Setembro 2023 vol. 10 num. 1 - XXX Simpósio Internacional de Engenharia Automotiva

Trabalho completo - Open Access.

Idioma principal | Segundo idioma

Sistemas de visão aplicados em processos de qualidade automotiva: Uma revisão sistemática

Vision systems applied in automotive quality processes: a systematic review

MARTINS, Vinicios da Silva ; SANTOS JR, Francisco Magalhães dos ; MELO, Leonardo Mendes De ; SILVA, Giovani Costa ; SOUSA, Gabriel Estevam ; BIRAIS, Pedro Henrique ; SOUSA, Bruno Araujo de ;

Trabalho completo:

A indústria automobilística produz milhões de veículos anualmente, tornando-se um setor vital globalmente. Processos de inspeção, incluindo verificações visuais e sistemas de visão computacional, garantem a qualidade e segurança dos produtos. Avanços recentes em técnicas de aprendizado de máquina e visão computacional impulsionaram o uso de sistemas de visão computacional para inspeção automotiva. Esses sistemas aumentam a eficiência e a qualidade da produção ao detectar defeitos e anomalias que passam despercebidos pelos operadores humanos. Para obter uma compreensão atualizada e identificar pesquisas científicas relevantes sobre inspeção automotiva usando visão computacional e aprendizado de máquina, uma revisão sistemática intitulada "Sistemas de Visão Computacional Aplicados em Processos de Qualidade Automotiva" analisou o banco de dados da Web of Science. Inicialmente, foram encontrados 220 artigos, mas após a aplicação de filtros, foram selecionados 16 trabalhos. O estudo utilizou análise de rede semântica para identificar tendências de inspeção e examinou patentes e registros de software. Esta pesquisa tem como objetivo consolidar e atualizar o conhecimento científico nessa área, ao mesmo tempo em que auxilia na tomada de decisões e serve como referência para pesquisadores

Trabalho completo:

The automotive industry produces millions of vehicles annually, making it a vital sector globally. Inspection processes, including visual checks and computer vision systems, ensure product quality and safety. Recent advances in machine learning and computer vision techniques have propelled the use of computer vision systems for automotive inspection. These systems enhance efficiency and production quality by detecting defects and anomalies missed by human operators. To gain a current understanding and identify relevant scientific research on automotive inspection using computer vision and machine learning, a systematic review titled "Computer Vision Systems Applied in Automotive Quality Processes" analyzed the Web of Science database. Initially, 220 articles were found, but after applying filters, 16 works were selected. The study employed semantic network analysis to identify inspection trends and examined patents and software records. This research aims to consolidate, organize, and update scientific knowledge in this field while assisting decision-making and serving as a reference for other researchers.

Download (PDF)

Palavras-chave: ,

Palavras-chave: ,

DOI: 10.5151/simea2023-PAP35

Referências bibliográficas
  • [1] " G. Cavaliere, Y. Borgianni, and E. Rampone,
  • [2] “Development of a System for the Analysis of
  • [3] Surface Defects in Die-Cast Components Using
  • [4] Machine Vision,” 2022, pp. 74–86. doi:
  • [5] 10.1007/978-3-031-14317-5_7.
  • [6] [2] N. D. Hoang, Q. L. Nguyen, and D. T. Bui, “Image
  • [7] Processing-Based Classification of Asphalt
  • [8] Pavement Cracks Using Support Vector Machine
  • [9] Optimized by Artificial Bee Colony,” JOURNAL OF
  • [10] COMPUTING IN CIVIL ENGINEERING, vol. 32,
  • [11] no. 5, 2018, doi: 10.1061/(ASCE)CP.1943-
  • [12] 5487.0000781.
  • [13] [3] H. Nishiura et al., “Machine-learning-based qualitylevel-estimation system for inspecting steel
  • [14] microstructures,” Microscopy, vol. 71, no. 4, pp.
  • [15] 214–221, 2022.
  • [16] [4] J. P. Yun, W. C. Shin, G. Koo, M. S. Kim, C. Lee,
  • [17] and S. J. Lee, “Automated defect inspection system
  • [18] for metal surfaces based on deep learning and data
  • [19] augmentation,” J Manuf Syst, vol. 55, pp. 317–324,
  • [20] 2020, doi: 10.1016/j.jmsy.2003.009.
  • [21] [5] E. Protopapadakis, A. Voulodimos, A. Doulamis, N.
  • [22] Doulamis, and T. Stathaki, “Automatic crack
  • [23] detection for tunnel inspection using deep learning
  • [24] and heuristic image post-processing,” APPLIED
  • [25] INTELLIGENCE, vol. 49, no. 7, pp. 2793–2806,
  • [26] 2019, doi: 10.1007/s10489-018-01396-y.
  • [27] [6] R. Rayhana, Y. T. Jiao, Z. Bahrami, Z. Liu, A. Wu,
  • [28] and X. J. Kong, “Valve Detection for Autonomous
  • [29] Water Pipeline Inspection Platform,” IEEE-ASME
  • [30] TRANSACTIONS ON MECHATRONICS, vol. 27,
  • [31] no. 2, pp. 1070–1080, 2022, doi:
  • [32] 10.1109/TMECH.2021.3079409.
  • [33] [7] P. J. Navarro-Lorente and I. J. Moreo-Lopez,
  • [34] “Machine learning system based on computer vision
  • [35] for the automatic inspection of magnetic particles in
  • [36] marine structures,” Dyna (Medellin), vol. 93, no. 6,
  • [37] pp. 636–642, 2018, doi: 10.6036/8820.
  • [38] [8] C. Beltrán-González, M. Bustreo, and A. Del Bue,
  • [39] “External and internal quality inspection of
  • [40] aerospace components,” in 2020 IEEE 7th
  • [41] International Workshop on Metrology for
  • [42] AeroSpace (MetroAeroSpace), IEEE, 2020, pp. 351–
  • [43] 355.
  • [44] [9] Y. Minglei, Z. Xingjie, and Z. Xiaoling, “Research
  • [45] on character correction method based on machine
  • [46] learning,” in Proceedings - 2020 3rd World
  • [47] Conference on Mechanical Engineering and
  • [48] Intelligent Manufacturing, WCMEIM 2020, Institute
  • [49] of Electrical and Electronics Engineers Inc., Dec.
  • [50] 2020, pp. 136–142. doi:
  • [51] 10.1109/WCMEIM52463.2020.00035.
  • [52] [10] R. Pramudita and F. I. Hariadi, “Development of
  • [53] techniques to determine object shifts for PCB board assembly automatic optical inspection (AOI),” in
  • [54] 2018 International Symposium on Electronics and
  • [55] Smart Devices (ISESD), IEEE, 2018, pp. 1–4.
  • [56] [11] Y. T. Li et al., “An integrated underwater structural
  • [57] multi-defects automatic identification and
  • [58] quantification framework for hydraulic tunnel via
  • [59] machine vision and deep learning,” STRUCTURAL
  • [60] HEALTH MONITORING-AN INTERNATIONAL
  • [61] JOURNAL, doi: 10.1177/14759217221122316.
  • [62] [12] R. K. W. Vithanage, C. S. Harrison, and A. K. M. M.
  • [63] DeSilva, “Enhance 3D point cloud accuracy through
  • [64] supervised machine learning for automated rolling
  • [65] stock maintenance: A railway sector case study,” in
  • [66] 2018 International Conference on Computing,
  • [67] Electronics & Communications Engineering
  • [68] (iCCECE), IEEE, 2018, pp. 242–246.
  • [69] [13] J. Karangwa, L. H. Kong, D. R. Yi, and J. S. Zheng,
  • [70] “Automatic optical inspection platform for real-time
  • [71] surface defects detection on plane optical
  • [72] components based on semantic segmentation,” Appl
  • [73] Opt, vol. 60, no. 19, pp. 5496–5506, 2021, doi:
  • [74] 10.1364/AO.424547.
  • [75] [14] S. Y. Zhang, Q. J. Zhang, J. F. Gu, L. Su, K. Li, and
  • [76] M. Pecht, “Visual inspection of steel surface defects
  • [77] based on domain adaptation and adaptive
  • [78] convolutional neural network,” Mech Syst Signal
  • [79] Process, vol. 153, 2021, doi:
  • [80] 10.1016/j.ymssp.2020.107541.
  • [81] [15] L. Jiang and X. Li, “An efficient and accurate object
  • [82] detection algorithm and its application,” in 2020
  • [83] IEEE 5th Information Technology and Mechatronics
  • [84] Engineering Conference (ITOEC), IEEE, 2020, pp.
  • [85] 656–661"
Como citar:

MARTINS, Vinicios da Silva; SANTOS JR, Francisco Magalhães dos; MELO, Leonardo Mendes De; SILVA, Giovani Costa; SOUSA, Gabriel Estevam; BIRAIS, Pedro Henrique; SOUSA, Bruno Araujo de; "Sistemas de visão aplicados em processos de qualidade automotiva: Uma revisão sistemática", p. 249-256 . In: Anais do XXX Simpósio Internacional de Engenharia Automotiva . São Paulo: Blucher, 2023.
ISSN 2357-7592, DOI 10.5151/simea2023-PAP35

últimos 30 dias | último ano | desde a publicação


downloads


visualizações


indexações