一种基于小波变换的遥感影像道路自动提取方法

遥感影像上道路的自动提取是摄影测量与遥感、计算机视觉等领域的重要课题。本文通过对图像小波系数的分析，完成道路节点的提取。随后利用基于小波变换的边缘提取算子，对道路进行边缘提取。边缘提取出来之后，以道路节点为种子点，对道路进行Snake跟踪。为保证提取效果，对图像进行了预处理。实验表明，利用本文提出的理论与方法能为道路网的自动提取与识别提供一个可靠的依据，同时也对其他地物（如建筑物）的自动提取提供一个有价值的参考。

Automatic Road Extraction from Remote Sensing Imagery Based on Wavelet Transform

Automatic road extraction from remote sensing imagery is a very important technology for construction of digital cities and inspection of road networks. Based on the traditional one, a method for automatic road extraction from remote sensing images is presented, which consists of three pruccssing levels. Firstly, node points are extracted by utilizing the wavelet transform. Secondly, an edge extraction algorithm based on wavelet is used. At last, we use a Snake tracing approach to trace roads and optimize the delineation process. We also do some image pretreatment to make the road extraction work well. Experiments show the method can effectively extract roads. Finally, the research issues are presented for further studies.