Abstract: The rapid development of mobile internet technology makes it possible to collect GPS data when people are travelling.To extract trajectory information based on the GPS data quickly and accurately is very important for road planning and route selecting in scenic areas.Based on analysis on GPS data of hiking,a method to eliminate noise and redundancy is proposed.Based on it,a algorithm for mining frequent sequences to identify key paths and POI (Point of Interest) is proposed.A mining algorithm of road network based on Density Based Spatial Clustering of Application with Noise (DBSCAN) is then developed.The results show that the proposed methods improve efficiency and accuracy of the mining algorithm on road network.They can accurately generate the road network in scenic areas,and be able to achieve new traveling paths.