论文部分内容阅读
道路信息是一种重要的基础地理信息,在城市规划、GIS更新等方面都有很重要的应用,从光学遥感图像中提取道路信息可以方便的获得并更新交通数据。文章提出了一种基于均值漂移的高分辨率遥感影像道路提取方法。首先利用均值漂移法对图像进行初步分割;然后以对分割后的图像利用道路局部灰度值相近的特点进行标记,通过这个步骤会将灰度较均匀的道路提取出来,但是与道路有共同灰度一致性特点的建筑物、停车场等也被提取出来;接下来根据道路一般面积不会太小且呈长条状的特点剔除非道路信息达到进一步的提纯;最后利用区域增长法对断裂的道路段进行连接形成道路网。根据实验结果验证该方法能较好的提取出道路信息。
Road information is an important basic geographic information, which has important applications in urban planning and GIS updating. It can easily obtain and update traffic data by extracting road information from optical remote sensing images. This paper presents a method of high resolution remote sensing image road extraction based on mean shift. Firstly, the image is preliminarily segmented by means of mean shift method. Then the local gray value of the road is marked by using the local grayscale value of the segmented image. Through this step, the road with more uniform gray level is extracted, but the road is same as gray The buildings with the same degree of consistency and the parking lot are also extracted. Next, the non-road information is removed according to the characteristic that the general area of the road is not too small and strip-like, so as to further purify the non-road information. Finally, Road sections are connected to form a road network. According to the experimental results, it is verified that this method can extract the road information well.