본 연구는 공조 덕트 청소 로봇 자율 주행을 위해 교차점 검출 알고리듬 개발에 대한 것이다. 자율 주행은 청소 로봇에 장착된 카메라 영상에서 교차점 검출 알고리듬을 통해 추출된 점과 중심점 사이의 거리 및 각도를 계산하여 로봇을 제어하도록 구성된다. 교차점 검출을 위한 데이터는 3D CAD 프로그램을 이용한 공조 덕트 내부 이미지를 Python을 이용해 교차점 좌표 및 두 경계선 각도를 추출하여 생성했다. 검출 알고리듬은 딥러닝 중 CNN 모델이 학습에 사용됐으며 학습 모델은 입력이미지에서 교차점 정보를 추출하며 학습 모델 정확도는 면적과 거리를 이용해 판단했다. 알고리듬 검증을 위해 청소 로봇을 제작했으며 로봇은 몸체, Raspberry Pi, 카메라 및 초음파 센서를 포함한 제어부, 모터와 바퀴를 포함한 구동부로 구성된다. 알고리듬을 탑재한 로봇 청소기 주행 영상을 통해 알고리듬을 검증했다. 향후 공조 덕트뿐만 아니라 에스컬레이터 등 다양한 환경에서 적용 가능할 것으로 기대된다.

Keywords

References

  1. H. S. Choi, B. Park, W. Jeong, S. B. Kwon, H. S. Kim, "Development of a Duct Cleaning Robot Platform", 2013 28th ICROS Annual Conference (ICROS 2013), Institute of Control, Robotics and Systems, Changwon-si, Korea, pp.185-186, May 2013.
  2. S. W. Jeon, W. Jeong, D. Park, S. B. Kwon, "Design of an Intelligent Duct Cleaning Robot with Force Compliant Brush", 2012 12th International Conference on Control (IEEE), Automation and Systems, Jeju island, Korea, pp.2033-2037, Oct. 2012.
  3. W. Jeong, S. W. Jeon, D. S. Park, "Force control method of duct cleaning robot brush to improve cleaning efficiency", Proceeding of KSME Annual Meeting, The Korean Society of Mechanical Engineers, Changwon-si, Korea, pp.2303-2308. Nov. 2012.
  4. W. Jeong, D. S. Park, S. B. Park, S. W. Jeon, "Development of Duct Cleaning System Using a Robot", 2013 28th ICROS Annual Conference (ICORS 2013), Institute of Control, Robotics and System, Changwon-si, Korea, pp.183-184, May 2013.
  5. W. Jeong, S. B. Park, S. W. Jeon, H. S. Kim, "Development of realtime control system for duct cleaning robot", Proceeding of 2014 KSMTE Spring Conference, The Korean Society of Manufacturing Technology Engineers, Changwon-si, Korea, pp.245-245, April 2014.
  6. K. T. Park, W. J. Jang, J. H. Ha, M. I. Seo, D. H. Kim, "Position Estimation and Control for Autonomous Duct Cleaning Robot", KSMTE Annual Autumn Conference 2018, The Korean Society of Manufacturing Technology Engineers, Seoul, Korea, pp.99-99, Oct. 2013.
  7. W. Jeong, D. S. Park, "Development of a Duct Cleaning Robot and Technology Trends for Subway Stations", Proceeding of 2011 Autumn Conference & Annual Meeting of the Korea Society for Railway, The Korea Society For Railway, Mokpo-si, Korea, pp.1335-1341, May 2011.
  8. T. Q. Hieu, S. Yeon, J. Kim, "Korean License Plate Recognition Using CNN", Journal of Institute of Korean Electrical and Electronics Engineers, Vol.23, No.4, pp.1337-1342, 2019. DOI: http://dx.doi.org/10.7471/ikeee.2019.23.4.1337
  9. 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, 2020. DOI: http://dx.doi.org/10.5762/KAIS.2020.21.2.189
  10. D. H. Kim, J. E. Ha, "Lane detection through epipole estimation by convolutional neural networks", Journal of Institute of Control, Vol.24, No.9, pp.851-856, 2018. DOI: http://dx.doi.org/10.5302/J.ICROS.2018.18.0111
  11. F. Chollet, "Keras", 2015.
  12. D. S. Ko, H. K. Lee, D. J. Hur "Using Topology Optimization, Light Weight Design of Vehicle Mounted Voltage Converter for Impact Loading", Computational Structural Engineering Institute of Korea, Vol.31, No.6, pp.353-358, 2018. DOI: https://doi.org/10.7734/COSEIK.2018.31.6.353