Granted Patent B2

US 12,532,003 · App. 18/342,669 · Granted Jan 20, 2026

Inventors: Xiaobo Li (Shanghai, CN); Tianxiao Ye (Shanghai, CN)

Assignee: SHANGHAI BILIBILI TECHNOLOGY CO., LTD.

H04N19/182H04N19/119H04N19/132H04N19/44H04N19/89H04N19/895H04N19/96

Abstract

Claims (100)

1 . A method, comprising:

acquiring a target reconstructed image;

dividing the target reconstructed image into a plurality of non-overlapping coding tree blocks;

traversing all reconstructed pixels in each coding tree block to calculate a band value of each traversed reconstructed pixel and a residual value between each traversed reconstructed pixel and a corresponding original pixel;

storing the calculated band value of each reconstructed pixel in a first array and the calculated residual value between each reconstructed pixel and the corresponding original pixel in a second array;

in response to determining that m consecutive band values in the first array belong to a same band, wherein m is an integer greater than or equal to 2:

calculating, at a time based on the residual values in the second array of m reconstructed pixels corresponding to the m consecutive band values, a first cumulative residual value of the m reconstructed pixels; and

calculating a first cumulative number of the m reconstructed pixels;

in response to determining that the m consecutive band values in the first array do not belong to the same band:

in response to determining that n consecutive band values of the m consecutive band values belong to the same band, wherein n<m and n is an integer greater than or equal to 2:

calculating, at a time based on the residual values in the second array of n reconstructed pixels corresponding to the n consecutive band values, a second cumulative residual value of the n reconstructed pixels; and

calculating a second cumulative number of the n reconstructed pixels; adding the first cumulative residual value or the second cumulative residual value to a sum variable of said same band, and adding the first cumulative number or the second cumulative number to a count variable of said same band; and

performing a pixel compensation on the target reconstructed image based on the sum variable and the count variable.

2 . The method according to claim 1 , further comprising:

reading the m consecutive band values from the first array;

storing the read band values in a first variable with preset bits;

fetching one of the band values from the first variable;

copying the fetched band value m−1 times to obtain the copied band values;

storing the copied band values in a second variable with the preset bits;

in response to determining that the first variable is equal to the second variable, determining that the m consecutive band values in the first array belong to the same band; and

in response to determining that the first variable is not equal to the second variable, determining that the m consecutive band values in the first array do not belong to the same band.

3 . The method according to claim 2 , further comprising: when the first variable is not equal to the second variable,

acquiring, from each of the first variable and the second variable, values in a preset number of bits that correspond to the n consecutive band values; and

in response to determining that the values acquired from the first variable are equal to the values acquired from the second variable, determining that the n consecutive band values of the m consecutive band values belong to the same band.

4 . The method according to claim 3 ,

wherein m=8, n=4,

wherein each band value is represented by an 8-bit value,

wherein the first variable with the preset bits is a 64-bit first variable, and

wherein acquiring, from each of the first variable and the second variable, values in the preset number of bits that correspond to the n consecutive band values comprises:

acquiring, from each of the first variable and the second variable, values in 32 least significant bits or 32 most significant bits.

5 . The method according to claim 2 , wherein fetching one of the band values from the first variable comprises:

fetching a band value at a preset location from the first variable.

6 . A computer device, comprising:

a processor; and

a memory, wherein the memory stores computer executable instructions that, when executed by the processor, cause the processor to:

acquire a target reconstructed image

divide the target reconstructed image into a plurality of non-overlapping coding tree blocks;

traverse all reconstructed pixels in each coding tree block to calculate a band value of each traversed reconstructed pixel, and a residual value between each traversed reconstructed pixel and a corresponding original pixel;

store the calculated band value of each reconstructed pixel in a first array and the calculated residual value between each reconstructed pixel and the corresponding original pixel in a second array;

in response to determining that m consecutive band values in the first array belong to a same band, wherein m is an integer greater than or equal to 2:

calculate a first cumulative residual value of m reconstructed pixels corresponding to the m consecutive band values at a time based on the residual values in the second array of the m reconstructed pixels; and

calculate a first cumulative number of the m reconstructed pixels;

in response to determining that the m consecutive band values in the first array do not belong to the same band:

in response to determining that n consecutive band values of the m consecutive band values belong to the same band, wherein n<m and n is an integer greater than or equal to 2:

calculate a second cumulative residual value of n reconstructed pixels corresponding to the n consecutive band values at a time based on the residual values in the second array of the n reconstructed pixels; and

calculate a second cumulative number of the n reconstructed pixels;

adding the first cumulative residual value or the second cumulative residual value to a sum variable of said same band, and adding the first cumulative number or the second cumulative number to a count variable of said same band; and

performing a pixel compensation on the target reconstructed image based on the sum variable and the count variable.

7 . The computer device according to claim 6 , wherein the computer executable instructions, when executed by the processor, further cause the processor to:

read the m consecutive band values from the first array;

store the read band values in a first variable with preset bits;

fetch one of the band values from the first variable;

copy the fetched band value m−1 times to obtain the copied band values;

store the copied band values in a second variable with the preset bits;

in response to determining that the first variable is equal to the second variable, determine that the m consecutive band values in the first array belong to the same band; and

in response to determining that the first variable is not equal to the second variable, determine that the m consecutive band values in the first array do not belong to the same band.

8 . The computer device according to claim 7 , wherein the computer executable instructions, when executed by the processor, further cause the processor to: when the first variable is not equal to the second variable,

acquire, from each of the first variable and the second variable, values in a preset number of bits that correspond to the n consecutive band values; and

in response to determining that the values acquired from the first variable are equal to the values acquired from the second variable, determine that the n consecutive band values of the m consecutive band values belong to the same band.

9 . The computer device according to claim 8 ,

wherein m=8, n=4,

wherein each band value is represented by an 8-bit value,

wherein the first variable with the preset bits is a 64-bit first variable, and

wherein acquiring, from each of the first variable and the second variable, values in the preset number of bits that correspond to the n consecutive band values comprises:

acquiring, from each of the first variable and the second variable, values in 32 least significant bits or 32 most significant bits.

10 . The computer device according to claim 7 , wherein fetching one of the band values from the first variable comprises:

fetching a band value at a preset location from the first variable.

11 . A non-transitory computer-readable storage medium having computer instructions stored thereon that, when executed by a processor, cause the processor to:

acquire a target reconstructed image

divide the target reconstructed image into a plurality of non-overlapping coding tree blocks;

traverse all reconstructed pixels in each coding tree block to calculate a band value of each traversed reconstructed pixel, and a residual value between each traversed reconstructed pixel and a corresponding original pixel;

store the calculated band value of each reconstructed pixel in a first array and the calculated residual value between each reconstructed pixel and the corresponding original pixel in a second array;

in response to determining that m consecutive band values in the first array belong to a same band, wherein m is an integer greater than or equal to 2:

calculate a first cumulative residual value of m reconstructed pixels corresponding to the m consecutive band values at a time based on the residual values in the second array of the m reconstructed pixels; and

calculate a first cumulative number of the m reconstructed pixels;

in response to determining that the m consecutive band values in the first array do not belong to the same band:

in response to determining that n consecutive band values of the m consecutive band values belong to the same band, wherein n<m and n is an integer greater than or equal to 2:

calculate a second cumulative residual value of n reconstructed pixels corresponding to the n consecutive band values at a time based on the residual values in the second array of the n reconstructed pixels; and

calculate a second cumulative number of the n reconstructed pixels;

add the first cumulative residual value or the second cumulative residual value to a sum variable of said same band, and add the first cumulative number or the second cumulative number to a count variable of said same band; and

perform a pixel compensation on the target reconstructed image based on the sum variable and the count variable.

12 . The non-transitory computer-readable storage medium according to claim 11 , wherein the computer instructions, when executed by the processor, further cause the processor to:

read the m consecutive band values from the first array;

store the read band values in a first variable with preset bits:

fetch one of the band values from the first variable;

copy the fetched band value m−1 times to obtain the copied band values;

store the copied band values in a second variable with the preset bits;

in response to determining that the first variable is equal to the second variable, determine that the m consecutive band values in the first array belong to the same band; and

in response to determining that the first variable is not equal to the second variable, determine that the m consecutive band values in the first array do not belong to the same band.

13 . The non-transitory computer-readable storage medium according to claim 12 , wherein the computer instructions, when executed by the processor, further cause the processor to: when the first variable is not equal to the second variable,

acquire, from each of the first variable and the second variable, values in a preset number of bits that correspond to the n consecutive band values; and

in response to determining that the values acquired from the first variable is equal to the values acquired from the second variable, determine that the n consecutive band values of the m consecutive band values belong to the same band.

14 . The non-transitory computer-readable storage medium according to claim 13 ,

wherein m=8, n=4,

wherein each band value is represented by an 8-bit value,

wherein the first variable with the preset bits is a 64-bit first variable, and

wherein acquiring, from each of the first variable and the second variable, values in the preset number of bits that correspond to the n consecutive band values comprises:

acquiring, from each of the first variable and the second variable, values in 32 least significant bits or 32 most significant bits.

15 . The non-transitory computer-readable storage medium according to claim 12 , wherein fetching one of the band values from the first variable comprises:

fetching a band value at a preset location from the first variable.

Assignments (1)

Priority Claims (1)

Continuity (2)

References Cited (35)