Repainting road markings is an important road maintenance task. An automatic repainting module (called "MRA") is being developed in France by the Laboratoire Central des Ponts-et-Chaussees. In this context, our goal is to design image processing methods for automatically extracting the parameters of the repainting process (e.g. line width or dashes modulation) from the scene images. We describe a three-level road marking recognition algorithm: 1) a road sides extraction procedure reduces the analysis area to the roadway and a road markings detection is performed in this area; 2) the segmentation is simplified using a projection-based method, leading to a polygonal representation of road markings; and 3) the parameters to be transmitted to the repainting system are extracted.
展开▼
机译：重新绘制的道路标记是一个重要的道路维护任务。在法国的法国由Laboratoire Central des Ponts-et-et-et-et-et-et-edssees开发了自动重新粉刷模块（称为“MRA”）。在这种情况下，我们的目标是设计用于自动提取从场景图像中自动提取重绘过程的参数（例如线宽或破折号调制）的图像处理方法。我们描述了一种三级道路标记识别算法：1）路边提取程序将分析区域减少到道路，在该区域进行道路标记检测; 2）使用基于投影的方法简化分割，导致道路标记的多边形表示; 3）提取要传输到重新绘制系统的参数。
展开▼
