Methods for palette size signaling and conditional palette escape flag signaling

1. A method of palette coding for a picture, wherein the picture is divided into multiple coding units (CUs), the method comprising:

receiving input data associated with a current coding unit (CU);

determining an escape indication flag regarding whether there is any sample in the current CU coded as an escape pixel, wherein the escape indication flag is asserted if there is any sample in the current CU coded as the escape pixel and the escape indication flag is not asserted if there is no sample in the current CU coded as the escape pixel; and

signaling size information related to palette size of a current palette table of the current CU conditionally based on the escape indication flag, wherein determining the escape indication flag is performed no later than determining the size information,

wherein if the escape indication flag is not asserted and predicted palette size is zero, then a size syntax representing new palette size minus one is signaled for the current palette table, wherein the predicted palette size indicates a first number of entries of the current palette table that are predicted by a palette predictor and the new palette size indicates a second number of entries of the current palette table that are not predicted by the palette predictor, and

wherein an actual new palette size corresponding to the second number of entries of the current palette table that are not predicted by the palette predictor is derived based on the size syntax parsed plus (numPredPreviousPalette ==0 && palette_escape_val_present_flag ==0), wherein numPredPreviousPalette corresponds to a first syntax representing the predicted palette size and palette_escape_val_present_flag corresponds to a second syntax representing the escape indication flag.

2. The method of claim 1 , wherein if the escape indication flag is not asserted, a syntax palette_size_minus_one signaled for the current palette table of the current CU is limited to a range from zero to a maximum palette size minus one inclusive, wherein the syntax palette_size_minus_one indicates a current palette size of the current palette table minus one.

3. The method of claim 2 , wherein the current palette size is reconstructed according to (palette_size_minus_one +one).

4. The method of claim 1 , wherein if the escape indication flag is not asserted and predicted palette size is zero, then new palette size related to the current palette table is restricted to be greater than zero, wherein the predicted palette size indicates a first number of entries of the current palette table that are predicted by a palette predictor and the new palette size indicates a second number of entries of the current palette table that are not predicted by the palette predictor.

5. The method of Claim 1 , wherein an actual new palette size corresponding to the second number of entries of the current palette table that are not predicted by the palette predictor is determined according to the size syntax parsed from a bitstream plus one, wherein the bitstream includes coded data of the current CU.

6. The method of claim 1 , wherein a size syntax representing new palette size minus one is signaled after the escape indication flag in a bitstream that includes coded data of the current CU.

7. The method of claim 1 , wherein a syntax numPredPreviousPalette representing a predicted palette size is signaled after the escape indication flag in a bitstream that includes coded data of the current CU, wherein the predicted palette size indicates a first number of entries of the current palette table that are predicted by a palette predictor.

8. The method of claim 1 , wherein the palette size of the current palette table of the current CU is directly restricted to be greater than one by an encoder.

9. The method of claim 8 , wherein if a bitstream that results in a decoded palette size of the current palette table being zero is determined as a non-conforming bitstream.

10. A method of palette coding for a picture, wherein the picture is divided into multiple coding units (CUs), the method comprising:

receiving input data associated with a current coding unit (CU);

determining current palette size of a current palette table of the current CU, or both predicted palette size and new palette size associated with the current palette table, wherein the predicted palette size indicates a first number of entries of the current palette table that are predicted by a palette predictor and the new palette size indicates a second number of entries of the current palette table that are not predicted by the palette predictor; and

performing one of:

determining whether or not to signal a CU-level escape pixel presence flag according to the current palette size or both of the predicted palette size and the new palette size and signaling a CU-level escape pixel presence flag according to the determination of whether or not to signal the CU-level escape pixel presence flag, wherein the CU-level escape pixel presence flag indicates whether there is any sample in the current CU coded as an escape pixel; and

determining whether or not to constrain the CU-level escape pixel presence flag to one according to the current palette size or both of the predicted palette size and the new palette size, and constrain the CU-level escape pixel presence flag to one according to the determination of whether or not to constrain the CU-level escape pixel presence flag to one.

11. The method of claim 10 , wherein the CU-level escape pixel presence flag is not signaled if the current palette size is zero or both the predicted palette size and the new palette size are zero.

12. The method of claim 11 , wherein the CU-level escape pixel presence flag is inferred as one.

13. The method of claim 10 , wherein the CU-level escape pixel presence flag is constrained to one if the current palette size is zero or both the predicted palette size and the new palette size are zero.

14. The method of claim 10 , wherein the current palette size of the current palette table is set to zero if both the predicted palette size and the new palette size are zero.

15. The method of claim 10 , wherein the current palette size of the current palette table relates to a previous palette table of a last palette-coded CU.

16. A video coding system implementing palette coding for a picture, wherein the picture is divided into multiple coding units (CUs), the system comprising:

a receiver configured to receive input data associated with a current coding unit (CU);

a processor configured to determine current palette size of a current palette table of the current CU, or both predicted palette size and new palette size associated with the current palette table, wherein the predicted palette size indicates a first number of entries of the current palette table that are predicted by a palette predictor and the new palette size indicates a second number of entries of the current palette table that are not predicted by the palette predictor; and

the processor further configured to perform one of:

determining whether or not to signal a CU-level escape pixel presence flag according to the current palette size or both of the predicted palette size and the new palette size and signaling a CU-level escape pixel presence flag according to the determination of whether or not to signal the CU-level escape pixel presence flag, wherein the CU-level escape pixel presence flag indicates whether there is any sample in the current CU coded as an escape pixel; and

determining whether or not to constrain the CU-level escape pixel presence flag to one according to the current palette size or both of the predicted palette size and the new palette size, and constrain the CU-level escape pixel presence flag to one according to the determination of whether or not to constrain the CU-level escape pixel presence flag to one.