Exploiting Machine Learning for Vision and Motion Planning of Autonomous Vehicles Navigation
DOI:
https://doi.org/10.22213/2410-9304-2021-3-95-104Keywords:
autonomous vehicles, perception, control, localisation, navigation, motion planningAbstract
The object of this paper is to create a system that can control any vehicle in any gaming environment to simulate, study, experiment and improve how self-driving vehicles operate. It is to be taken as the bases for future work on autonomous vehicles with real hardware devices. The long-term goal is to eliminate human error. Perception, localisation, planning and control subsystems were developed. LiDAR and RADAR sensors were used in addition to a normal web Camera. After getting information from the perception module, the system will be able to localise where the vehicle is, then the planning module is used to plan to which location the vehicle will move, using localisation module data to draw up the best path to use. After knowing the best path, the system will control the vehicle to move autonomously without human help. As a controller a Proportional Integral Derivative PID controller was used. Python programming language, computer vision, and machine learning were used in developing the system, where the only hardware required is a computer with a GPU and powerful graphical card that can run a game which has a vehicle, roads with lane lines and a map of the road. The developed system is intended to be a good tool in conducting experiments for achieving reliable autonomous vehicle navigation.References
Feng, L., Everett, H. R., Borenstein, J. (1994). "Where am I?": sensors and methods for autonomous mobile robot positioning. University of Michigan.
Katrakazas, C., Quddus, M., Chen, W.-H., & Deka, L. (2015). Real-time motion planning methods for autonomous on-road driving: State-of-the-art and future research directions. Transportation Research Part C: Emerging Technologies, 60, 416-442. doi.org/10.1016/j.trc.2015.09.011.
Pendleton, S., Andersen, H., Du, X., Shen, X., Meghjani, M., Eng, Y., Rus, D., & Ang, M. (2017). Perception, Planning, Control, and Coordination for Autonomous Vehicles. Machines, 5(1), 6. Available at: https://doi.org/10.3390/machines5010006.
Badue, C., et al. (2019). Self-Driving Cars: A Survey. arXiv:1901.04407v2.
Minton, S. (1993). The Morgan KAUFMANN Series in Machine Learning. Machine Learning Methods for Planning, p. ii. doi:10.1016/b978-1-4832-0774-2.50001-9.
Ganguly, S. (2021). Machine Learning. Quantum Machine Learning: An Applied Approach. Apress. pp. 41-97.
Brownlee, J. (2020). 10 Clustering Algorithms with Python. Machine Learning Mastery. Available at: https://machinelearningmastery.com/clustering-algorithms-with-python.
Cohen, J. (2021). Deep Learning Algorithms in Self-Driving Cars. Medium. Available at: https://becominghuman.ai/deep-learning-algorithms-in-self-driving-cars-14b13a895068.
Road Lane line detection-Computer Vision Project in Python. DataFlair. (2021). Available at: https://data-flair.training/blogs/road-lane-line-detection.
Ferdinand, N. (2020). A Deep Dive into Lane Detection with Hough Transform. Medium. Available at: https://towardsdatascience.com/a-deep-dive-into-lane-detection-with-hough-transform-8f90fdd1322f.
Johnson M. (2005). PID Control Technology. In: Johnson M.A., Moradi M.H. (eds) PID Control. Springer, London. Available at: https://doi.org/10.1007/1-84628-148-2_1.
Crowe J., Johnson M. (2005). Phase-Locked Loop Methods and PID Control. In: Johnson M.A., Moradi M.H. (eds) PID Control. Springer, London. Available at: https://doi.org/10.1007/1-84628-148-2_7.
Semmlow J. L. (2008). Fundamentals of Image Processing: Matlab Image Processing Toolbox. Bio-signal and Medical Image Processing, pp. 293-325.
Kittler, J. (1985). Image Processing System Architectures. Research Studies Press, Wiley.
