Search in Medwell
 
 
Journal of Engineering and Applied Sciences
Year: 2019 | Volume: 14 | Issue: 1 SI | Page No.: 3946-3949
DOI: 10.36478/jeasci.2019.3946.3949  
Human Detection using Improved YOLOv2: Images Captured by the UAV
Juwon Kwon and Soonchul Kwon
 
Abstract: In recent years, the technology of Unmanned Aerial Vehicles (UAV) has developed. The UAV which were initially used for military purposes have recently begun to be commercialized. The market for the UAV has grown and small UAV have been created. Many people can enjoy a variety of hobbies such as taking pictures using the UAV. Therefore, the technology utilizing the images captured by the UAV is also developing. Existing human detection algorithms have resulted in unsatisfactory results in images taken by the UAV. There is a problem that distortions that do not occur in ordinary images occur due to the difference of angle of view. When the distance between the UAV and the human is long, there is a problem that the human is small and has a low resolution. In this study, we improve the YOLOv2 algorithm which shows good performance for object detection, to improve human detection performance in images taken by the UAV.
 
How to cite this article:
Juwon Kwon and Soonchul Kwon, 2019. Human Detection using Improved YOLOv2: Images Captured by the UAV. Journal of Engineering and Applied Sciences, 14: 3946-3949.
DOI: 10.36478/jeasci.2019.3946.3949
URL: http://medwelljournals.com/abstract/?doi=jeasci.2019.3946.3949