Aiming at the problem of the influence of abnormal data on the test accuracy during the SINS measurement process, this paper proposes a high-precision inertial navigation method based on morphological filtering. The system firstly uses the proposed triangular morphology filtering method to correct the measurement data of the inertial measurement unit (IMU), reduce the outliers in the measurement data, and then use the astigmatism calibration method based on the odometer calibration to predict the morphological filtering. The processed data is calculated to achieve precise positioning of the measured pipeline path. In order to understand the inertial navigation positioning effect after triangular morphology filtering, a simulation comparison test was carried out on the 104 m pipeline. The analysis of the test results shows that the measurement accuracy of the horizontal direction of the system is increased from 0.58% to 0.29%, which is doubled by the morphological filtering of the triangle. The measurement accuracy in the height direction is increased from 0.096% to 0.077%, an increase of 24.6%, achieved good results.