Project Abstract

This research focuses on the effectiveness of the YOLO algorithm in small target detection and its improvement methods to address the important challenges currently facing real-time target detection systems. As various applications place higher demands on accurate and rapid target detection, especially in areas such as autonomous driving and urban surveillance, it is particularly urgent to investigate how to improve the ability to detect small targets.The proposition of this study is to enhance the detection accuracy of YOLO model for small targets in complex environments through structural optimisation and algorithmic improvement. The main methods used include introducing multi-scale training techniques, optimising the loss function of the model, and applying data enhancement techniques to improve the model’s recognition and generalisation ability for small targets.The main conclusions of the study show that these improvements significantly improve the performance of the YOLO algorithm in small target detection, especially in the detection accuracy when the target size is small or in complex backgrounds. These results not only validate the feasibility of the improved YOLO algorithm in practical applications, but also provide new ideas for the advancement of real-time target detection technology.Through this study, the world now understands that the YOLO algorithm is still able to perform the target detection task effectively even in the face of the challenges of small size and complex environments through appropriate technical adjustments and algorithmic optimisation. This discovery significantly advances the technological progress in the field of real-time video analytics and automated surveillance, and lays the foundation for future development of related technologies.

Keywords: computer vision, target detection, YOLO

 Conference Details

Session: Poster Session B at Poster Stand 130

Location: Sir Stanley Clarke Auditorium at Wednesday 8th 09:00 – 12:30

Markers: Yuanbo Wu, Jay Morgan

Course: BSc Computer Science, 3rd Year

Future Plans: I’m continuing studies