Color image watermarking

Patent metadata

Field	Value
Publication number	US-10037587-B2
Application number	US-201615360936-A
Country	US
Kind code	B2
Filing date	Nov 23, 2016
Priority date	Nov 23, 2016
Publication date	Jul 31, 2018
Grant date	Jul 31, 2018

Abstract

One example embodiment is a method to embed a watermark image into a host image such that a watermarked image is generated. The method divides a host image into an Red (R) component, a Green (G) component, and a Blue (B) component, and partitions each component of the R, G, and B components into non-overlapping blocks from which embedding blocks are selected to embed watermark information such that a watermarked image is generated.

First claim

What is claimed is:

1. A method that is executed by a computer system to embed a watermark image into a host image to generate a watermarked image, the method comprising:
providing the host image as a color image;
providing the watermark image as a color image;
dividing, by the computer system, the watermark image into an Red (R) component, a Green (G) component, and a Blue (B) component;
converting, by the computer system and for each component of the R, G, and B components of the watermark image, each pixel value into a binary sequence to generate a plurality of binary sequences for each component of the R, G, and B components of the watermark image;
combining, by the computer system and for each component of the R, G, and B components of the watermark image, the plurality of binary sequences to form a component binary sequence such that three component binary sequences are formed that correspond to the R, G, and B components of the watermark image;
dividing, by the computer system, the host image into an R component, a G component, and a B component;
partitioning, by the computer system, each component of the R, G, and B components of the host image into non-overlapping blocks with a non-uniform subdivision treatment;
selecting, by the computer system and for each component of the R, G, and B components of the host image, embedding blocks of the host image;
performing, by the computer system and for each component of the R, G, and B components of the host image, iteration steps that embed watermark information into the embedding blocks until the embedding blocks are converted into watermarked blocks such that a watermarked component image is formed that includes the watermarked blocks; and
reconstructing, by the computer system, three watermarked component images that correspond to the R, G, and B components of the host image such that the watermark image is embedded into the host image to obtain the watermarked image.

2. The method of claim 1 , wherein the iteration steps comprising:
decomposing, by the computer system, an embedding block with QR decomposition to obtain a matrix R and a matrix Q, wherein the matrix R is an upper triangular matrix and the matrix Q is an orthogonal matrix;
embedding, by the computer system and for the embedding block, watermark information into the matrix R;
conducting, by the computer system and for the embedding block, an inverse QR operation to obtain a watermarked block; and
moving, by the computer system, to next embedding block.

3. The method of claim 1 , further comprising:
storing, by the computer system, results of the partitioning for each component of the R, G, and B components of the host image into an array RNO, wherein the results include pixel locations X and Y, blocks' width W and height H for the non-overlapping blocks; and
wherein the iteration steps further comprising:
calculating, by the computer system and based on the array RNO, an area size RCOUNT of the embedding block;
calculating, by the computer system and for the embedding block, quantization step Δ with equation (1):
Δ=0.01+RCOUNT Δ a/ 16  (1),

wherein a is a coefficient of x in an approximate polynomial f(x,y) that is used for the non-uniform subdivision treatment;

determining, by the computer system and based on watermark information w, for the embedding block, modifying values T 1 and T 2 with equations (2) and (3):

if
⁢

⁢
w

=
0

,

{

T
1

=

-
0.5

⁢
Δ

T
2

=

1.5
⁢
Δ

,

(
2
)

if
⁢

⁢
w

=
1

,

{

T
1

=

0.5
⁢
Δ

T
2

=

-
1.5

⁢
Δ

;

(
3
)

determining, by the computer system and for the embedding block, quantization results C 1 and C 2 with equations (4) and (5):
C 1 =2 kΔ+T 1   (4)

C 2 =2 kΔ+T 2   (5),

wherein

k
=

Assignees

• Univ Macau Sci & Tech

Inventors

• Cai Zhanchuan

Classifications

• G06T7/90 Primary

  Determination of colour characteristics · CPC title

• G06T2201/0061

  Embedding of the watermark in each block of the image, e.g. segmented watermarking · CPC title

• G06T1/005

  Robust watermarking, e.g. average attack or collusion attack resistant · CPC title

• G06T1/0021 Primary

  Image watermarking · CPC title

• G06T2207/20021

  Dividing image into blocks, subimages or windows · CPC title