딥러닝 기반 LNGC 화물창 스캐닝 점군 데이터의 비계 시스템 객체 탐지 및 후처리
Object Detection and Post-processing of LNGC CCS Scaffolding System using 3D Point Cloud Based on Deep Learning
1
조회
0
다운로드
4
인용
0
참고문헌
키워드
초록
Recently, quality control of the Liquefied Natural Gas Carrier (LNGC) cargo hold and block-erection interference areas using 3D scanners have been performed, focusing on large shipyards and the international association of classification societies. In this study, as a part of the research on LNGC cargo hold quality management advancement, a study on deep-learning-based scaffolding system 3D point cloud object detection and post-processing were conducted using a LNGC cargo hold 3D point cloud. The scaffolding system point cloud object detection is based on the PointNet deep learning architecture that detects objects using point clouds, achieving 70% prediction accuracy. In addition, the possibility of improving the accuracy of object detection through parameter adjustment is confirmed, and the standard of Intersection over Union (IoU), an index for determining whether the object is the same, is achieved. To avoid the manual post-processing work, the object detection architecture allows automatic task performance and can achieve stable prediction accuracy through supplementation and improvement of learning data. In the future, an improved study will be conducted on not only the flat surface of the LNGC cargo hold but also complex systems such as curved surfaces, and the results are expected to be applicable in process progress automation rate monitoring and ship quality control.
Abstract
Recently, quality control of the Liquefied Natural Gas Carrier (LNGC) cargo hold and block-erection interference areas using 3D scanners have been performed, focusing on large shipyards and the international association of classification societies. In this study, as a part of the research on LNGC cargo hold quality management advancement, a study on deep-learning-based scaffolding system 3D point cloud object detection and post-processing were conducted using a LNGC cargo hold 3D point cloud. The scaffolding system point cloud object detection is based on the PointNet deep learning architecture that detects objects using point clouds, achieving 70% prediction accuracy. In addition, the possibility of improving the accuracy of object detection through parameter adjustment is confirmed, and the standard of Intersection over Union (IoU), an index for determining whether the object is the same, is achieved. To avoid the manual post-processing work, the object detection architecture allows automatic task performance and can achieve stable prediction accuracy through supplementation and improvement of learning data. In the future, an improved study will be conducted on not only the flat surface of the LNGC cargo hold but also complex systems such as curved surfaces, and the results are expected to be applicable in process progress automation rate monitoring and ship quality control.
참고문헌 (13)
- Charles, R.Q, "Frustum PointNets for 3D object detection from rgb-d data", Proceeding of the IEEE Conference on Computer Vision and Pattern Recognition (2018)
- Charles, R.Q, "PointNet:Deep learning on point sets for 3D classification and segmentation", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition
- Ham, D, "A study of method for machine learning application for using shipyard big data", Journal of Computational Design and Engineering, pp.186~190 (2016)
- Jaderberg, M, "Spatial transformer networks" (2015)
- 정용철, "LNG 화물창 정밀 구조해석을 위한 에폭시와 폴리우레탄 접착제 기계적 물성치 분석", 대한조선학회 논문집, 58(2), pp.66~72 (2021), 대한조선학회[DOI]
- 김연중, "딥러닝을 이용한 연안방재 시스템 구축에 관한 연구", 한국해양공학회지, 33(6), pp.590~596 (2019), 한국해양공학회[DOI]
- Kirillov, A, "Panoptic segmentation", Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (2019)
- 권혁천, "Stacking Ensemble Learning을 활용한 블록 탑재 시수 예측", 대한조선학회 논문집, 56(6), pp.488~496 (2019), 대한조선학회[DOI]
- Lee, D. K, "Simulation-based work plan verification in shipyards", Journal of Ship Production and Design, 30(2), pp.49~57 (2014)
- Maturana, D, "VoxNet: A 3D convolution neural network for real-time object recognition", IEEE, 2015, 922~928
- Russakovsky, O, "ImageNet large scale visual recognition challenge", Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (2015)
- 우종훈, "중립파일 기반의 조선소 시뮬레이션 환경에 관한 연구", 대한조선학회 논문집, 53(1), pp.18~28 (2016), 대한조선학회
- Yin, Z, "VoxelNet: End-to-end learning for point cloud based 3D object detection", Proceeding of the IEEE Conference on Computer Vision and Pattern Recognition (2017)