Every year people in Taiwan have been facing earthquake and typhoon nature disaster challenges which are seriously affecting the people's lives and property. During summer season typhoons bring abundant rainfall resulting in landslides and debris flows. Because the landslide attributes to several environmental factors, we could monitor and analyze endangered areas to prevent damages. In this study Taiwan Provincial Highway 14 branch at 34K~38K+500m is chosen, which has devastated by Earthquake 921 and Typhoon Mindulle. A decision analysis is conducted by utilizing combined ortho-photo and LiDAR Digital Elevation Model (DEM) data to classify landslide areas. The result showed that the performance of overall accuracy could be increased when terrain factors are considered.