Abstract: A novel algorithm is proposed，which may recognize and extract efficaciously the road network from paper-based color highway traffic map images.This algorithm includes three key steps.Firstly，according to the features of the non-road regions，the luminance of the non-road regions is extracted.Secondly，based on the color and shape features of roads，the colors of the road network are recognized and extracted，which taking advantage of some digital image processing technologies such as connection component.Finally，some methods are put forward to connect some road segments in order that the whole road networks are obtained.This algorithm has been applied to many highway traffic map images.It is shown that the algorithm is effective in various successful cases.

摘要： 根据彩色公路交通地图的图像特征，提出一种新颖的道路识别与提取方法。这种方法包括三个关键步骤。首先，根据区域的特征，提取出区域的灰度值；其次，根据道路的颜色和形状特征以及数字图像处理的一些方法（如对象的连通成分等），识别并提取出道路的颜色；最后，为了获得完整的道路网络，一些道路连接方法被提出。这种算法已经被应用于许多彩色公路地图图像中去提取道路网络。大量成功的实例表明这个算法是非常有效的。