Rate control based on human visual perception characteristics and its application in encoding time-lapse video

To prevent the loss of the information embedded in the generated variable bit-rate data that is transmitted over a constant bit-rate channel, several methods were proposed in MPEG TMN5, H.263+ TMN5, and TMN8. In these methods, the quantity of coded data is controlled by adjusting the coding parameters according to the amount of data remaining in the buffer. Because this control is based on the past coded information and does not reflect the nature of the image being coded, there is no assurance that sufficient image quality will be obtained. This dissertation presents new rate control schemes to control the encoder generated Variable Bit-rate (VBR) data over a Constant Bit-rate (CBR) channel for both real-time and non real-time streaming video applications. The proposed rate control schemes integrate frame and block layer rate control to regulate the bit-rate using the combination of feed-forward control, feed-backward control, and model-based approaches together with characteristics of human visual perception, Smooth Pursuit Eye Movement (SPEM). As a result, the proposed rate control methods encode the video sequence with higher encoded video quality compared to UBC (University of British Columbia) H.263+ codec.; In this dissertation, we also investigate the use of our proposed Scene Context Dependent Key Frame Selection (SCDKFS) Scheme in combination with the proposed rate control to reduce the amount of video data recording and to control the rate of the video data generated by the encoder while producing the best streaming time-lapse video quality. We will show that our proposed SCDKFS and Rate Control (SCDKFS&RC) scheme is very effective in delivering the best streaming time-lapse video quality to the end users while reducing the amount of recorded video data as compared to the traditional time-lapse and the existing UBC H.263+ offline rate control.