# US 2018/0324426 A1 · Video processing method

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2018/0324426 A1
- Document ID: US20180324426A1
- Status: Published application
- Kind: A1
- Application number: 15/585,422
- Filed: 2017-05-03
- Published: 2018-11-08
- Earliest claimed: 2017-05-03
- Assignees: Novatek Microelectronics Corp.
- Applicants: Novatek Microelectronics Corp.
- Inventors: Xin Huang; Fan-Di Jou
- Web page: https://patentgrep.com/patent/US20180324426A1

## CPC

- H04N 19/124: Quantisation
- H04N 19/137: Motion inside a coding unit, e.g. average field, frame or block difference
- H04N 19/14: Coding unit complexity, e.g. amount of activity or edge presence estimation (H04N19/146 takes precedence)
- H04N 19/176: the region being a block, e.g. a macroblock
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION

## Abstract

A video processing method includes: an i-th block of a target frame partitioned into a plurality of blocks is received; at least one of a spatial attribute parameter and a temporal attribute parameter of an i-th block of the target frame is determined; and at least one of a delta quantization parameter and a modified Lagrange multiplier is determined according to at least one of the spatial attribute parameter and the temporal attribute parameter, for encoding the i-th block of the target frame.

## Claims

### Claim 1

A video processing method, adapted to a video encoder, comprising:
receiving an i-th block of a target frame partitioned into a plurality of blocks;
determining at least one of a spatial attribute parameter and a temporal attribute parameter of an i-th block of the target frame, wherein the spatial attribute parameter is a logarithmic approximation of a first value associated with the variance of pixel data in the i-th block of the target frame, and the temporal attribute parameter is a logarithmic approximation of a second value associated with the variance of an absolute difference between the pixel data of the i-th block of the target frame and pixel data of a reference block, which is at the same position as the i-th block of the target frame, of a reference frame; and
determining at least one of a delta quantization parameter and a modified Lagrange multiplier according to at least one of the spatial attribute parameter and the temporal attribute parameter, for encoding the i-th block of the target frame.

### Claim 2 (depends on 1)

The video processing method according to claim 1, wherein determining at least one of the spatial attribute parameter and the temporal attribute parameter of an i-th block of the target frame comprises:
determining the spatial attribute parameter of the i-th block of the target frame by performing arithmetic operations on a value, wherein the value is determined by searching a look up table based on the first value associated with the variance of pixel data in the i-th block of the target frame.

### Claim 3 (depends on 1)

The video processing method according to claim 1, wherein determining at least one of the spatial attribute parameter and the temporal attribute parameter of an i-th block of the target frame comprises:
determining the temporal attribute parameter of the i-th block of the target frame by performing arithmetic operations on a value, wherein the value is determined by searching a look up table based on the second value associated with the variance of the absolute difference between the pixel data of the i-th block of the target frame and pixel data of a reference block.

### Claim 4 (depends on 1)

The video processing method according to claim 1, wherein determining at least one of the delta quantization parameter and the modified Lagrange multiplier according to at least one of the spatial attribute parameter and the temporal attribute parameter comprises:
determining the delta quantization parameter based on at least one of (1) a difference between the spatial attribute parameter of the i-th block of the target frame and a global spatial attribute parameter of the target frame and (2) a difference between the temporal attribute parameter of the i-th block of the target frame and a global temporal attribute parameter of the target frame;
wherein the global spatial attribute parameter is determined by an average of spatial attribute parameters of the plurality of blocks of the target frame, and the global temporal attribute parameter is determined by an average of temporal attribute parameters of the plurality of blocks of the target frame.

### Claim 5 (depends on 4)

The video processing method according to claim 4, wherein the delta quantization parameter, ΔQP, is determined as:
Δ


QP
=
a

s
_

(
s
i
-
s
_
)
+
b

t
_

(
t
i
-
t
_
)
s
_
+
t
_
,
where s is the global spatial attribute parameter, t is the global temporal spatial attribute parameter, si is the spatial attribute parameter of the i-th block of the target frame, ti is the temporal attribute parameter of the i-th block of the target frame, and a and b are constants.

### Claim 6 (depends on 1)

The video processing method according to claim 1, further comprising:
comparing the i-th block of the target frame with a prediction result with respect to the i-th block of the target frame to generate residual data for the i-th block;
transforming the residual data into transformed data;
quantizing the transformed data by using a quantization step to generate quantized data, wherein the quantization step is determined according to a quantization parameter adjusted by the delta quantization parameter; and
performing entropy encoding on the quantized data to generate output data of the video encoder.

### Claim 7 (depends on 1)

The video processing method according to claim 1, wherein determining at least one of the delta quantization parameter and the modified Lagrange multiplier according to at least one of the spatial attribute parameter and the temporal attribute parameter comprises:
determining the modified Lagrange multiplier to be as a linear combination of an original Lagrange multiplier multiplied by a spatial weighting factor of the i-th block of the target frame and the original Lagrange multiplier multiplied by a temporal weighting factor of the i-th block of the target frame, as λi=a*ws*λ+b*wt*A,
wherein the spatial weighting factor, ws, is determined as
w
s
=
k
1

σ
x
,
i
2
+
c
1
exp

(
s
_
)
,
where s is a global spatial attribute parameter determined by an average of spatial attribute parameters of the plurality of blocks of the target frame, σx,i2 is the variance of the pixel data in the i-th block of the target frame, and k1 and c1 are constants; and
wherein the temporal weighting factor, wt, is determined as
w
t
=
k
2

σ
xD
,
i
2
+
c
2
exp

(
t
_
)
,
where t is a global temporal attribute parameter determined by an average of temporal attribute parameters of the plurality of blocks of the target frame, and σxD,i2 is the variance of the absolute difference between the pixel data of the i-th block of the target frame and the pixel data of the reference block of the reference frame, and k2 and c2 are constants; and
wherein a and b are constants.

### Claim 8 (depends on 1)

The video processing method according to claim 1, further comprising:
selecting a partition type for the i-th block of the target frame according to the modified Lagrange multiplier before the i-th block of the target frame is encoded.

### Claim 9

A video processing method, adapted to a video encoder, comprising:
receiving an i-th block of a target frame partitioned into a plurality of blocks;
determining a spatial attribute parameter and a temporal attribute parameter of an i-th block of the target frame, wherein the spatial attribute parameter is dependent on the variance of pixel data in the i-th block of the target frame, and the temporal attribute parameter is dependent on the variance of an absolute difference between the pixel data of the i-th block of the target frame and pixel data of a reference block, which is at the same position as the i-th block of the target frame, of a reference frame; and
determining at least one of a delta quantization parameter and a modified Lagrange multiplier according to the spatial attribute parameter and the temporal attribute parameter, for encoding the i-th block of the target frame.

### Claim 10 (depends on 9)

The video processing method according to claim 9, wherein determining the spatial attribute parameter of the i-th block of the target frame comprises:
determining the spatial attribute parameter of the i-th block of the target frame by performing arithmetic operations on a value, wherein the value is determined by searching a look up table based on a first value associated with the variance of pixel data in the i-th block of the target frame;
wherein the spatial attribute parameter is a logarithmic approximation of the first value.

### Claim 11 (depends on 9)

The video processing method according to claim 9, wherein determining the temporal attribute parameter of the i-th block of the target frame comprises:
determining the temporal attribute parameter of the i-th block of the target frame by performing arithmetic operations on a value, wherein the value is determined by searching a look up table based on a second value associated with the variance of an absolute difference between the pixel data of the i-th block of the target frame and pixel data of the reference block;
wherein the temporal attribute parameter is a logarithmic approximation of the second value.

### Claim 12 (depends on 9)

The video processing method according to claim 9, wherein determining at least one of the delta quantization parameter and the modified Lagrange multiplier according to the spatial attribute parameter and the temporal attribute parameter comprises:
determining the delta quantization parameter based on at least one of (1) a difference between the spatial attribute parameter of the i-th block of the target frame and a global spatial attribute parameter of the target frame and (2) a difference between the temporal attribute parameter of the i-th block of the target frame and a global temporal attribute parameter of the target frame;
wherein the global spatial attribute parameter is determined by an average of spatial attribute parameters of the plurality of blocks of the target frame, and the global temporal attribute parameter is determined by an average of temporal attribute parameters of the plurality of blocks of the target frame.

### Claim 13 (depends on 12)

The video processing method according to claim 12, wherein the delta quantization parameter, ΔQP, is determined as:
Δ


QP
=
a

s
_

(
s
i
-
s
_
)
+
b

t
_

(
t
i
-
t
_
)
s
_
+
t
_
,
where s is the global spatial attribute parameter, t is the global temporal spatial attribute parameter, si is the spatial attribute parameter of the i-th block of the target frame, ti is the temporal attribute parameter of the i-th block of the target frame, and a and b are constants.

### Claim 14 (depends on 9)

The video processing method according to claim 9, further comprising:
comparing the i-th block of the target frame with a prediction result with respect to the i-th block of the target frame to generate residual data for the i-th block;
transforming the residual data into transformed data;
quantizing the transformed data by using a quantization step to generate quantized data, wherein the quantization step is determined according to a quantization parameter adjusted by the delta quantization parameter; and
performing entropy encoding on the quantized data to generate output data of the video encoder.

### Claim 15 (depends on 9)

The video processing method according to claim 9, wherein determining at least one of the delta quantization parameter and the modified Lagrange multiplier according to the spatial attribute parameter and the temporal attribute parameter comprises:
determining the modified Lagrange multiplier to be as a linear combination of an original Lagrange multiplier multiplied by a spatial weighting factor of the i-th block of the target frame and the original Lagrange multiplier multiplied by a temporal weighting factor of the i-th block of the target frame, as λi=a*ws*A+b*wt*A,
wherein the spatial weighting factor, ws, is determined as
w
s
=
k
1

σ
x
,
i
2
+
c
1
exp

(
s
_
)
,
where s is a global spatial attribute parameter determined by an average of spatial attribute parameters of the plurality of blocks of the target frame, σx,i2 is the variance of the pixel data in the i-th block of the target frame, and k1 and c1 are constants; and
wherein the temporal weighting factor, wt, is determined as
w
t
=
k
2

σ
xD
,
i
2
+
c
2
exp

(
t
_
)
,
where t is a global temporal attribute parameter determined by an average of temporal attribute parameters of the plurality of blocks of the target frame, and σxD,i2 is the variance of the absolute difference between the pixel data of the i-th block of the target frame and the pixel data of the reference block of the reference frame, and k2 and c2 are constants; and
wherein a and b are constants.

### Claim 16 (depends on 9)

The video processing method according to claim 9, further comprising:
selecting a partition type for the i-th block of the target frame according to the modified Lagrange multiplier before the i-th block of the target frame is encoded.

### Claim 17

A video processing method, adapted to a video encoder, comprising:
receiving an i-th block of a target frame partitioned into a plurality of blocks;
determining a temporal attribute parameter of an i-th block of the target frame, wherein the temporal attribute parameter is dependent on the variance of an absolute difference between pixel data of the i-th block of the target frame and pixel data of a reference block, which is at the same position as the i-th block of the target frame, of a reference frame; and
determining at least one of a delta quantization parameter and a modified Lagrange multiplier according to the temporal attribute parameter, for encoding the i-th block of the target frame.

### Claim 18 (depends on 17)

The video processing method according to claim 17, wherein determining the temporal attribute parameter of the i-th block of the target frame comprises:
determining the temporal attribute parameter of the i-th block of the target frame by performing arithmetic operations on a value, wherein the value is determined by searching a look up table based on a first value associated with the variance of the absolute difference between the pixel data of the i-th block of the target frame and pixel data of the reference block;
wherein the temporal attribute parameter is a logarithmic approximation of the first value.

### Claim 19 (depends on 17)

The video processing method according to claim 17, wherein determining at least one of the delta quantization parameter and the modified Lagrange multiplier according to the temporal attribute parameter comprises:
determining the delta quantization parameter based on a difference between the temporal attribute parameter of the i-th block of the target frame and a global temporal attribute parameter of the target frame;
wherein the global temporal attribute parameter is determined by an average of temporal attribute parameters of the plurality of blocks of the target frame.

### Claim 20 (depends on 19)

The video processing method according to claim 19, wherein the delta quantization parameter, ΔQP, is determined as:
Δ


QP
=
b

t
_

(
t
i
-
t
_
)
t
_
,
where t is the global temporal spatial attribute parameter, ti is the temporal attribute parameter of the i-th block of the target frame, and b is a constant.

### Claim 21 (depends on 17)

The video processing method according to claim 17, further comprising:
comparing the i-th block of the target frame with a prediction result with respect to the i-th block of the target frame to generate residual data for the i-th block;
transforming the residual data into transformed data;
quantizing the transformed data by using a quantization step to generate quantized data, wherein the quantization step is determined according to a quantization parameter adjusted by the delta quantization parameter; and
performing entropy encoding on the quantized data to generate output data of the video encoder.

### Claim 22 (depends on 17)

The video processing method according to claim 17, wherein determining at least one of the delta quantization parameter and the modified Lagrange multiplier according to the temporal attribute parameter comprises:
determining the modified Lagrange multiplier to be as an original Lagrange multiplier multiplied by a temporal weighting factor of the i-th block of the target frame, as λi=a*wt*λ,
wherein the temporal weighting factor, wt, is determined as
w
t
=
k
2

σ
xD
,
i
2
+
c
2
exp

(
t
_
)
,
where t is a global temporal attribute parameter determined by an average of temporal attribute parameters of the plurality of blocks of the target frame, and σxD,i2 is the variance of the absolute difference between the pixel data of the i-th block of the target frame and the pixel data of the reference block of the reference frame, and k2 and c2 are constants; and
wherein a is a constant.

### Claim 23 (depends on 17)

The video processing method according to claim 17, further comprising:
selecting a partition type for the i-th block of the target frame according to the modified Lagrange multiplier before the i-th block of the target frame is encoded.
