Fuel Tank Position Localization Based on Machine Vision
DOI:
https://doi.org/10.62051/ijcsit.v4n2.06Keywords:
Machine Vision, ResNet, Convolutional Neural Networks, Kalman Filter (EKF), Bayesian Filtering Algorithm, Refueling Robot, RobustnessAbstract
Refueling robots, as an important part of intelligent service systems, greatly enhance the safety and efficiency of the refueling process. This study focuses on the machine vision module in refueling robots, specifically exploring the application of high-resolution cameras and LiDAR in data acquisition. We used Convolutional Neural Networks (CNNs) such as ResNet and MobileNet for feature extraction, which ensured high-precision recognition and classification in a variety of environments. Meanwhile, the target detection module uses YOLOv4 and MobileNet3 with fast and accurate target localization capabilities to effectively identify and calibrate the location of fueling ports. In addition, we introduced Extended Kalman Filter (EKF) and Bayesian Filter algorithms for data fusion and state estimation, which improves the robustness and reliability of the system. By combining these advanced vision techniques and algorithms, the refueling robot realizes efficient and accurate automatic refueling operation. This research provides theoretical and technical support for the further development of intelligent refueling robots so that they can still operate stably in complex environments.
Downloads
References
[1] Rega, G., Settimi, V., & Lenci, S. (2020b). Chaos in one-dimensional structural mechanics. Nonlinear Dynamics, 102(2), 785–834. https://doi.org/10.1007/s11071-020-05849-3Rega, G., Settimi, V., & Lenci, S. (2020b). Chaos in one-dimensional structural mechanics. Nonlinear Dynamics, 102(2), 785–834. https://doi.org/10.1007/s11071-020-05849-3Thermocouples: Figliola and Beasley, pp. 342-357.
[2] C. Szegedy, S. Ioffe, V. Vanhoucke, A.A. Alemi. Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning. Proceedings of the Thirty-First AAAI Conference on Artificial IntelligenceFebruary, pp. 4278–4284, 2017.
[3] Huang, L., Xiang, Z., Yun, J., Sun, Y., Liu, Y., Jiang, D., Ma, H., & Yu, H. (2023). Target detection based on Two-Stream Convolution neural network with Self-Powered Sensors information. IEEE Sensors Journal, 23(18), 20681–20690. https://doi.org/10.1109/jsen.2022.3220341
[4] D.T. Nguyen, T.N. Nguyen, H. Kim, H-J Lee. A High-Throughput and Power-Efficient FPGA Implementa-tion of YOLO CNN for Object Detection. IEEE Transactions on Very Large Scale Integration (VLSI) Systems; vol. 27, pp. 1861–1873, 2019.
[5] MASAZADE E, NIU R, VARSHNEY P K, et al. Energy aware iterative source localization for wireless sensor networks [J]. IEEE Transactions on Signal Processing, 2010, 58(9)4824-4835.
[6] CARMI A. Sensor scheduling via compressed sensing[C]// Proceedings of 13th international Conference on Information Fusion. Edinburgh: IEEE, 2010:1-8.
[7] BOYD S, PARIKH N, CHU E, et al. Distributed optimization and statistical learning via the alternating direction method of multipliers [J]. Found. Trends Mach. Learn, 2011,3(1):1-122.
[8] Ziar ko W. Variable Precisio n Roug h Set Mo del [J]. J of Comp uter Sy stems and Science, 1993, 46(1): 39-59.
[9] Bo gdan K. Finding Locat ion Using a Part icle Filter and Histo g ram M atching [A]. P roc of A rtif icial I ntelligence and Sof t Comp uting [C]. Poland: Spring er, 2004: 786-791.
[10] Andr ew H, Sa jid S. An Ex perimental Study of Lo calization Using Wir eless Ether net [A]. Pr oc of the I nt Conf on Field and Serv ice Robotics [C]. Lake Yamanaka, 2003: 201-206.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
How to Cite
