中文摘要：

山区公路往往面临多种地质灾害的威胁，这些地质灾害往往对公路的通行能力、行车人的安全以及公路的维护造成严重的影响，甚至造成巨大的财产损失和人员伤亡。为了及时、科学地部署应急救灾工作，有效降低灾害影响，迫切需要对公路灾害进行快速调查为其提供辅助决策信息。遥感技术具有宏观、快捷、直观、动态的特点，适用于大范围灾害的动态监测和调查。通过分析公路遥感影像特征，研究提出了适用于不同数据获取情况的3种公路灾害检测方法，即基于灾害识别、基于图像灰度均值与基于公路断点识别的公路灾害检测方法。试验证明，这些方法切实可行，快速有效，为缺少灾前遥感影像数据情况下的公路灾害快速检测提供了技术支撑，对交通应急救援具有实际应用价值。

英文摘要：

Owing to their geographical locations, mountain highways are facing the threat of a variety of geological disasters, resulting in not only huge property losses and human casualties but also highway maintenance difficulties. To deploy emergency relief work scientifically and effectively in order to reduce the damage of disasters to a minimum, a quick survey on the highway disasters is needed. One technology, regarded as the most effective way in disaster investigation and monitoring, is remote sensing. Based on features of post-disaster remote sensing images, this paper introduced various methods of highway disaster detection depending on whether or not the vector data of road network pre-disaster is offered. The methods can be divided into three classes： methods based on disaster identification, average gray value of image and highway breakpoint identification. The test results with aerial images show that the methods are quick, efficient and accurate. All these methods provide a complete set of data processing and information extraction flow for rapid highway disaster detection in the absence of pre-disaster remote sensing images. The study can provide a technical support to the diverse needs of information extraction in response to serious emergencies and disasters.