Vextex correction method and apparatus for rotated three-dimensional (3D) components
View Patent ↗A 3D model can be modeled using pattern-instance representation, wherein an instance component may be represented as transformation (for example, rotation, translation, and scaling) of a pattern. Quantization errors may be introduced when encoding rotation information, causing different vertex coordinate errors at different vertices of an instance. To efficiently compensate the vertex coordinate errors, an upper bound can be estimated for the vertex coordinate error of a vertex. Based on the upper bound, the codec decides whether the vertex coordinate error of the vertex needs to be compensated, and decides a quantization parameter for compensating the vertex coordinate error if compensation is needed. The upper bound can be estimated at both the encoder and decoder, and thus, no explicit signaling is needed to indicate whether vertex coordinate error compensation is used or to indicate the quantization parameter for the vertex coordinate error.
1. A method for generating or decoding a bitstream representing a 3D model, comprising:
accessing a reconstructed instance corresponding to an instance associated with a pattern, the pattern representative of a plurality of repetitive structures in the 3D model, and the instance being one of the plurality of repetitive structures and represented as a transformation of the pattern;
accessing a first quantization parameter used for quantizing a rotation part of the transformation, wherein the reconstructed instance is determined in response to the pattern and the first quantization parameter;
estimating an upper bound of a vertex coordinate error between a vertex of the instance and a corresponding vertex of the reconstructed instance, in response to at least one of (1) the first quantization parameter and (2) a distance between the corresponding vertex of the reconstructed instance and a rotation center of the reconstructed instance;
determining a second quantization parameter for quantizing the vertex coordinate error in response to the estimated upper bound; and
performing encoding or decoding of the vertex coordinate error between the vertex of the instance and the corresponding vertex of the reconstructed instance based on the second quantization parameter.
2. The method of claim 1 , further comprising:
determining whether the vertex coordinate error needs to be encoded or decoded in response to the at least one of (1) the first quantization parameter and (2) the distance between the corresponding vertex of the reconstructed instance and the rotation center of the reconstructed instance.
3. The method of claim 1 , wherein the upper bound corresponds to
2
π
2
n
r
,
wherein n corresponds to the first quantization parameter for the rotation part, and r is the distance between the corresponding vertex of the reconstructed instance and the rotation center of the reconstructed instance.
4. The method of claim 1 , wherein the second quantization parameter is determined further in response to a maximum tolerable error.
5. The method of claim 4 , wherein the maximum tolerable error is determined in response to a quantization error associated with quantizing a translation part of the transformation or quantizing geometry of the pattern.
6. The method of claim 5 , wherein the quantization error associated with quantizing the translation part of the transformation is determined in response to a third quantization parameter for quantizing the translation part and a range for the translation part.
7. The method of claim 6 , wherein the second quantization parameter is a function of
Q
translation
-
log
2
range
e
est
,
wherein Q translation is the third quantization parameter for quantizing the translation part, range is the maximum range for the translation part, and e est is the estimated upper bound.
8. The method of claim 1 , further comprising:
determining the vertex coordinate error between the vertex of the instance and the corresponding vertex of the reconstructed instance;
quantizing the determined vertex coordinate error in response to the second quantization parameter to form a quantized error; and
encoding the quantized error into the bitstream.
9. The method of claim 1 , further comprising:
determining the vertex coordinate error between the vertex of the instance and the corresponding vertex of the reconstructed instance from the bitstream;
de-quantizing the determined vertex coordinate error in response to the second quantization parameter to form a de-quantized error; and
refining the reconstructed instance in response to the de-quantized error.
10. An apparatus for generating or decoding a bitstream representing a 3D model, comprising at least a memory and one or more processors, the one or more processors being configured to:
access a reconstructed instance corresponding to an instance associated with a pattern, the pattern representative of a plurality of repetitive structures in the 3D model and the instance being one of the plurality of repetitive structures and represented as a transformation of the pattern, and the reconstruction being in response to the pattern and a first quantization parameter used for quantizing a rotation part of the transformation;
estimate an upper bound of a vertex coordinate error, between a vertex of the instance and a corresponding vertex of the reconstructed instance, along a particular axis in response to at least one of (1) the first quantization parameter and (2) a distance between the corresponding vertex of the reconstructed instance and the rotation center of the reconstructed instance;
determine a second quantization parameter for quantizing the vertex coordinate error along the particular axis, in response to the estimated upper bound for the vertex coordinate error along the particular axis; and
encode or decode the vertex coordinate error between the vertex of the instance and the corresponding vertex of the reconstructed instance based on the second quantization parameter.
11. The apparatus of claim 10 , wherein the one or more processors are further configured to determine whether the vertex coordinate error needs to be encoded or decoded in response to the at least one of (1) the first quantization parameter and (2) the distance between the corresponding vertex of the reconstructed instance and the rotation center of the reconstructed instance.
12. The apparatus of claim 10 , wherein the upper bound corresponds to
2
π
2
n
r
,
wherein n corresponds to the first quantization parameter for the rotation part, and r is the distance between the corresponding vertex of the reconstructed instance and the rotation center of the reconstructed instance.
13. The apparatus of claim 10 , wherein the one or more processors are configured to determine a maximum tolerable error in response to a quantization error associated with quantizing a translation part of the transformation or quantizing geometry of the pattern.
14. The apparatus of claim 13 , wherein the one or more processors are configured to determine the quantization error associated with quantizing the translation part of the transformation in response to a third quantization parameter for quantizing the translation part and a range for the translation part.
15. The apparatus of claim 14 , wherein the second quantization parameter is a function of
Q
translation
-
log
2
range
e
est
,
wherein Q translation is the third quantization parameter for quantizing the translation part, range is the maximum range for the translation part, and e est is the estimated upper bound.
16. The apparatus of claim 10 , wherein the one or more processors are further configured to:
determine the vertex coordinate error between the vertex of the instance and the corresponding vertex of the reconstructed instance; and
encode the determined vertex coordinate error in response to the second quantization parameter.
17. The apparatus of claim 10 , wherein the one or more processors are further configured to:
determine the vertex coordinate error between the vertex of the instance and the corresponding vertex of the reconstructed instance from the bitstream;
decode the determined vertex coordinate error in response to the second quantization parameter to form a de-quantized error; and
refine the reconstructed instance in response to the de-quantized error.