A high performance VBR coding algorithm for fixed size storage applications

We propose a novel variable bit rate (VBR) algorithm that adopts two-pass VBR encoding to implement high performance coding for fixed size storage applications. In the first pass encoding, we use the original frame as the reference frame to compute motion estimation so that the exact relationship between the amount of bits generated for each frame (Rate) and each possible quantization factor (Q) can be obtained. With this R-Q function, we are able to optimize the quantization factor throughout an entire sequence according to the constraints of the storage size and the decoder buffer size before the second pass encoding is executed. In addition, the decoder is able to prevent the buffer from overflowing by pausing the operation of reading the bit stream into the buffer at some specific time. In the second pass encoding, the selected quantization factors for all the frames will be used to encode the entire video sequence as a free VBR encoder. Experimental results show that the proposed VBR encoding can provide more consistent visual quality as well as significantly improved coding efficiency