An adaptive quantization algorithm for video coding

This paper proposes an adaptive quantization algorithm for video coding usi ng the information obtained from the previously encoded image, Before quant izing the discrete cosine transform coefficients, the properties of reconst ruction error of each macro block (MB) are estimated from the previous fram e. For the estimation of the error of current MB, a block with the size of MB in the previous frame is chosen. Since the original and reconstructed im ages of the previous frame are available in the encoder, me can evaluate th e tendency of reconstruction error of this block in advance. Then, this err or is considered as the expected error of the current MB if it is quantized with the same step size and bit rate. Comparing the error of the MB with t he average of overall MB's, if it is larger than the average, a small step size is given for this MB, and vice versa, As a result, the error distribut ion of the MB is more concentrated to the average, yielding low variance an d improved image quality, Especially for low bit application, the proposed algorithm yields much smaller error variance and higher peak signal-to-nois e ratio compared to the conventional TM5. We also propose a modified algori thm for efficient hardware implementation.