INTEGRASI VISUAL COMPUTING UNTUK SISTEM PENGEREMAN ADAPTIF ROBOT INSPEKSI REL OTONOM
Keywords:
Night Vision Zero-DCE , YOLOv8n, Human Pose Estimation, Stereo Vision, Autonomous Braking SystemAbstract
The continuous operation of railway inspection systems over a 24-hour cycle necessitates the implementation of supporting technologies that can function optimally under all lighting conditions. This study develops an autonomous braking system powered by artificial intelligence and computer vision, incorporating Zero-DCE++ Night Vision enhancement, YOLOv8n object detection, and Human Pose Estimation on the LuminoLynx Inspection Robot. The system adopts a Stereo Vision triangulation approach for distance estimation and is integrated with an ESP32 microcontroller to enable real-time decision-making based on object proximity. Under low-light static testing, the system detected human objects with 100% accuracy, achieving the smallest distance error of 0.9 meters (3.37%) and a braking response within 3 seconds. For vehicle objects, the system yielded a confidence score of 0.62, a 3.61-meter error (36.1%), and a braking time of 6 seconds. These results confirm that integrating night vision and AI-based detection improves detection accuracy and braking responsiveness in low-light conditions. Although detection remained functional during dynamic testing, camera vibration reduced the precision of stereo matching, impacting distance estimation accuracy. Future improvements such as camera stabilization or depth sensor integration are recommended. This research contributes to advancing autonomous braking technology for nighttime railway inspections.
References
DAFTAR PUSTAKA
Ciptaningrum, A. (2023). Design of smart human following on rail inspection using human pose estimation marker-less motion capture based on blazepose. Journal Geuthee of Engineering and Energy (JOGE), 2(2), 106–118. https://doi.org/10.52626/joge.v2i2.26
Guo, C., Li, C., Guo, J., Loy, C. C., Hou, J., Kwong, S., & Cong, R. (2020). Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 1777–1786. https://doi.org/10.1109/CVPR42600.2020.00185
Lei, S., Yi, H., & Sarmiento, J. S. (2024). Synchronous End-to-End Vehicle Pedestrian Detection Algorithm Based on Improved YOLOv8 in Complex Scenarios. Sensors, 24(18), 6116. https://doi.org/10.3390/s24186116
Li, J., Durazo-Cardenas, I., Ruiz-Carcel, C., He, F., Anderson, R., Hall, A., Burbridge, D., & Starr, A. (2024). Smart railways: The design and construction of an autonomous inspection and maintenance vehicle. Procedia CIRP, 128, 61–65. https://doi.org/10.1016/j.procir.2024.03.003
Indonesian Society of Applied Science (ISAS)
