CN113409206A - High-precision digital printing color space conversion method - Google Patents

High-precision digital printing color space conversion method Download PDF

Info

Publication number
CN113409206A
CN113409206A CN202110654803.9A CN202110654803A CN113409206A CN 113409206 A CN113409206 A CN 113409206A CN 202110654803 A CN202110654803 A CN 202110654803A CN 113409206 A CN113409206 A CN 113409206A
Authority
CN
China
Prior art keywords
color space
neural network
space conversion
rbf neural
vector
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.)
Pending
Application number
CN202110654803.9A
Other languages
Chinese (zh)
Inventor
苏泽斌
赵思源
李鹏飞
景军锋
张缓缓
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Polytechnic University
Original Assignee
Xian Polytechnic University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xian Polytechnic University filed Critical Xian Polytechnic University
Priority to CN202110654803.9A priority Critical patent/CN113409206A/en
Publication of CN113409206A publication Critical patent/CN113409206A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/90—Dynamic range modification of images or parts thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Facsimile Image Signal Circuits (AREA)
  • Color Image Communication Systems (AREA)

Abstract

The invention discloses a high-precision digital printing color space conversion method, which comprises the steps of firstly, collecting color lump CMYK values and LAB values from a standard color card to construct a data set; establishing an RBF neural network model; training an RBF color space conversion model by using a data set; optimizing a central layer vector, a base width vector and a connection weight of the RBF neural network by using a whale optimization algorithm; after the color space conversion is finished, training the optimized RBF color space conversion model by using a training sample to obtain a final color space conversion model, and realizing the conversion from CMYK to LAB color space; testing the final color space conversion model by using the test sample, and checking the precision of the established color space conversion model; aiming at the problems of insufficient precision and difficult key parameter selection of the neural network method, the RBF neural network is optimized through the whale algorithm, so that the conversion precision of color space conversion is improved, and the method has strong adaptivity.

Description

High-precision digital printing color space conversion method
Technical Field
The invention belongs to the technical field of image processing, and relates to a high-precision digital printing color space conversion method.
Background
With the development of equipment manufacturing technology and the improvement of environmental protection requirements, the digital jet printing technology is widely applied to multiple fields of textile printing, paper printing and the like. In the digital printing process of textiles, because the color characteristics of different devices are different, the color information reproduced when the patterns are printed on the textiles can be influenced, and the color difference between a finished product and an original manuscript can be generated, so that the final quality of a printed product is influenced. Therefore, the printing process adopts the color management technology to ensure the accurate transmission of color information, and the core process is to convert colors among various input, output and display devices by taking a color space which is irrelevant to the devices as a medium, thereby realizing high-quality color reproduction. The CMYK color space is a color space commonly used in the digital printing industry at present, and represents abundant colors by mixing four colors of cyan (C), magenta (M), yellow (Y), and black (K). The LAB color space is a device-independent color space that is currently widely used by the international color consortium and whose principle is to describe color features by digitizing human visual perception. The L component in the LAB color space represents the brightness of the pigment, and the value range represents pure black to pure white; the a component and the B component in the LAB color space represent from red (high luminance value) to green (low luminance value) and from yellow (high luminance value) to blue (low luminance value), respectively, according to the difference in luminance. By establishing an accurate CMYK to LAB color space conversion model, accurate transmission of original color information can be ensured, and the final quality of a digital printing product is improved.
The nature of color space conversion is the problem of nonlinear mapping, so a Radial Basis Function (RBF) neural network with high-performance nonlinear mapping capability is very suitable for color space conversion, but the RBF neural network has the problems of insufficient precision and difficult selection of key parameters. A Whale Optimization Algorithm (WOA) is combined with the RBF neural network, three parameters of a central vector, a base width vector, a hidden layer and an output layer of the RBF neural network are optimized through direct connection weight, and the adaptivity and the conversion precision of the RBF neural network for color space conversion are improved.
Disclosure of Invention
The invention aims to provide a high-precision digital printing color space conversion method, wherein a whale optimization algorithm is used for optimizing a radial basis function neural network, and high-precision conversion from CMYK to LAB color space is realized.
The technical scheme adopted by the invention is that the high-precision digital printing color space conversion method is implemented according to the following steps:
step 1, collecting CMYK values and LAB values of color blocks from a standard color card to construct a data set;
step 2, establishing an RBF neural network model; initializing parameters in the model, setting learning rate and momentum factor of the network, and facilitating the data set constructed in the step 1 to train the RBF color space conversion model;
step 3, optimizing a central layer vector, a base width vector and a connection weight of the RBF neural network by using a whale optimization algorithm;
step 4, after the optimization in the step 3 is completed, training the optimized RBF color space conversion model by using the training sample in the step 1 to obtain a final color space conversion model, and realizing the conversion from CMYK to LAB color space;
and 5, testing the final color space conversion model by using the test sample in the step 1, and checking the accuracy of the established color space conversion model.
The invention is also characterized in that:
wherein the CMYK values are used as inputs to the neural network and the LAB values are used as outputs from the neural network in step 1; the data set is divided into a training sample and a testing sample, the training sample is used for training the color space conversion model, and the testing sample is used for evaluating the conversion precision of the trained model;
the specific process of establishing the RBF neural network model in the step 2 is as follows:
setting the learning rate of the RBF neural network to be 0.05 and the momentum factor to be 0.85, and initializing a central vector, a base width vector and a connection weight of the RBF neural network; selecting a Gaussian function as a base function of a hidden layer node, wherein the formula is shown as (1):
Figure BDA0003112222160000031
in the formula, xnFor the nth input sample, cn、σnRespectively representing the base function center and the base function width of the nth node of the hidden layer;
the output of the RBF neural network can be obtained from equation (2):
Figure BDA0003112222160000032
wherein the specific process optimized in the step 3 is implemented according to the following steps:
step 3.1, initializing a whale optimization algorithm, setting the number of whale populations to be N-10, and setting the maximum iteration number T of the algorithm max50, the position of each individual whale in the whale population comprises a central vector, a base width vector and a connection weight in the RBF neural network, error statistics is carried out on the calculation result of each individual whale, the position information of the individual with the minimum error is used as the current optimal solution and is recorded as X*(t)；
Step 3.2, converting X in step 3.1*(t) reserving, and updating the whale population position, wherein the updating is shown as a formula (3):
Figure BDA0003112222160000033
wherein A is 2a r-a, a is 2-2 (t/t)max) R is in [0,1 ]]Random vector of (d), tmaxRespectively representing the current iteration number and the maximum iteration number, X (t) representing the current position vector, b being a logarithmic spiral constant, l being located at [ -1,1]P is a random number located in [0,1 ]]The calculation of the random numbers between, D and D' is shown in equations (4) and (5):
D＝|C·X*(t)-X(t)| (4)
D′＝|X*(t)-X(t)| (5)
wherein, C is 2 r;
to avoid the algorithm falling into the local optimum, a random walk is added to update the position, as shown in equations (6) and (7):
X(t+1)＝Xr(t)-A·D (6)
D＝|C·Xr(t)-X(t)| (7)
in the formula, Xr(t) is a position vector of randomly selected whale individuals;
step 3.3, error calculation is carried out on the newly generated whale population, and the error calculation is carried out on the whale population and the X in the previous generation population*(t) comparing, and replacing the individuals with larger errors in the previous generation with the individuals with smaller errors in the whale population of the new generation to generate new X*(t)；
Step 3.4, for the current X*(t) evaluating, and if the iteration number N or the conversion precision set in the step 3.1 is reached, taking the individual position as an optimal solution
Figure BDA0003112222160000041
Assigning the position information to a central vector, a base width vector and a connection weight of the RBF neural network respectively, and returning to the step 3.2 to continuously seek an optimal solution if the iteration number N or the conversion precision is not met;
wherein the specific training process in the step 4 is as follows:
step 4.1, the optimal solution in step 3 is obtained
Figure BDA0003112222160000042
Decomposing the parameters into a center vector, a base width vector and a connection weight and inputting the parameters into the RBF neural network to obtain an optimized RBF neural network;
and 4.2, taking any CMYK value as an input value of the RBF neural network, and outputting the converted LAB value by the RBF neural network after the conversion by the RBF neural network.
The invention has the beneficial effects that:
the high-precision digital printing color space conversion method provided by the invention aims at the problems of insufficient precision and difficulty in key parameter selection of a neural network method, and not only is the conversion precision of color space conversion improved, but also the conversion method has strong adaptivity by optimizing the RBF neural network through a whale algorithm.
Drawings
FIG. 1 is a schematic flow chart of a high-precision digital printing color space conversion method according to the present invention;
FIG. 2 is a diagram of an RBF neural network structure in the high-precision digital printing color space conversion method of the present invention;
FIG. 3 is a flow chart of RBF neural network training in the high-precision digital printing color space conversion method of the present invention;
FIG. 4 is a flow chart of an algorithm for optimizing RBF neural network by whale optimization algorithm in the high-precision digital printing color space conversion method of the invention;
fig. 5 is a diagram of a conversion color difference distribution for verifying a designed color space conversion method according to an embodiment of the present invention.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
The invention provides a high-precision digital printing color space conversion method, which is implemented by the following steps as shown in figure 1:
step 1, collecting CMYK values and LAB values of color blocks from a standard color card to construct a data set, wherein the CMYK values are used as input of a neural network, and the LAB values are used as output of the neural network; the data set is divided into a training sample and a testing sample, the training sample is used for training the color space conversion model, and the testing sample is used for evaluating the conversion precision of the trained model;
step 2, as shown in fig. 2, establishing an RBF neural network model; initializing parameters in the model, and setting the learning rate and momentum factor of the network; training an RBF color space conversion model by using the data set constructed in the step 1:
setting the learning rate of the RBF neural network to be 0.05 and the momentum factor to be 0.85, and initializing the central vector, the base width vector and the connection weight of the RBF neural network. In the step 2, a gaussian function is selected as a basis function of the hidden layer node, and the formula (1) is as follows:
Figure BDA0003112222160000061
in the formula, xnFor the nth input sample, cn、σnThe output of the RBF neural network is obtained from equation (2), where the vectors are the basis function center and the basis width of the nth node of the hidden layer:
Figure BDA0003112222160000062
the RBF neural network training flow chart is shown in FIG. 3;
and 3, optimizing the central layer vector, the base width vector and the connection weight of the RBF neural network by using a whale optimization algorithm:
as shown in fig. 4, the specific process of step 3 is as follows:
step 3.1, initializing a whale optimization algorithm, setting the number N of whale populations and the maximum iteration time T of the algorithmmaxThe position of each individual whale in the whale population comprises a central vector, a base width vector and a connection weight in the RBF neural network, error statistics is carried out on the calculation result of each individual whale, the position information of the individual with the minimum error is used as the current optimal solution and is recorded as X*(t)；
Step 3.2, converting X in step 3.1*(t) reserving, and updating the whale population position, wherein the updating is shown as a formula (3):
Figure BDA0003112222160000063
wherein, A is 2a r-a, a is 2-2 (t/t)max) R is in [0,1 ]]Random vector of (d), tmaxRespectively representing the current iteration number and the maximum iteration number, X (t) representing the current position vector, b is a logarithmic spiralA spin constant, l is in [ -1,1 [ ]]P is a random number located in [0,1 ]]The calculation of the random numbers between, D and D' is shown in equations (4) and (5):
D＝|C·X*(t)-X(t)| (4)
D′＝|X*(t)-X(t)| (5)
wherein, C is 2 r;
to avoid the algorithm falling into the local optimum, a random walk is added to update the position, as shown in equations (6) and (7):
X(t+1)＝Xr(t)-A·D (6)
D＝|C·Xr(t)-X(t)| (7)
wherein, Xr(t) is a position vector of randomly selected whale individuals;
step 3.3, error calculation is carried out on the newly generated whale population, and the error calculation is carried out on the whale population and the X in the previous generation population*(t) comparing, replacing the individuals with larger errors in the previous generation with the individuals with smaller errors in the whale population of the new generation to generate new X*(t)；
Step 3.4, for the current X*(t) evaluating, and if the iteration number N or the conversion precision set in the step 3.1 is reached, taking the individual position as an optimal solution
Figure BDA0003112222160000071
Assigning the position information to a central vector, a base width vector and a connection weight of the RBF neural network respectively, and returning to the step 3.2 to continuously seek an optimal solution if the iteration times or the conversion precision are not met;
step 4, after the step 3 is completed, training the optimized RBF color space conversion model by using the training sample in the step 1 to obtain a final color space conversion model, and realizing the conversion from CMYK to LAB color space;
the specific process of step 4 is as follows:
step 4.1, the optimal solution in step 3 is obtained
Figure BDA0003112222160000072
Decomposed into a center vector, a base width vector sumConnecting the three parameters of the weight and inputting the three parameters into the RBF neural network to obtain an optimized RBF neural network;
step 4.2, taking any CMYK value as an input value of the RBF neural network, and outputting a converted LAB value by the RBF neural network after the conversion by the RBF neural network;
and 5, testing the final color space conversion model by using the test sample in the step 1, and checking the accuracy of the established color space conversion model.
Examples
The operating system of the example is Windows 10, MATLAB R2020a is used as simulation software, the sample source is PANTONE standard color card, 2300 color blocks of the color card are all numbered, then MATLAB is used for generating 700 random numbers in the range of [1,2300], the color blocks with the numbers corresponding to the 700 random numbers are used as training samples, then 80 random numbers are generated by the same method, and 80 color blocks are selected from the rest color blocks to be used as test samples. The CMYK values of the selected color patches are used as input values of the neural network, and the LAB values are used as output values of the neural network. Training the RBF neural network by using a training sample, after the training is finished, performing color difference evaluation on the model by using a test sample, calculating the evaluation between an LAB conversion value and a true value, and taking the color difference delta E as an evaluation standard, wherein the calculation formula is shown as a formula (13):
Figure BDA0003112222160000081
in the formula (I), the compound is shown in the specification,
Figure BDA0003112222160000082
for the value LAB value obtained by the model transformation,
Figure BDA0003112222160000083
standard LAB values for color blocks;
counting 100 color differences delta E of the test, as shown in FIG. 5, it can be seen that the CMYK to LAB color space conversion precision is higher;
the high-precision digital printing color space conversion method provided by the invention uses an RBF neural network optimized by a whale algorithm to complete the conversion from a CMYK color space to an LAB color space; and inputting any CMYK value into the trained conversion model, thereby realizing the conversion from CMYK to LAB color space. The working process is as follows: making a training sample and a testing sample; determining a neural network structure and learning rate and momentum factors of the neural network; taking CMYK values of color blocks in the sample as input values of the RBF neural network, and taking LAB values as output values of the RBF neural network; training the RBF neural network, and optimizing a central vector, a base width vector and a connection weight of the RBF neural network by a whale optimization algorithm; obtaining a final color space conversion model after training, and converting any CMYK value into a corresponding LAB value through the conversion model to realize a color space conversion function; the method optimizes the RBF neural network by using a whale optimization algorithm, solves the problems of insufficient precision and difficult key parameter selection of the RBF neural network, further improves the precision of color space conversion, and has good adaptivity.

Claims (5)

1. A high-precision digital printing color space conversion method is characterized by comprising the following steps:
step 1, collecting CMYK values and LAB values of color blocks from a standard color card to construct a data set;
step 2, establishing an RBF neural network model; initializing parameters in the model, setting learning rate and momentum factor of the network, and facilitating the data set constructed in the step 1 to train the RBF color space conversion model;
step 3, optimizing a central layer vector, a base width vector and a connection weight of the RBF neural network by using a whale optimization algorithm;
step 4, after the optimization in the step 3 is completed, training the optimized RBF color space conversion model by using the training sample in the step 1 to obtain a final color space conversion model, and realizing the conversion from CMYK to LAB color space;
and 5, testing the final color space conversion model by using the test sample in the step 1, and checking the accuracy of the established color space conversion model.
2. A high precision digital printing color space conversion method according to claim 1, wherein CMYK values are used as input of a neural network and LAB values are used as output of the neural network in step 1; the data set is divided into training samples and testing samples, the training samples are used for training the color space conversion model, and the testing samples are used for evaluating the conversion accuracy of the trained model.
3. The method for converting color space of digital printing with high precision according to claim 1, wherein the specific process of establishing the RBF neural network model in the step 2 is as follows:
setting the learning rate of the RBF neural network to be 0.05 and the momentum factor to be 0.85, and initializing a central vector, a base width vector and a connection weight of the RBF neural network; selecting a Gaussian function as a base function of a hidden layer node, wherein the formula is shown as (1):
Figure FDA0003112222150000021
in the formula, xnFor the nth input sample, cn、σnRespectively representing the base function center and the base function width of the nth node of the hidden layer;
the output of the RBF neural network can be obtained from equation (2):
Figure FDA0003112222150000022
4. the method for converting color space of digital printing with high precision according to claim 1, wherein the specific process optimized in the step 3 is implemented as the following steps:
step 3.1, initializing a whale optimization algorithm, setting the number of whale populations to be N-10, and setting the maximum iteration number T of the algorithmmax50, whaleThe position of each individual whale in the population comprises a central vector, a base width vector and a connection weight in the RBF neural network, error statistics is carried out on the calculation result of each individual whale, the position information of the individual with the minimum error is used as the current optimal solution and is recorded as X*(t)；
Step 3.2, converting X in step 3.1*(t) reserving, and updating the whale population position, wherein the updating is shown as a formula (3):
Figure FDA0003112222150000023
wherein A is 2a r-a, a is 2-2 (t/t)max) R is in [0,1 ]]Random vector of (d), tmaxRespectively representing the current iteration number and the maximum iteration number, X (t) representing the current position vector, b being a logarithmic spiral constant, l being located at [ -1,1]P is a random number located in [0,1 ]]The calculation of the random numbers between, D and D' is shown in equations (4) and (5):
D＝|C·X*(t)-X(t)| (4)
D′＝|X*(t)-X(t)| (5)
wherein, C is 2 r;
to avoid the algorithm falling into the local optimum, a random walk is added to update the position, as shown in equations (6) and (7):
X(t+1)＝Xr(t)-A·D (6)
D＝|C·Xr(t)-X(t)| (7)
in the formula, Xr(t) is a position vector of randomly selected whale individuals;
step 3.3, error calculation is carried out on the newly generated whale population, and the error calculation is carried out on the whale population and the X in the previous generation population*(t) comparing, and replacing the individuals with larger errors in the previous generation with the individuals with smaller errors in the whale population of the new generation to generate new X*(t)；
Step 3.4, for the current X*(t) evaluating, and if the iteration number N or the conversion precision set in the step 3.1 is reached, taking the individual position as an optimal solution
Figure FDA0003112222150000031
And assigning the position information to a central vector, a base width vector and a connection weight of the RBF neural network respectively, and returning to the step 3.2 to continuously seek the optimal solution if the iteration number N or the conversion precision is not met.
5. The method for converting color space of digital printing with high precision according to claim 1 or 4, wherein the specific training process in step 4 is as follows:
step 4.1, the optimal solution in step 3 is obtained
Figure FDA0003112222150000032
Decomposing the parameters into a center vector, a base width vector and a connection weight and inputting the parameters into the RBF neural network to obtain an optimized RBF neural network;
and 4.2, taking any CMYK value as an input value of the RBF neural network, and outputting the converted LAB value by the RBF neural network after the conversion by the RBF neural network.
CN202110654803.9A 2021-06-11 2021-06-11 High-precision digital printing color space conversion method Pending CN113409206A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110654803.9A CN113409206A (en) 2021-06-11 2021-06-11 High-precision digital printing color space conversion method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110654803.9A CN113409206A (en) 2021-06-11 2021-06-11 High-precision digital printing color space conversion method

Publications (1)

Publication Number Publication Date
CN113409206A true CN113409206A (en) 2021-09-17

Family

ID=77683647

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110654803.9A Pending CN113409206A (en) 2021-06-11 2021-06-11 High-precision digital printing color space conversion method

Country Status (1)

Country Link
CN (1) CN113409206A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116587759A (en) * 2023-07-13 2023-08-15 江苏龙达纺织科技有限公司 Visual and efficient digital printing color correction management method and system

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102110428A (en) * 2009-12-23 2011-06-29 新奥特（北京）视频技术有限公司 Method and device for converting color space from CMYK to RGB
CN106937018A (en) * 2017-01-12 2017-07-07 浙江大学 It is used for the color mapping method of textile inkjet printing and dyeing based on RBF neural
CN110377957A (en) * 2019-06-20 2019-10-25 浙江大学 A kind of overhead crane neural network modeling approach of whale search strategy grey wolf algorithm
CN110475043A (en) * 2019-07-31 2019-11-19 西安工程大学 A kind of conversion method of CMYK to Lab color space
CN112232493A (en) * 2020-11-09 2021-01-15 江苏科技大学 RBF neural network optimization method based on improved whale algorithm
CN112270397A (en) * 2020-10-26 2021-01-26 西安工程大学 Color space conversion method based on deep neural network
CN112578312A (en) * 2020-12-04 2021-03-30 燕山大学 Micro-grid fault diagnosis method based on whale algorithm optimization extreme learning machine

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102110428A (en) * 2009-12-23 2011-06-29 新奥特（北京）视频技术有限公司 Method and device for converting color space from CMYK to RGB
CN106937018A (en) * 2017-01-12 2017-07-07 浙江大学 It is used for the color mapping method of textile inkjet printing and dyeing based on RBF neural
CN110377957A (en) * 2019-06-20 2019-10-25 浙江大学 A kind of overhead crane neural network modeling approach of whale search strategy grey wolf algorithm
CN110475043A (en) * 2019-07-31 2019-11-19 西安工程大学 A kind of conversion method of CMYK to Lab color space
CN112270397A (en) * 2020-10-26 2021-01-26 西安工程大学 Color space conversion method based on deep neural network
CN112232493A (en) * 2020-11-09 2021-01-15 江苏科技大学 RBF neural network optimization method based on improved whale algorithm
CN112578312A (en) * 2020-12-04 2021-03-30 燕山大学 Micro-grid fault diagnosis method based on whale algorithm optimization extreme learning machine

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
ZHANG XIN等: "Research on Transformer Oil Kinematic Viscosity Detection Method Based on IWOA-RBF and Multi-frequency Ultrasonic Technology", 2020 3RD INTERNATIONAL CONFERENCE ON CONTROL AND ROBOTS (ICCR), pages 199 - 203 *
季鹏等: "基于ＲBF神经网络的船舶电站PID控制器研究", 《重庆理工大学学报( 自然科学)》 *
季鹏等: "基于ＲBF神经网络的船舶电站PID控制器研究", 《重庆理工大学学报( 自然科学)》, vol. 34, no. 2, 28 February 2020 (2020-02-28), pages 2 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116587759A (en) * 2023-07-13 2023-08-15 江苏龙达纺织科技有限公司 Visual and efficient digital printing color correction management method and system
CN116587759B (en) * 2023-07-13 2023-09-29 江苏龙达纺织科技有限公司 Visual and efficient digital printing color correction management method and system

Similar Documents

Publication Publication Date Title
US10171706B2 (en) Configuring an imaging system
US9047804B1 (en) Lexical-based processing of color images
CN108830912A (en) A kind of interactive grayscale image color method of depth characteristic confrontation type study
US6381036B1 (en) Method for determining printer colorant levels
CN110223359A (en) It is a kind of that color model and its construction method and application on the dual-stage polygamy colo(u)r streak original text of network are fought based on generation
CN110475043B (en) Method for converting CMYK to Lab color space
CN1946563A (en) N-ink color gamut construction
EP3135026B1 (en) Generation of color mapping representative of a color halftone processing pipeline
CN102110428B (en) Method and device for converting color space from CMYK to RGB
CN109816714B (en) Point cloud object type identification method based on three-dimensional convolutional neural network
CN105184370B (en) A kind of river mouth river basin water quality evaluation method based on SOM sorting techniques
CN108021881A (en) A kind of skin color segmentation method, apparatus and storage medium
CN113409206A (en) High-precision digital printing color space conversion method
CN112270397B (en) Color space conversion method based on deep neural network
CN115713634A (en) Color collocation evaluation method combining similarity measurement and visual perception
CN113807194A (en) Enhanced power transmission line fault image identification method
WO2021092807A1 (en) Self-adaptive pre-coding model training method, self-adaptive pre-coding method and base station
CN113408798B (en) Barrier-free VR teaching resource color optimization method for people with abnormal color vision
CN114998454B (en) Dynamic multi-color code generation and analysis method
Yanzhe A novel algorithm based on improved BP neural network and its application in color management
US11321026B2 (en) Color value prediction system and generation method for color value predictor
CN112183546B (en) Image segmentation method based on spatial nearest neighbor with weight constraint
KR102474170B1 (en) Apparatus and method for generating synthetic data for model training
JPH06339016A (en) Automatic converting method and device for color reproduction area and automatic color converter
CN112651932A (en) Method for calibrating metamerism maximum gray level equivalent substitution range

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