Accelerated Extraction Technology Research on Damaged Building Information by Earthquake Based on LiDAR Image

To obtain the damaged building information rapidly and accurately can provide support for the disaster-relief work after the earthquake, and it is also an important part of the evaluation of the earthquake disaster losses. The high-resolution remote sensing image and the satellite radar data are important technical means for the disaster monitoring, but the accuracy of the automatic extraction of information is subject to certain restrictions. However in recent years, the newly-presented LiDAR technology can provide the elevation information for ground targets, which can be applied to rapidly and accurately extract the earthquake disaster information under the circumstances that the remote-sensing image is absent before the earthquake. Adopting the LiDAR data and remote sensing data in the Yushu disaster area, the damaged building information after the earthquake is automatically extracted through the Arcmap collection category samples and by using the object-oriented SVM classification, the overall accuracy can reach 80.96% in the research.