视觉引导AGV中的数字图像处理方法

DIGITAL IMAGE PROCESSING METHODS IN VISUAL NAVIGATION AGV

利用计算机视觉技术引导自动运输车AGV(Automatic Guided Vehicle)是一种重要的AGV导航手段。该方法通过摄像头获取道路图像信息并进行处理,获得道路参数,以此来指引AGV导航。现有的视觉导航算法大多复杂度较大,难以满足实际应用的实时性需求。提出一种基于道路参数识别的视觉引导方法,将道路中的道路识别为曲线参数,结合预测进行修正,在保证精度的前提下,大幅提高算法效率。理论分析和实验结果表明,该方法具有识别精度高,执行速度快,能达到实用化的要求。

It is an important AGV approach to guide the AGV with computer visual technology.The method obtains road image information through cameras and processes it to get road parameters to guide AGV navigations.Most existing visual navigation algorithms are too complicated to satisfy real time demands in actual applications.The paper proposes a visual navigation approach based on road parameter identification,which identifies roads as curving parameters to be modified together with the combination of predictions.Under the premise of promising its precision,the efficiency of the algorithm is significantly promoted.Theoretical analysis and experimental results demonstrate that the approach bears such characteristics as high identification resolution,speedy execution and practicality.