CN113077381B - Color image description method based on ternary number continuous orthogonal moment - Google Patents
Color image description method based on ternary number continuous orthogonal moment Download PDFInfo
- Publication number
- CN113077381B CN113077381B CN202110427867.5A CN202110427867A CN113077381B CN 113077381 B CN113077381 B CN 113077381B CN 202110427867 A CN202110427867 A CN 202110427867A CN 113077381 B CN113077381 B CN 113077381B
- Authority
- CN
- China
- Prior art keywords
- color image
- theta
- ternary
- continuous orthogonal
- moments
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 17
- 239000011159 matrix material Substances 0.000 claims description 6
- 239000000126 substance Substances 0.000 claims description 4
- 238000007792 addition Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G06T3/04—
Abstract
The invention provides a color image description method based on ternary number continuous orthogonal moments, which comprises the following steps: a. representing a sub-polar color image f (r, theta)Is a set of pure ternary numbers; b. constructing three-element continuous orthogonal moment of the color image f (r, theta); c. using a finite number of ternary continuous orthogonal moments (with the highest order being n)maxThe maximum repetition degree is lmax) The original image f (r, θ) is approximately reconstructed. The invention describes the color image by constructing the ternary number continuous orthogonal moment, not only fully utilizes the relation among all channels, but also avoids information redundancy and reduces the calculation cost.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an image description method, and specifically relates to a color image description method based on ternary continuous orthogonal moments.
Background
The concept of moments originally appeared in the research field of statistics and classical mechanics, and Hu proposed Hu invariant moment for the first time in 1962, introduced into the image processing field, and proposed the image moment theory for describing image features. Image moments, which are a invariant feature describing an image, find wide application in many fields of image processing, such as object recognition, image retrieval, and image watermarking. However, the definition of the continuous orthogonal moments of the image is for grayscale images. As color images can provide more information, how to calculate the continuous orthogonal moments of color images attracts great attention.
In order to apply image continuous orthogonal moments to color images, there are generally two processing methods: one is to convert the color image into a gray image and then calculate the continuous orthogonal moment; the other is to calculate the successive orthogonal moments of the three channels of the color image separately and then simply combine them. Obviously, the two methods cannot capture the correlation between channels, lose image information and influence the calculation accuracy of moments.
For this reason, the quaternion theory is applied to the continuous orthogonal moments of the color image, and the internal relation between color image channels is fully preserved. Quaternion continuous orthogonal moments preserve the relationship between the three channels of the color image, but the fourth dimension of quaternions creates information redundancy when processing color images and is computationally expensive.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides the color image description method based on the ternary number continuous orthogonal moment, which makes full use of the relationship among all channels, avoids information redundancy and reduces the calculation cost.
The continuous orthogonal moment of the image is the projection of the image function on the continuous basis function, is a stable image characteristic, and has strong image description capability and geometric invariance. Continuous orthogonal moment M of image with order n of repetition degree l of polar coordinate image f (r, theta)nlIs defined as:
wherein C is a constant, and C is a constant,denotes the conjugated complex number, Pn(r) is a radial basis function, theta is a polar angle, the value range of theta is more than or equal to 0 and less than or equal to 2 pi, and the range of r is more than or equal to 0 and less than or equal to 1.
As can be seen from the theory of the orthogonal integral function system, the original image function f (r, θ) can be approximately reconstructed by using a finite number of moment-weighted superposition, and the greater the number of moments used, the higher the degree of approximation:
and a ternary number is an extension of a complex number, defined as μ ═ a + ib + jc, μ is a ternary number, where a, b, and c are real numbers, and i and j are operators that satisfy the following rules:
i2=j,ij=ji=-1,j2=-i
the three basic elements {1, i, j } of a ternary number form an abelian group, where 1 is the multiplicative unit bit and μ is a pure ternary number when a is 0.
The ternary number μ ═ a + ib + jc can be written as follows:
μ=|μ|(cos(φ)+μsin(φ))
wherein the content of the first and second substances,is the magnitude of the ternary number, and when | μ | ═ 1, μ is the unit ternary number.
Mu.s of1=a1+ib1+jc1And mu2=a2+ib2+jc2Is two ternary numbers, then μ is defined1And mu2Satisfy addition and multiplication of
μ1+μ2=(a1+a2)+i(b1+b2)+j(c1+c2)
μ1μ2=(a1a2-b1c2-c1b2)+i(a1b2+b1a2-c1c2)+j(a1c2+b1b2+c1a2)
Unlike quaternions, quaternions are interchangeable under both addition and multiplication:
μ1+μ2=μ2+μ1
μ1μ2=μ2μ1
note that the ternary numbers are not reversible.
The invention describes the color image by constructing the ternary number continuous orthogonal moment, not only fully utilizes the relation among all channels, but also avoids information redundancy and reduces the calculation cost.
The invention is realized by the following technical scheme, and provides a color image description method based on ternary continuous orthogonal moments, which comprises the following steps:
a. a polar color image f (r, theta) is represented as a set of pure ternary numbers,
f(r,θ)=R(r,θ)+G(r,θ)i+B(r,θ)k
wherein, R (R, θ), G (R, θ) and B (R, θ) represent the red, green and blue three channel components of the color image f (R, θ), respectively;
b. constructing three-element continuous orthogonal moments of the color image f (r, theta),
c. using a finite number of ternary continuous orthogonal moments (with the highest order being n)maxThe maximum repetition degree is lmax) Approximate reconstructed original image function f (r, θ):
Further, the calculation process of the three-element continuous orthogonal moment of the color image f (r, θ) in step b is as follows:
wherein the content of the first and second substances,
wherein, Mnl(R)、Mnl(G) And Mnl(B) The successive orthogonal moments of the three channels R (R, θ), G (R, θ) and B (R, θ), Re (·) and Im (·) of the color image f (R, θ), respectively, represent the real and imaginary parts, respectively, of the complex number.
Further, the calculation process of the reconstructed color image f (r, θ) according to the three-element continuous orthogonal moments in the step c is as follows:
where R '(R, θ), G' (R, θ) and B '(R, θ) are the three red, green and blue channels of the reconstructed color image f' (R, θ), respectively, and are derived by the following equation:
R′(r,θ)=Re(A(r,θ))+α2Im(A(r,θ))-γ2Im(D(r,θ))-β2Im(E(r,θ))
G′(r,θ)=Re(D(r,θ))+β2Im(A(r,θ))+α2Im(D(r,θ))-γ2Im(E(r,θ))
B′(r,θ)=Re(E(r,θ))+γ2Im(A(r,θ))+β2Im(D(r,θ))+α2Im(E(r,θ))
wherein A (r, theta), D (r, theta) and E (r, theta) are respectively Anl、DnlAnd EnlThe resulting matrix is reconstructed and the matrix is reconstructed,
in conclusion, the color image is described by constructing the ternary continuous orthogonal moment, so that the relationship among channels is fully utilized, information redundancy is avoided, and the calculation cost is reduced.
Drawings
FIG. 1 is a flow chart of the steps of calculating and reconstructing a color image description method based on three-component continuous orthogonal moments according to the present invention;
Detailed Description
In order to clearly illustrate the technical features of the present invention, the present invention is further illustrated by the following specific embodiments.
A color image description method based on ternary number continuous orthogonal moment is characterized by comprising the following steps:
a. a polar color image f (r, theta) is represented as a set of pure ternary numbers,
f(r,θ)=R(r,θ)+G(r,θ)i+B(r,θ)k
wherein, R (R, θ), G (R, θ) and B (R, θ) represent the red, green and blue three channel components of the color image f (R, θ), respectively;
b. constructing three-element continuous orthogonal moments of the color image f (r, theta),
c. using a finite number of ternary continuous orthogonal moments (with the highest order being n)maxThe maximum repetition degree is lmax) Approximate reconstructed original image function f (r, θ):
In this embodiment, the process of calculating the three-element continuous orthogonal moment of the color image f (r, θ) in step b is as follows:
wherein the content of the first and second substances,
Anl=Re(Mnl(R))-γ1Im(Mnl(G))-β1Im(Mnl(B))
Dnl=β1Im(Mnl(R))+Re(Mnl(G))-γ1Im(Mnl(B))
Enl=γ1Im(Mnl(R))+β1Im(Mnl(G))+Re(Mnl(B))
wherein, Mnl(R)、Mnl(G) And Mnl(B) The successive orthogonal moments of the three channels R (R, θ), G (R, θ) and B (R, θ), Re (·) and Im (·) of the color image f (R, θ), respectively, represent the real and imaginary parts, respectively, of the complex number.
The calculation process of the reconstructed color image f (r, theta) according to the ternary number continuous orthogonal moment in the step c is as follows:
where R '(R, θ), G' (R, θ) and B '(R, θ) are the three red, green and blue channels of the reconstructed color image f' (R, θ), respectively, and are derived by the following equation:
R′(r,θ)=Re(A(r,θ))+α2Im(A(r,θ))-γ2Im(D(r,θ))-β2Im(E(r,θ))
G′(r,θ)=Re(D(r,θ))+β2Im(A(r,θ))+α2Im(D(r,θ))-γ2Im(E(r,θ))
B′(r,θ)=Re(E(r,θ))+γ2Im(A(r,θ))+β2Im(D(r,θ))+α2Im(E(r,θ))
wherein A (r, theta), D (r, theta) and E (r, theta) are respectively Anl、DnlAnd EnlThe resulting matrix is reconstructed and the matrix is reconstructed,
finally, it should be further noted that the above examples and descriptions are not limited to the above embodiments, and technical features of the present invention that are not described may be implemented by or using the prior art, and are not described herein again; the above embodiments and drawings are only for illustrating the technical solutions of the present invention and not for limiting the present invention, and the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that changes, modifications, additions or substitutions within the spirit and scope of the present invention may be made by those skilled in the art without departing from the spirit of the present invention, and shall also fall within the scope of the claims of the present invention.
Claims (1)
1. A color image description method based on ternary continuous orthogonal moments is characterized by comprising the following steps:
a. a polar color image f (r, theta) is represented as a set of pure ternary numbers,
f(r,θ)=R(r,θ)+G(r,θ)i+B(r,θ)k
wherein, R (R, theta), G (R, theta) and B (R, theta) respectively represent red, green and blue channel components of the color image f (R, theta), theta is a polar angle and has a value range of 0-2 pi, and R has a value range of 0-1R;
b. constructing three-element continuous orthogonal moments of the color image f (r, theta),
wherein, mu1=β1i+γ1j is any unit pure ternary number, beta1 2+γ1 21 andPn(r) is a radial basis function;
c. reconstructing the original image f (r, θ) using a finite number of successive orthogonal moments approximations:
The calculation process of the three-element continuous orthogonal moment of the color image f (r, theta) in the step b is as follows:
wherein the content of the first and second substances,
Anl=Re(Mnl(R))-γ1Im(Mnl(G))-β1Im(Mnl(B))
Dnl=β1Im(Mnl(R))+Re(Mnl(G))-γ1Im(Mnl(B))
Enl=γ1Im(Mnl(R))+β1Im(Mnl(G))+Re(Mnl(B))
wherein M isnl(R)、Mnl(G) And Mnl(B) The successive orthogonal moments of the three channels R (R, θ), G (R, θ) and B (R, θ), Re (·) and Im (·) of the color image f (R, θ), respectively, represent the real and imaginary parts, respectively, of the complex number.
The calculation process of the reconstructed color image f (r, theta) according to the three-element continuous orthogonal moment in the step c is as follows:
where R '(R, θ), G' (R, θ) and B '(R, θ) are the three red, green and blue channels of the reconstructed color image f' (R, θ), respectively, and are derived by the following equation:
R′(r,θ)=Re(A(r,θ))+α2Im(A(r,θ))-γ2Im(D(r,θ))-β2Im(E(r,θ))
G′(r,θ)=Re(D(r,θ))+β2Im(A(r,θ))+α2Im(D(r,θ))-γ2Im(E(r,θ))
B′(r,θ)=Re(E(r,θ))+γ2Im(A(r,θ))+β2Im(D(r,θ))+α2Im(E(r,θ))
wherein A (r, θ), D (r, θ) and E (r, θ) are each Anl、DnlAnd EnlThe resulting matrix is reconstructed and the matrix is reconstructed,
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110427867.5A CN113077381B (en) | 2021-04-21 | 2021-04-21 | Color image description method based on ternary number continuous orthogonal moment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110427867.5A CN113077381B (en) | 2021-04-21 | 2021-04-21 | Color image description method based on ternary number continuous orthogonal moment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113077381A CN113077381A (en) | 2021-07-06 |
CN113077381B true CN113077381B (en) | 2022-06-07 |
Family
ID=76618198
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110427867.5A Active CN113077381B (en) | 2021-04-21 | 2021-04-21 | Color image description method based on ternary number continuous orthogonal moment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113077381B (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103761701A (en) * | 2013-12-28 | 2014-04-30 | 辽宁师范大学 | Color image watermarking method based on quaternion index matrix |
CN109726577A (en) * | 2019-01-10 | 2019-05-07 | 首都师范大学 | Image encryption method and device |
CN112614196A (en) * | 2020-12-16 | 2021-04-06 | 湖南科技大学 | Image robustness Hash authentication method based on quaternion convolution neural network |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
AUPQ055999A0 (en) * | 1999-05-25 | 1999-06-17 | Silverbrook Research Pty Ltd | A method and apparatus (npage01) |
-
2021
- 2021-04-21 CN CN202110427867.5A patent/CN113077381B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103761701A (en) * | 2013-12-28 | 2014-04-30 | 辽宁师范大学 | Color image watermarking method based on quaternion index matrix |
CN109726577A (en) * | 2019-01-10 | 2019-05-07 | 首都师范大学 | Image encryption method and device |
CN112614196A (en) * | 2020-12-16 | 2021-04-06 | 湖南科技大学 | Image robustness Hash authentication method based on quaternion convolution neural network |
Also Published As
Publication number | Publication date |
---|---|
CN113077381A (en) | 2021-07-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zhang et al. | Learning deep CNN denoiser prior for image restoration | |
CN108961186B (en) | Old film repairing and reproducing method based on deep learning | |
CN111369440A (en) | Model training method, image super-resolution processing method, device, terminal and storage medium | |
US11216913B2 (en) | Convolutional neural network processor, image processing method and electronic device | |
CN113298716B (en) | Image super-resolution reconstruction method based on convolutional neural network | |
CN110532928B (en) | Facial key point detection method based on facial region normalization and deformable hourglass network | |
Wang et al. | Image splicing detection based on convolutional neural network with weight combination strategy | |
CN112733795A (en) | Method, device and equipment for correcting sight of face image and storage medium | |
CN113096017A (en) | Image super-resolution reconstruction method based on depth coordinate attention network model | |
Cao et al. | New architecture of deep recursive convolution networks for super-resolution | |
Niu et al. | Effective image restoration for semantic segmentation | |
WO2023284608A1 (en) | Character recognition model generating method and apparatus, computer device, and storage medium | |
CN112257727B (en) | Feature image extraction method based on deep learning self-adaptive deformable convolution | |
CN111986092B (en) | Dual-network-based image super-resolution reconstruction method and system | |
CN111476133A (en) | Unmanned driving-oriented foreground and background codec network target extraction method | |
CN115713462A (en) | Super-resolution model training method, image recognition method, device and equipment | |
CN109993701B (en) | Depth map super-resolution reconstruction method based on pyramid structure | |
CN110047038B (en) | Single-image super-resolution reconstruction method based on hierarchical progressive network | |
CN113077381B (en) | Color image description method based on ternary number continuous orthogonal moment | |
CN114663297A (en) | Underwater image enhancement method based on multi-scale intensive generation countermeasure network and training method of network model | |
Hung et al. | Video super resolution via deep global-aware network | |
CN111611962A (en) | Face image super-resolution identification method based on fractional order multi-set partial least square | |
CN112907446B (en) | Image super-resolution reconstruction method based on packet connection network | |
CN115294424A (en) | Sample data enhancement method based on generation countermeasure network | |
Akram et al. | Integrating GAN with CNN for face sketch synthesis |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |