Abstract: In view of the characteristics of the vehicle trajectory data, which has the same variation trend, this paper proposed a feature extraction algorithm in consideration of the trend changing of track. Firstly, the trend set to confirm and the trend set to pend are set up, and the state of the trend set could be determined by calculating the trend of the trajectory point. Further, the effective trend would be obtained by judging the states of the 2 trends. Finally, the constraint conditions of the feature of the trend are applied to finish the feature extraction of the trend. Experimental results show that the proposed algorithm is superior to the traditional method of curve fitting in efficiency. The interference features due to special circumstances are ruled out effectively. The proposed algorithm is not only simple and feasible, but also has high efficiency, meeting the application requirements of integrated navigation.