基于YOLO算法的自动驾驶汽车检测研究综述

Research Review of Autonomous Vehicle Detection Based on YOLO Algorithm

摘要: 目标检测是自动驾驶汽车环境感知的重要内容。YOLO系列算法在检测性能领域表现突出，对目标检测的研究有重要意义。YOLO算法能够实时监测自动驾驶车辆中的目标，包括车辆、行人、交通标志、灯光和车道线等。同时，自动驾驶汽车的发展对于提高交通安全、节能减排以及减少交通事故有着重要意义。在自动驾驶中，目标检测是一项基础且关键的技术，需要实时准确地检测和识别道路上的各类目标。本文首先介绍了目标检测中常用的评价指标；其次，总结了单阶段和双阶段目标检测算法的思想及其优缺点；综述了单阶段目标检测算法-YOLO算法在自动驾驶汽车检测领域的应用，从交通标志、交通灯、行人识别和交通车辆四个方面分开阐述和总结研究现状以及应用情况；最后展望了现阶段目标检测存在的问题和未来发展方向，以及YOLO算法可以在自动驾驶汽车检测方面做出哪些更具有挑战性的研究。

Abstract: Object detection is an important part of the environmental perception of autonomous vehicles. The YOLO series algorithms have outstanding performance in the field of detection performance and are of great significance to the research of object detection. The YOLO algorithm is able to monitor targets in autonomous vehicles in real time, including vehicles, pedestrians, traffic signs, lights, and lane lines. At the same time, the development of autonomous vehicles is of great significance to improve traffic safety, save energy and reduce emissions, and reduce traffic accidents. In autonomous driving, object detection is a fundamental and critical technology that requires real-time accurate detection and identification of various types of objects on the road. This paper first introduces the commonly used evaluation indicators in object detection. Secondly, the ideas of single-stage and two-stage object detection algorithms and their advantages and disadvantages are summarized. The application of single-stage object detection algorithm-YOLO algorithm in the field of autono-mous vehicle detection is reviewed, and the research status and application status are summarized from four aspects: traffic signs, traffic lights, pedestrian recognition and traffic vehicles. Finally, the current problems and future development directions of object detection are prospected, and the more challenging research that YOLO algorithm can make in autonomous vehicle detection.