Velocity-based object detection in dynamic environment using YOLO-based deep learning algorithm
AUTHORS
Ihn-Sik Weon,Dept of Mechanical Engineering, Graduated School, Kyung Hee University, Yongin-si, Korea
Soon-Geul Lee*,Dept of Mechanical Engineering, Kyung Hee University, Yongin-si, Korea
ABSTRACT
In this study, to solve the constraints of an image sensor and resolve an obstacle detection error according to the traveling speed of an autonomous vehicle, we applied the object recognition technology of the single-shot detector technique based on the you only look once (YOLO) algorithm to pedestrians, bicycles, traffic lights, and pedestrian crossings for effective obstacle avoidance and perception. The proposed technique was experimentally proven on campus at Kyung Hee University, with the results confirming the accuracy of object recognition using a number of learning datasets.
KEYWORDS
Single-shot detector(SSD), Object detection, Deep-learning, Vision System, FMCW Radar
REFERENCES
[1] G. H. Anh.; S. W. Lee.; Technology Trends of Self-Driving Vehicle. Report of ETRI, (2013), 135(4), 35-54.
[2] J. M. Kim.; J. M Heo.; S. Y. Jung.; S. S. Kim.; Path-Planning Using Modified Genetic Algorithm and SLAM Based on Feature Map for Autonomous Vehicle. Journal of Korean Institute of Intelligent Systems, (2009), 19(3), 381-387.
[3] D. J. Yoon.; J. H. Kim.; J. H. Kim.; LiDAR Point Cloud Data Clustering and Classification for Obstacle Recognition of UGV. Conference of Korean Society of Automotive Engineers, (2012), 1339-1343.
[4] A. Ahmad.; L. Dey.; A K-Means Clustering Algorithm for Mixed Numeric and Categorical Data. Data & Knowledge Engineering, (2007), 63(2), 503-527.
[5] S. H. Yang.; B. S Song.; J. Y. Um.; Radar and Vision Sensor Fusion for Primary Vehicle Detection. Journal of Institute of Control, Robotics and Systems, (2010), 16(7), 639-649.
[6] O. Chavez-Garcia.; O. Aycard.; Multiple Sensor Fusion and Classification for Moving Object Detection and Tracking. IEEE Transactions on Intelligent Transportation Systems, (2016), 17(2), 525-534.
[7] I. S. Weon.; J. K. Ryu.; S. G. Lee.; Recognition of Object Tracking with Radar for USV. Conference of Korean Society for Precision Engineering, (2017), 299-300.
[8] I. S. Weon.; J. K. Ryu.; S. G. Lee.; Obstacle Avoidance of Unmanned Surface Vehicle Based on 3D Lidar for VFH Algorithm. Asia-Pacific Journal of Multimedia Service Convergent with Art, Humanities, and Sociology, (2018), 8(3), 945-953.
[9] I. S. Weon.; J. K. Ryu.; S. G. Lee.; Virtual Bubble Filtering Based on Heading Angle and Velocity for Unmanned Surface Vehicle (USV). International Conference on Control, Automation, and Systems (ICCAS), (2017), 1954-1958.
[10] J. Redmon.; You Only Look Once: Unified, Real-Time Object Detection. IEEE Conference on Computer Vision and Pattern Recognition, (2016).