图基导航的小型道路标线划线车无人驾驶系统设计与实现
Design and Implementation of a Driverless Road Marking System for a Small-Scale Vehicle Based on Vision-Guided Navigation

道路标线划线车是标记或划定道路边界、车道、停车线等交通标线的专业车辆，其作用是施划道路标线，确保道路标线整齐清晰，从而提高道路安全性和车辆行驶的有序性。针对使用传统手推划线车存在施工效率和施工精度偏低、人眼对齐施工道路引导水线(以下简称引导水线)难的问题，设计了小型划线车无人驾驶改造方案。该系统首先采用图像处理技术改进Haar-like模版进行引导水线特征识别；其次，结合GNSS定位信息将图像中引导水线特征中点转换到全局坐标系下，并根据图像所得到的全局定位信息进行模型预测控制器(MPC)轨迹跟踪设计；最后通过实车实验进行验证，本文所设计的小型无人驾驶划线车改造方案能够清晰识别引导水线特征，无人驾驶跟踪精度满足道路标线工艺要求。
The road marking vehicle is a specialized apparatus for marking or delineating road boundaries, lanes, and parking lines, with the primary function of applying road markings to ensure clarity and neatness, thereby enhancing road safety and facilitating orderly vehicle movement. To address the issues of low construction efficiency and precision, as well as the difficulty for manual alignment of the guide waterline (hereafter referred to as the guide waterline) during traditional hand-push road marking operations, a solution for the driverless transformation of a small road marking vehicle has been designed. This system first utilizes image processing techniques to improve the Haar-like template for guide waterline feature recognition. Next, it integrates GNSS positioning information to transform the midpoint of the guide waterline features from the image into the global coordinate system. Based on the globally positioned information derived from the image, a model predictive controller (MPC) is designed for trajectory tracking. Finally, through real vehicle experiments, it is demonstrated that the intelligent modification solution for the small autonomous road marking vehicle can accurately recognize guide waterline features, and its unmanned tracking precision meets the requirements of road marking processes.