Abstract: Visual SLAM is a widely used navigation technology, but its performance in practical applications is often constrained by multiple factors. Among them, image jitter is one of the main bottlenecks. To address this issue, this paper proposes a jitter suppression optimization algorithm that combines ORB feature point tracking with Kalman filtering. The method calculates the jitter amplitude by block-based feature point matching, then applies Kalman filtering to fit the amplitude and estimate the compensation value, which is used to correct the jitter. In this way, high-precision SLAM mapping can be achieved under jitter interference. In experiments, the proposed algorithm was compared with the original ORB-SLAM algorithm. Results show that under small-amplitude jitter, the improved algorithm reduces the average error and root mean square error (RMSE) by up to 7.68% and 11.19%, respectively, compared with the original algorithm. Under large-amplitude jitter, the reductions reach up to 74.89% and 56.93%, respectively. These findings demonstrate that the proposed algorithm achieves robust jitter suppression under varying conditions and significantly improves the mapping accuracy of visual SLAM.