Computer vision, particularly vehicle and pedestrian identification is
critical to the evolution of autonomous driving, artificial intelligence, and
video surveillance. Current traffic monitoring systems confront major
difficulty in recognizing small objects and pedestrians effectively in
real-time, posing a serious risk to public safety and contributing to traffic
inefficiency. Recognizing these difficulties, our project focuses on the
creation and validation of an advanced deep-learning framework capable of
processing complex visual input for precise, real-time recognition of cars and
people in a variety of environmental situations. On a dataset representing
complicated urban settings, we trained and evaluated different versions of the
YOLOv8 and RT-DETR models. The YOLOv8 Large version proved to be the most
effective, especially in pedestrian recognition, with great precision and
robustness. The results, which include Mean Average Precision and recall rates,
demonstrate the model's ability to dramatically improve traffic monitoring and
safety. This study makes an important addition to real-time, reliable detection
in computer vision, establishing new benchmarks for traffic management systems.

通过训练和评估 YOLOv8 和 RT-DETR 模型的不同版本，我们的研究项目旨在创建和验证一个先进的深度学习框架，能够处理复杂视觉输入，以实时识别各种环境中的汽车和行人。YOLOv8 Large 版本被证明是最有效的，在行人识别方面具有很高的准确性和鲁棒性。研究结果表明，该模型能够显著提高交通监控和安全性，成为计算机视觉领域实时可靠检测的重要贡献，并为交通管理系统建立了新的基准。

实时深度学习车辆和行人检测与分析

Real-Time Detection and Analysis of Vehicles and Pedestrians using Deep  Learning

Overtaking is a critical maneuver in driving that requires accurate
information about the location and distance of other vehicles on the road. This
study suggests a real-time overtaking assistance system that uses a combination
of the You Only Look Once (YOLO) object detection algorithm and stereo vision
techniques to accurately identify and locate vehicles in front of the driver,
and estimate their distance. The system then signals the vehicles behind the
driver using colored lights to inform them of the safe overtaking distance. The
proposed system has been implemented using Stereo vision for distance analysis
and You Only Look Once (YOLO) for object identification. The results
demonstrate its effectiveness in providing vehicle type and the distance
between the camera module and the vehicle accurately with an approximate error
of 4.107%. Our system has the potential to reduce the risk of accidents and
improve the safety of overtaking maneuvers, especially on busy highways and
roads.

本研究提出了一种可以实时辅助驾驶员超车的系统，结合 YOLO 算法和立体视觉技术精确识别和定位前方车辆并估算其距离，再使用彩色灯光通知后续车辆安全超车距离，实验结果表明该系统精度可达误差率 4.107％，具有减少事故风险，提高道路行车安全的潜力。