Joint base layer and enhancement layer quantizer adaptation in EDR video coding
An encoder receives one or more input pictures of enhanced dynamic range (EDR) to be encoded in a coded bit stream comprising a base layer and one or more enhancement layer. The encoder comprises a base layer quantizer (BLQ) and an enhancement layer quantizer (ELQ) and selects parameters of the BLQ and the ELQ by a joint BLQ-ELQ adaptation method which given a plurality of candidate sets of parameters for the BLQ, for each candidate set, computes a joint BLQ-ELQ distortion value based on a BLQ distortion function, an ELQ distortion function, and at least in part on the number of input pixels to be quantized by the ELQ. The encoder selects as the output BLQ parameter set the candidate set for which the computed joint BLQ-ELQ distortion value is the smallest. Example ELQ, BLQ, and joint BLQ-ELQ distortion functions are provided.
1. A method comprising:
receiving a plurality of input pixels to be encoded by an encoder, the encoder comprising a base layer quantizer (BLQ) and an enhancement layer quantizer (ELQ);
selecting a first clipping mode;
for the first clipping mode selecting a first plurality of candidate sets of parameters for the BLQ;
for each candidate set of BLQ parameters in the first plurality of candidate sets:
computing with the encoder a first BLQ distortion metric based on a first BLQ distortion function;
computing with the encoder a first ELQ distortion metric based on a first ELQ distortion function;
computing a first joint BLQ-ELQ distortion value based on the first BLQ distortion metric, the first ELQ distortion metric, and at least in part on the number of pixels to be quantized by the ELQ under the first clipping mode;
determining a first output BLQ parameter set among the first plurality of candidate sets, wherein the first output BLQ parameter set comprises the candidate set for which the computed first joint BLQ-ELQ distortion value is the smallest;
selecting a second clipping mode;
for the second clipping mode, selecting a second plurality of candidate sets of parameters for the BLQ;
for each candidate set of BLQ parameters in the second plurality of candidate sets:
computing with the encoder a second BLQ distortion metric based on a second BLQ distortion function;
computing with the encoder a second ELQ distortion metric based on a second ELQ distortion function;
computing a second joint BLQ-ELQ distortion value based on the second BLQ distortion metric, the second ELQ distortion metric, and at least in part on the number of pixels to be quantized by the ELQ under the second clipping mode;
determining a second output BLQ parameter set among the second plurality of candidate sets, wherein the second output BLQ parameter set comprises the candidate set for which the computed second joint BLQ-ELQ distortion value is the smallest;
computing a third joint output distortion value based on the first output BLQ parameter set and the first joint BLQ-ELQ distortion function;
computing a fourth joint output distortion value based on the second output BLQ parameter set and the second joint BLQ-ELQ distortion function; and
selecting the first output BLQ parameter set if the third joint output distortion value is smaller than the fourth joint distortion value, otherwise selecting the second output BLQ parameter set.
2. The method of claim 1 , wherein the first clipping mode is a low-clipping mode and the second clipping mode is a high-clipping mode.
3. The method of claim 1 , wherein the first ELQ distortion function and the second ELQ distortion function are different.
4. The method of claim 1 , wherein the BLQ is a linear quantizer and the candidate set of BLQ parameters in the first plurality of candidate sets comprises a fixed high-clipping output parameter (C H ) and a variable low-clipping output parameter (C L ).
5. The method of claim 4 , wherein the fixed high-clipping parameter is equal to 255.
6. The method of claim 4 , wherein the variable low-clipping output parameter is smaller than zero.
7. The method of claim 1 , wherein the BLQ is a linear quantizer and the candidate set of BLQ parameters in the second plurality of candidate sets comprises a variable high-clipping output parameter (C H ) and a fixed low-clipping output parameter (C L ).
8. The method of claim 7 , wherein the fixed low-clipping parameter is equal to zero.
9. The method of claim 7 , wherein the variable high-clipping output parameter is larger than 255.
10. A non-transitory computer-readable storage medium having stored thereon computer-executable instruction for executing a method with one or more processors in accordance with claim 1 .