Research on infrared video stabilization based on SLIC and path optimization

Aiming at the technical difficulties of infrared video stabilization due to parallax changes, an infrared video stabilization algorithm based on SLIC and path optimization is proposed in the paper, which mainly includes: preprocessing, improving the accuracy of motion estimation to get the camera paths, multi-path optimization, and motion compensation. In this paper, we perform SLIC processing on infrared video, using RANSAC algorithm as well as improved k-means algorithm to process the feature points to get accurate global motion estimation, and after that, we optimize multiple camera paths to get stable infrared video after compensation. After experimental analysis, our method achieves better results compared to some other traditional methods.