Aiming at the problems that the front-end drift error of the V-SLAM algorithm is accumulated and it is sensitive to changes in background brightness, a monocular visual odometry method based on the simulation of target characteristics is proposed. This method uses the dense 3D point cloud data to perform Poisson reconstruction of the target surface and the correlation of the physical properties of the material, and then builds a physical model-based imaging rendering engine (PBRT) according to the imaging principle of the vision sensor to generate simulated images of the target characteristics under different observation conditions. The simulation image of the target characteristics and the image captured by the optical camera are registered and the motion deviation is recovered, and the Extended Kalman Filter (EKF) is designed to output the optimal estimation value of the pose state. The prototype system development and experimental evaluation show that this method effectively overcomes the problems of accumulation of drift errors and sensitivity to background changes in the traditional method. Compared with the traditional ORB-SLAM2 algorithm, the front-end positioning accuracy is significantly improved, which provides a new method for the design of visual odometry. technical ideas.