Coding unit quantization parameters in video coding
A method is provided that includes receiving a coded largest coding unit in a video decoder, wherein the coded largest coding unit includes a coded coding unit structure and a plurality of coded quantization parameters, and decoding the coded largest coding unit based on the coded coding unit structure and the plurality of coded quantization parameters.
1. A method of video processing, comprising:
dividing a picture into a plurality of non-over-lapping blocks using a recursive quad-tree structure;
determining a minimum coding unit size for the plurality of non-over-lapping blocks of a first size for which a first quantization parameter will be determined wherein the minimum coding unit size is less than the first size;
transforming the plurality of non-over-lapping blocks into a plurality of transformed coefficients in a frequency domain using a transform function;
quantizing the plurality of transformed coefficients using a plurality of quantization parameters at least one of which is the first quantization parameter; and
encoding the plurality of quantized transformed coefficients into a compressed bit stream and signaling at a picture level in the compressed bit stream the minimum coding unit size for which the first quantization parameter is determined for the first non-over-lapping block.
2. The method of claim 1 further comprising transmitting the compressed bit stream on a communications channel.
3. The method of claim 1 further comprising:
performing an inverse quantization operation on the plurality of quantized transformed coefficients using the plurality of quantization parameters to form a plurality of reconstructed transformed coefficients; and
performing an inverse transform operation on the plurality of reconstructed transformed coefficients to form a plurality of reconstructed non-over-lapping blocks.
4. The method of claim 3 wherein the picture obtained using a digital video camera on a mobile cellular telephone.
5. The method claim 3 wherein the picture is obtained using a digital video camera.
6. The method of claim 3 wherein the picture is obtained from a video content provider.
7. A method of video processing, comprising:
obtaining a picture using a digital camera on a mobile cellular telephone;
processing the picture on the mobile cellular telephone using the steps of:
dividing the picture into a plurality of non-over-lapping blocks using a recursive quad-tree structure;
determining a minimum coding unit size for the plurality of non-over-lapping blocks of a first size for which a first quantization parameter will be determined wherein the minimum coding unit size is less than the first size;
transforming the plurality of non-over-lapping blocks into a plurality of transformed coefficients in a frequency domain using a transform function;
quantizing the plurality of transformed coefficients using a plurality of quantization parameters at least one of which is the first quantization parameter; and
encoding the plurality of quantized transformed coefficients into a compressed bit stream and signaling at a picture level in the compressed bit stream the minimum coding unit size for which the first quantization parameter is determined for the first non-over-lapping block.
8. The method of claim 7 further comprising:
performing an inverse quantization operation on the plurality of quantized transformed coefficients using the plurality of quantization parameters to form a plurality of reconstructed transformed coefficients; and
performing an inverse transform operation on the plurality of reconstructed transformed coefficients to form a plurality of reconstructed non-over-lapping blocks.