Research on Eye Tracking Process Optimization Based on Combined Kalman Filtering

The application of eye tracking is becoming increasingly widespread and significant in daily life. However, due to physiological and technological limitations, eye tracking often contains noise and inaccuracies, which poses a challenge to its efficient use. This study aims to enhance the accuracy and precision of eye tracking, thereby improving the performance of human-computer interaction systems. We explore the application effects of various algorithms on eye-tracking data processing, including the Simple Moving Average, Weighted Moving Average, Exponential Weighted Moving Average, Kalman Filter Algorithm, and Combined Kalman Filter Algorithm. The advantages and disadvantages of these five algorithms are analyzed using eye-tracking data evaluation metrics focused on accuracy and precision. The findings of this study have practical application value in fields requiring high-precision eye movement data.