Deep Learning-Based Instance Segmentation for Enhanced Navigation of Agricultural Vehicles

Renato de Avila Lopes; Marcus Vinicius Leal de Carvalho; Edson Kitani; Francisco de Assis Zampirolli; Leopoldo Yoshioka; Luiz Antonio Celiberto Junior; Ugo Ibusuki

doi:10.22456/2175-2745.143329

Authors

Renato de Avila Lopes Universidade Federal do ABC (UFABC) https://orcid.org/0009-0002-8108-1326
Marcus Vinicius Leal de Carvalho Universidade de São Paulo (USP) https://orcid.org/0000-0002-1298-8517
Edson Kitani Faculdade de Tecnologia do Estado de São Paulo (FATEC)
Francisco de Assis Zampirolli Universidade Federal do ABC (UFABC) https://orcid.org/0000-0002-7707-1793
Leopoldo Yoshioka Universidade de São Paulo (USP)
Luiz Antonio Celiberto Junior Universidade Federal do ABC (UFABC)
Ugo Ibusuki Universidade Federal do ABC (UFABC)

DOI:

https://doi.org/10.22456/2175-2745.143329

Keywords:

Instance Segmentation, Yolov8, Computer Vision, Autonomous Vehicle

Abstract

This paper presents the development of a computer vision application based on the YOLOv8 network, designed to assist the navigation of autonomous vehicles on rural roads, particularly those found in sugarcane fields. The application employs instance segmentation to differentiate between navigable and non-navigable areas and detect obstacles such as pedestrians, vehicles, and other potential hazards. This information is used to generate an occupancy map that helps the navigation planner identify the safest and most efficient routes. The system was trained on a dataset containing 1.018 images, and the results demonstrate that instance segmentation significantly enhances the precision and safety of autonomous navigation in complex rural environments. The proposed approach is compatible with the ROS2 framework, using its structure for data integration and enabling real-time decision making.

Downloads

Download data is not yet available.

References

BORGES, P. V. et al. A survey on terrain traversability analysis for autonomous ground vehicles: Methods, sensors, and challenges. Field Rob., v. 2, n. 1, p. 1567–1627, 2022.

YASUDA, Y. D.; MARTINS, L. E. G.; CAPPABIANCO, F. A. Autonomous visual navigation for mobile robots: A systematic literature review. ACM Comp. Surveys (CSUR), ACM New York, NY, USA, v. 53, n. 1, p. 1–34, 2020.

JOCHER, G.; CHAURASIA, A.; QIU, J. Ultralytics YOLOv8. 2023. Disponível em: ⟨https://github.com/ultralytics/ultralytics⟩.

SZELISKI, R. Computer vision: algorithms and applications. [S.l.]: Springer Nature, 2022.

GOODFELLOW, I.; BENGIO, Y.; COURVILLE, A. Deep Learning. [S.l.]: MIT Press, 2016. ⟨http://www.deeplearningbook.org⟩.

HAFIZ, A. M.; BHAT, G. M. A survey on instance segmentation: state of the art. Int. Jour. of Mult. Inf. Retr., Springer, v. 9, n. 3, p. 171–189, 2020.

RACINSKIS, P.; ARENTS, J.; GREITANS, M. Constructing maps for autonomous robotics: An introductory conceptual overview. Electronics, MDPI, v. 12, n. 13, p. 2925, 2023.

CRISTÓFORIS, P. D. et al. Hybrid vision-based navigation for mobile robots in mixed indoor/outdoor environments. P. R. Letters, Elsevier, v. 53, p. 118–128, 2015.

FOUNDATION, O. S. R. Robot Operating System 2 (ROS2). 2024. ⟨https://docs.ros.org/en/foxy/⟩. Acesso: 2024-08-20.

MACENSKI, S. et al. From the desks of ROS maintainers: A survey of modern & capable mobile robotics algorithms in the robot operating system 2. Robotics and Autonomous Systems, Elsevier, v. 168, p. 104493, 2023.

HUANG, Z. Semantic road segmentation based on adapted Poly-YOLO. In: J. of Phy.: Conf. Series. [S.l.: s.n.], 2023. v. 2580, n. 1, p. 012015.

HURTIK, P. et al. Poly-YOLO: higher speed, more precise detection and instance segmentation for YOLOv3. Neural Comp. and Appl., Springer, v. 34, n. 10, p. 8275–8290, 2022.

OJHA, A.; SAHU, S. P.; DEWANGAN, D. K. Vehicle detection through instance segmentation using Mask R-CNN for intelligent vehicle system. In: IEEE. Int. Conf. on Intel. Comp. and Control Systems (ICICCS). [S.l.], 2021. p. 954–959.

WU, W. et al. Instance segmentation of tea garden roads based on an improved YOLOv8n-seg model. Agricul., MDPI AG, v. 14, n. 7, p. 1163, 2024.

LOPES, R. de A. Segmentation Article. 2024. ⟨https://github.com/natoavilalopes/segmentationArticle⟩. Acesso: 2024-10-09.

ROBOFLOW. Roboflow. 2024. ⟨https://roboflow.com⟩. Acesso: 2024-10-09.

AVILA, R. et al. Comparisons of neural networks using computer vision for agricultural automation. In: 2023 15th IEEE International Conference on Industry Applications (INDUSCON). [S.l.: s.n.], 2023. p. 466–470.

KETKAR, N. et al. Introduction to PyTorch. Deep learning with Python: learn best practices of deep learning models with PyTorch, Springer, p. 27–91, 2021.

TEAM, O. OpenCV: Open Source Computer Vision Library. 2023. ⟨https://opencv.org⟩. Acesso: 2024-08-20.

DEVELOPERS, R. RViz. 2024. ROS Visualization Tool. ⟨http://wiki.ros.org/rviz⟩.

YUE, X. et al. Improved YOLOv8-seg network for instance segmentation of healthy and diseased tomato plants in the growth stage. Agriculture, MDPI, v. 13, n. 8, p. 1643, 2023.

HAIMER, Z. et al. Pothole detection: A performance comparison between YOLOv7 and YOLOv8. In: 2023 9th International Conference on Optimization and Applications (ICOA). [S.l.: s.n.], 2023. p. 1–7.

PADILLA, R.; NETTO, S. L.; SILVA, E. A. D. A survey on performance metrics for object-detection algorithms. In: IEEE. 2020 Int. Conf. on Systems, Signals and Image Processing (IWSSIP). [S.l.], 2020. p. 237–242.