Implicit view-dependent quantization

1. A system for encoding geometric data, the system comprising:

a processor; and

a memory, the memory including code as instructions that cause the processor to:

receive geometric data and a viewpoint associated with the geometric data,

quantize the geometric data based on the viewpoint using a plurality of view-dependent quantization values, quantizing the geometric data including:

transforming a distance z between a point and the viewpoint to a quantized distance z′=1/(1+z), the quantized distance having a density that is a function of a distance from the viewpoint,

mapping the point of the geometric data to one of the view-dependent quantization values based on the quantized distance,

each of the quantized distances having a different dimension based on the function, and

each of the view-dependent quantization values having a same bit length integer representation based on a corresponding view-dependent quantization value, and

encode the quantized geometric data.

2. The system of claim 1 , wherein the distances are less coarse close to the viewpoint and more coarse with increasing distance from the viewpoint.

3. The system of claim 1 , wherein the view-dependent quantization values are re-determined based on at least one of a change in the viewpoint and a change in a scene represented by the geometric data.

4. The system of claim 1 , wherein the geometric data represents a point cloud and the geometric construct is a two or more dimensional shape encompassing the point cloud.

5. The system of claim 1 , wherein the quantizing of the geometric data includes one of:

mapping a value associated with a point of the geometric data to the view-dependent quantization value that is less than the value associated with the point,

mapping a value associated with a point of the geometric data to the view-dependent quantization value that is greater than the value associated with the point, and

mapping a value associated with a point of the geometric data to the view-dependent quantization value that minimizes an error associated with the point.

6. A method for encoding geometric data, the method comprising:

receiving geometric data and a viewpoint associated with the geometric data;

quantizing a first point of the geometric data based on a distance z between the first point and the viewpoint using a first view-dependent quantization value, the first view-dependent quantization value being based on a transformed distance z′=1/(a*z{circumflex over ( )}c+b), with a>0, b>0, and c>=1;

quantizing a second point of the geometric data based on a distance between the second point and the viewpoint using a second view-dependent quantization value, the second point being further from the viewpoint than the first point, the second view-dependent quantization value corresponding to one of a plurality of quantized distances that is more coarse than one of the plurality of quantized distances corresponding to the first view-dependent quantization value, and the second view-dependent quantization value having a same bit length integer representation as the first view-dependent quantization value; and

encoding the quantized geometric data.

7. The method of claim 6 , wherein the quantizing of the geometric data includes converting a float value or double precision value associated with the geometric data into a b-bit integer representation.

8. The method of claim 6 , wherein the geometric data represents a point cloud and the geometric construct is a two or more dimensional shape encompassing the point cloud.

9. The method of claim 6 , wherein the quantizing of the geometric data includes one of:

mapping a value associated with at least one of the first point and the second point to a view-dependent quantization value that is less than a value associated with the first point or the second point,

mapping a value associated with at least one of the first point and the second point to a view-dependent quantization value that is greater than a value associated with the first point or the second point, and

mapping a value associated with at least one of the first point and the second point to a view-dependent quantization value that minimizes an error associated with the point.

10. The method of claim 6 , wherein an artifact associated with the first point is relatively small as compared to an artifact associated with the second point.

11. A non-transitory computer readable storage medium having code as instructions for encoding geometric data that cause a processor to:

receive geometric data and a viewpoint associated with the geometric data;

quantize the geometric data based on the viewpoint using a plurality of view-dependent quantization values, quantizing the geometric data including:

determining a distance between a point of the geometric data and the viewpoint,

transforming the distance z to a quantized distance z′=1/(z+d) for any number d>0 that is less coarse when closer to the viewpoint and more coarse when a greater distance from the viewpoint,

mapping the point of the geometric data to one of the view-dependent quantization values based on the quantized distance,

each of the quantized distances having a different dimension, and

each of the view-dependent quantization values having a same bit length integer representation based on a corresponding view-dependent quantization value; and

encode the quantized geometric data.

12. The non-transitory computer readable storage medium of claim 11 , wherein the quantizing of the geometric data includes one of:

mapping a value associated with the point of the geometric data to the view-dependent quantization value that is less than the value associated with the point,

mapping a value associated with the point of the geometric data to the view-dependent quantization value that is greater than the value associated with the point, and

mapping a value associated with the point of the geometric data to the view-dependent quantization value that minimizes an error associated with the point.

13. The non-transitory computer readable storage medium of claim 11 , wherein the view-dependent quantization values are re-determined based on at least one of a change in the viewpoint and a change in a scene represented by the geometric data.

14. The non-transitory computer readable storage medium of claim 11 , wherein an artifact associated with the quantized distance that is less coarse is relatively small as compared to an artifact associated with the quantized distance that is more coarse.