Design and Implementation of a Real-Time Object Detection System using YOLO and Deep Learning

The current research has achieved designing, implementing and testing a real-time object detection system by leveraging YOLO algorithm and deep learning techniques. Main goal in achieving faster and more accurate detection for images and real-time video streams has been successfully reached, it is able to detect multiple objects simultaneously while keeping a good trade-off between accuracy and processing speed suitable for real-time implementation.  The use of deep learning based backend and an interactive frontend, allow the system to display detections visually to the user and therefore increases its usability for real-time tasks. Detection results based on standard metrics (precision, recall, mAP, FPS) confirm that the model behaves correctly under several conditions and prove that YOLO-based approach shows a clear speed improvement in comparison with previous region-based methods. There are few limitations in the developed system such as: loss of accuracy in small and partially occluded objects and dependence on hardware. However, the results and behavior are suitable for most of the real-world applications such as surveillance, traffic management, and intelligent environments. In conclusion, this work shows a promising future of YOLO-based models for the implementation of efficient real-time object detection systems. The obtained results serve as a good base for further research and improvements in the field of computer vision.

[1]     J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, Real-Time Object Detection,” Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR), 2016.   [2] J. Redmon and A. Farhadi, “YOLO9000: Better, Faster, Stronger,” Proc. IEEE CVPR, 2017. [3]     S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards Real-Time Object Detection,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017.   [4]     J. Redmon and A. Farhadi, “YOLOv3: An Incremental Improvement,” 2018. [5]     A. Bochkovskiy, C.-Y. Wang, and H.-Y. M. Liao,“YOLOv4: Optimal Speed and Accuracy of Object Detection,”2020.   [6]     C.-Y. Wang et al.,“YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art,” 2022. [7]     W. Liu et al.,“SSD: Single Shot MultiBox Detector,”Proc. European Conference on Computer Vision (ECCV), 2016.   [8]     A. Krizhevsky, I. Sutskever, and G. Hinton,“ImageNet Classification with Deep Convolutional Neural Networks,”Proc. NIPS, 2012.   9] K. He, X. Zhang, S. Ren, and J. Sun,“Deep Residual Learning for Image Recognition,”Proc. CVPR, 2016.   [10]                   T.-Y. Lin et al.,“Microsoft COCO: Common Objects in Context,”Proc. ECCV, 2014.   [11]                   R. Al Amin and R. Obermaisser,“Real-Time Object Detection and Classification using YOLO for Edge FPGAs,”Proc. c7th International Symposium ELMAR, 2025.