Precision in Pixel Prediction: Video Interpolation through Polynomial Fitting

This paper presents an innovative method for video interpolation, focusing on enhancing temporal continuity in low frame-rate videos. The approach combines the power of the Polynomial Regression / Gaussian elimination with polynomial fitting to predict pixel values in intermediate frames, drawing data from k previous and k upcoming frames. Commencing with the challenge of improving video smoothness, the methodology involves transforming pixel data into the function domain, enabling the creation of polynomial representations that capture temporal dynamics accurately. Findings from this research underscore the effectiveness of the proposed technique in achieving higher frame rates while maintaining video quality. This method holds promise for addressing real-world video processing challenges, such as compression artifacts and noise. In conclusion, this abstract encapsulates the core elements of the paper, including its central idea, research method, key findings, and implications. The approach outlined herein offers a valuable contribution to video interpolation, with potential applications in multimedia, digital content creation, and video enhancement.