Using the Kalman Filter for Auto Bit-rate H.264 Streaming Based on Human Interaction

Human Machine Interface (HMI) and interactive systems applications are complex and diversified but human machine interaction in networks is even more complex. To design an interactive system, the designers must consider factors including design, technologies, people, activities and contexts. As the intermediary for human and machines, the interactive system takes charge of not only data transmission, display, storage and convertion, but also reactions to human behaviors. Based on human interaction, this paper focuses on the popular video streaming. As one of the most commonly used compression video formats, H.264 provides better quality at lower bit-rates than its previous standards in transmitting video/audio data. However, the quality of networked multimedia streaming easily fluctuates with the bit-rate variation. In order to maintain good video quality, this paper proposes to use the Kalman filter to implement the Auto bit-rate technique, which can regulate the bit-rate of the video/audio data automatically when the bit-rate is insufficient, and simultaneously guarantee the video quality. The experimental result proves that our proposed Auto bit-rate scheme can regulate the bit-rate to achieve the optimal visual quality and offer the best quality of service at the same time.