Control method for controlling color generation equipment for digital printing to generate color

Abstract

The invention relates to the technical field of color generation control, in particular to a control method for controlling a color generation device for digital printing to generate color, which comprises the following steps: acquiring an image to be subjected to color generation, acquiring RGB color data of each pixel point in the image to be subjected to color generation, and converting the RGB color data of each pixel point in the image to be subjected to color generation into Lab color data; and converting the Lab color data of each pixel point in the to-be-colored image into CMYKW color data based on a pre-trained color conversion model. The color generation device can be controlled according to the ink formula sequence in the image, the needed ink formula is accurately provided for each pixel point, optimization operation is carried out according to formula changes, the color generation operation of the pixel points one by one is suitable for most complex images, jumping operation can be achieved in the repeated color area, the printing speed is increased, and the printing efficiency is improved. And the printing ink formula of each pixel point is accurately controlled, so that excessive printing ink waste can be avoided, and the cost is saved.

Description

Technical Field

[0001] The present invention relates to the technical field of color rendering control, and in particular to a control method for controlling a color rendering device for digital printing to render color. Background Art

[0002] Color rendering devices refer to devices used to generate, display or print colors. These devices can convert raw image data into colors that we can see. The core function of color rendering devices is to ensure accurate reproduction of image colors. Digital printing is a modern printing technology. It is different from traditional printing methods. It uses digitized image data to directly drive the printer for printing.

[0003] In traditional methods, color conversion often relies on simple color models and lacks precise color matching mechanisms. The conversion between RGB and CMYK color gamuts is not precise enough, which may lead to color deviation, especially when the color capability of the printer is insufficient, the color reproduction is not as expected; and traditional methods usually rely on manual ink mixing or color matching based on simple ink libraries. Although some devices have built-in color libraries, these libraries are usually not fully optimized, and when encountering special or complex colors, manual intervention is often required; and in the traditional printing process, the ink formula of each pixel is usually calculated and ejected separately without precise control of the use of ink. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the disadvantages of the above-mentioned prior art and provide a control method for controlling a color rendering device for digital printing to render color.

[0005] The technical solution adopted to solve the above technical problem is: a control method for controlling a color rendering device for digital printing to render color, comprising:

[0006] Acquire an image to be rendered, acquire RGB color data of each pixel in the image to be rendered, and convert the RGB color data of each pixel in the image to be rendered into Lab color data;

[0007] Convert the Lab color data of each pixel in the image to be rendered into CMYKW color data based on a pre-trained color conversion model;

[0008] Matching the CMYKW color data of each pixel in the image to be rendered with a preset ink database, and if the CMYKW color data matches the preset ink database, obtaining a matching ink formula corresponding to each pixel in the image to be rendered;

[0009] If the CMYKW color data does not match the preset ink database, ink color matching is performed on the CMYKW color data of each pixel in the unmatched image to be rendered based on a pre-trained ink color matching model to obtain an ink formula corresponding to each pixel in the unmatched image to be rendered;

[0010] Sorting the ink formula corresponding to each pixel in the image to be rendered according to the position of the pixel in the image to be rendered, so as to obtain an ink formula sequence corresponding to the image to be rendered;

[0011] The color rendering device is controlled to render color based on the ink formula sequence corresponding to the image to be rendered.

[0012] Preferably, converting the RGB color data of each pixel in the image to be rendered into Lab color data comprises:

[0013] The RGB color data of each pixel in the image to be rendered is linearized to obtain the linear RGB color data of each pixel in the image to be rendered, wherein the linearization formula of the RGB color data is as follows:

[0014] ;

[0015] in, Represents linear single-channel color data, Represents single-channel color data;

[0016] Calculate the three-dimensional coordinates of the linear RGB color data of each pixel in the image to be rendered, wherein the calculation formula of the three-dimensional coordinates of the linear RGB color data is as follows:

[0017] ;

[0018] in, Three-dimensional coordinates representing linear RGB color data, , and Represents R, G and B channel color data;

[0019] The linear RGB color data is converted into Lab color data based on the three-dimensional coordinates of the linear RGB color data of each pixel in the image to be rendered.

[0020] Preferably, the color conversion model adopts an improved BP neural network, and the BP neural network is optimized based on an optimization algorithm to obtain an improved BP neural network, and the R, G and B channel color data of the RGB color data of each pixel in the image to be rendered are used as input data, and the color conversion model outputs the C, M, Y, K and W channel color data of the CMYKW color data of each pixel in the image to be rendered.

[0021] Preferably, the BP neural network is optimized based on an optimization algorithm to obtain an improved BP neural network, including:

[0022] The individual population is randomly initialized, and the position of each individual represents the combination of each weight coefficient and bias coefficient of the BP neural network;

[0023] For each combination of weights and bias values ​​represented by an individual, the output of the BP neural network is calculated by forward propagation, and the fitness of the individual is calculated according to the error between the output of the BP neural network and the actual target, wherein the fitness is the mean square error;

[0024] Randomly generate a position update coefficient, where the position update coefficient is a constant ranging from 0 to 1, determine whether the current number of iterations is greater than a preset number of iterations threshold, and determine whether the position update coefficient is greater than 0.5, if the current number of iterations is greater than the preset number of iterations and the position update coefficient is greater than 0.5, then update the position of the individual using a first position formula, if the current number of iterations is greater than the preset number of iterations and the position update coefficient is not greater than 0.5, then update the position of the individual using a second position formula, if the current number of iterations is not greater than the preset number of iterations and the position update coefficient is greater than 0.5, then update the position of the individual using a third position formula, if the current number of iterations is not greater than the preset number of iterations and the position update coefficient is not greater than 0.5, then update the position of the individual using a fourth position formula;

[0025] The optimal individual is obtained based on the updated positions of all individuals, and it is determined whether the training error reaches the convergence value or the number of iterations reaches the maximum value. If the training error reaches the convergence value or the number of iterations reaches the maximum value, the iteration ends and the weight and bias value combination of the neural network is output, otherwise the number of iterations is increased by 1.

[0026] Preferably, the individual initialization formula is as follows:

[0027] ;

[0028] in, Indicates The position of the individual at the iteration, represents the lower bound of the search space, represents the upper bound of the search space, Indicates the generated random number between 0 and 1;

[0029] The expression of the first position formula is as follows:

[0030] ;

[0031] in, Indicates The first position of the individual at the iteration, represents the position of the current optimal individual, Indicates the current iteration number, represents the total number of iterations, Indicates The average position of the current individual at the iteration, represents the position update coefficient;

[0032] The expression of the second position formula is as follows:

[0033] ;

[0034] in, Indicates The second position of the individual at the iteration, represents the Levy flight distribution function, Indicates The worst position of the current individual at the iteration;

[0035] The expression of the third position formula is as follows:

[0036] ;

[0037] in, Indicates The third position of the individual at the iteration, and represents a development adjustment parameter fixed between 0 and 1;

[0038] The expression of the fourth position formula is as follows:

[0039] ;

[0040] in, Indicates The fourth position of the individual at the iteration, represents the quality function of the balanced search strategy, represents the motion coefficient, represents the flight slope, and , Represents a random number, ,and .

[0041] Preferably, the ink color matching model includes an input layer, a hidden layer and an output layer, wherein the input layer is used to input C, M, Y, K and W channel color data, the number of neurons in the input layer is 5, the number of layers of the hidden layer is 1, the number of neurons in the hidden layer is 20, and the output layer is used to output the types and concentrations of each ink, and the number of neurons in the output layer is 2.

[0042] Preferably, the output expression of the output layer is as follows:

[0043] ;

[0044] in, Indicates The output of a neuron, represents the connection weights between each neuron node in the hidden layer and each neuron node in the output layer, represents the output of the hidden layer, represents the bias of each neuron node in the output layer, represents the Purelin function, and , represents the tanh function, Represents the connection weights between each neuron node in the input layer and each neuron node in the hidden layer. represents the input of the input layer, Represents the bias of each neuron node in the hidden layer.

[0045] Preferably, the ink formula corresponding to each pixel in the image to be rendered is sorted according to the position of the pixel in the image to be rendered to obtain the ink formula sequence corresponding to the image to be rendered, including:

[0046] Recording the position of each pixel in the image to be rendered in color;

[0047] Based on the position of the pixel point in the image to be rendered, the ink formula corresponding to each pixel point is arranged in position order to obtain an ink formula sorting sequence;

[0048] Sliding a preset sliding window in the ink formula sorting sequence to determine whether the same ink formula exists in the sliding window;

[0049] If the same ink formula exists in the sliding window, the ink formula sorting sequence is updated to obtain an ink formula sequence;

[0050] If the same ink formula does not exist in the sliding window, the sliding window continues to slide in the ink formula sorting sequence.

[0051] Preferably, if the same ink formula exists in the sliding window, the ink formula sorting sequence is updated to obtain an ink formula sequence, including:

[0052] Overlapping the same ink formula in the sliding window;

[0053] The corresponding pixel points in the image to be rendered are added to the same ink formula to obtain an ink formula sequence.

[0054] Preferably, controlling the color rendering device to render color based on the ink formula sequence corresponding to the image to be rendered includes:

[0055] Traversing the ink formula sequence corresponding to the image to be rendered;

[0056] When the overlapping part in the ink formula sequence corresponding to the image to be rendered is not traversed, the rendering device is controlled pixel by pixel to render the color;

[0057] When the overlapping part in the ink formula sequence corresponding to the image to be rendered is traversed, the pixel points between the overlapping parts are jumped to control the rendering device to render the color.

[0058] The beneficial effects of the present invention are as follows: (1) The present invention can achieve more accurate color reproduction by converting the RGB color data of the image to be rendered into Lab color data, and then converting the Lab color data into CMYKW color data based on a pre-trained color conversion model, and using a preset ink database to match the CMYKW color data to ensure the accuracy of the ink formula. In particular, for common colors, a formula that matches the ink database can be directly obtained, thereby reducing color matching errors; (2) The present invention uses a pre-trained ink color matching model to perform intelligent color matching for unmatched color data, which can not only improve the accuracy of color matching, but also significantly reduce manual intervention, especially in In the case of complex or special colors, repeated manual mixing processes are avoided, and by sorting the ink formula sequence and deciding whether to skip color rendering based on whether the ink formulas of adjacent pixels are consistent, redundant operations of repeated color rendering can be avoided, which can greatly improve efficiency; (3) The present invention can control the color rendering device according to the ink formula sequence in the image, accurately provide the required ink formula for each pixel, and optimize the operation according to the change of the formula. The color rendering operation of each pixel is suitable for most complex images, and skipping operations can be implemented in repeated color areas, thereby improving printing speed. By accurately controlling the ink formula of each pixel, excessive ink waste can be avoided, saving costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A schematic flow chart of the steps of an overall method in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] Embodiment 1, as Figure 1 As shown, the present invention proposes a control method for controlling a color rendering device for digital printing to render color, comprising:

[0061] S1, obtaining an image to be rendered, obtaining RGB color data of each pixel in the image to be rendered, and converting the RGB color data of each pixel in the image to be rendered into Lab color data;

[0062] S2, converting the Lab color data of each pixel in the image to be rendered into CMYKW color data based on a pre-trained color conversion model;

[0063] S3, matching the CMYKW color data of each pixel in the image to be rendered with a preset ink database, if the CMYKW color data matches the preset ink database, then obtaining the ink formula corresponding to each pixel in the image to be rendered;

[0064] S4. If the CMYKW color data does not match the preset ink database, ink color matching is performed on the CMYKW color data of each pixel in the unmatched image to be rendered based on a pre-trained ink color matching model to obtain an ink formula corresponding to each pixel in the unmatched image to be rendered;

[0065] S5, sorting the ink formula corresponding to each pixel in the image to be rendered according to the position of the pixel in the image to be rendered, so as to obtain an ink formula sequence corresponding to the image to be rendered;

[0066] S6. Control the color rendering device to render color based on the ink formula sequence corresponding to the image to be rendered.

[0067] In the present invention, RGB (Red, Green, Blue) is a color model commonly used in computer graphics and display devices (such as monitors, televisions, etc.). The RGB model is based on three basic colors: red, green and blue. Various colors can be synthesized by adjusting the intensity of these three colors; the Lab color space is a color model related to the color perception of the human eye, L represents the brightness of the color, ranging from 0 (black) to 100 (white), a represents the red-green axis, negative values ​​represent green, and positive values ​​represent red, b represents the yellow-blue axis, negative values ​​represent blue, and positive values ​​represent yellow; CMYKW is an extended color model that includes the traditional CMYK model, C is complementary to the red in RGB, M is complementary to the green in RGB, Y is complementary to the blue in RGB, K is used to enhance the depth and contrast of the image and avoid the color mixing problem caused by using a combination of pure colors, and W is added in some printing or image processing to reflect the white ink or background color in the printing press.

[0068] Embodiment 2, a control method for controlling a color rendering device for digital printing to render color is proposed by the present invention. Compared with embodiment 1, this embodiment further includes: converting the RGB color data of each pixel in the image to be rendered into Lab color data, including:

[0069] A1. Linearize the RGB color data of each pixel in the image to be rendered to obtain the linear RGB color data of each pixel in the image to be rendered, wherein the linearization formula of the RGB color data is as follows:

[0070] ;

[0071] in, Represents linear single-channel color data, Represents single-channel color data;

[0072] A2. Calculate the three-dimensional coordinates of the linear RGB color data of each pixel in the image to be rendered, wherein the calculation formula of the three-dimensional coordinates of the linear RGB color data is as follows:

[0073] ;

[0074] in, Three-dimensional coordinates representing linear RGB color data, , and Represents R, G and B channel color data;

[0075] A3. Convert the linear RGB color data into Lab color data based on the three-dimensional coordinates of the linear RGB color data of each pixel in the image to be rendered.

[0076] In an optional embodiment, the color conversion model adopts an improved BP neural network, and optimizes the BP neural network based on an optimization algorithm to obtain an improved BP neural network. The R, G, and B channel color data of the RGB color data of each pixel in the image to be rendered are used as input data, and the color conversion model outputs the C, M, Y, K, and W channel color data of the CMYKW color data of each pixel in the image to be rendered.

[0077] BP neural network (back propagation neural network) is a common artificial neural network model used to solve various prediction and classification problems. Its core idea is to minimize the output error by continuously adjusting the weights of the neural network; the optimization algorithm is used to find the best solution to the problem, usually by adjusting the model parameters (such as the weights in the neural network) to optimize the performance of the model.

[0078] In an optional embodiment, the BP neural network is optimized based on an optimization algorithm to obtain an improved BP neural network, including:

[0079] B1. Randomly initialize the individual population, and the position of each individual represents the combination of each weight coefficient and bias coefficient of the BP neural network;

[0080] B2. For each combination of weights and bias values ​​represented by an individual, the output of the BP neural network is calculated through forward propagation, and the fitness of the individual is calculated based on the error between the output of the BP neural network and the actual target, where the fitness is the mean square error;

[0081] B3. Randomly generate a position update coefficient, which is a constant ranging from 0 to 1. Determine whether the current number of iterations is greater than a preset number of iterations threshold, and determine whether the position update coefficient is greater than 0.5. If the current number of iterations is greater than the preset number of iterations and the position update coefficient is greater than 0.5, then update the position of the individual using the first position formula. If the current number of iterations is greater than the preset number of iterations and the position update coefficient is not greater than 0.5, then update the position of the individual using the second position formula. If the current number of iterations is not greater than the preset number of iterations and the position update coefficient is greater than 0.5, then update the position of the individual using the third position formula. If the current number of iterations is not greater than the preset number of iterations and the position update coefficient is not greater than 0.5, then update the position of the individual using the fourth position formula.

[0082] B4. Obtain the optimal individual based on the updated positions of all individuals, and determine whether the training error reaches the convergence value or the number of iterations reaches the maximum value. If the training error reaches the convergence value or the number of iterations reaches the maximum value, the iteration ends and outputs the weight and bias value combination of the neural network, otherwise the number of iterations increases by 1.

[0083] In an optional embodiment, the individual initialization formula is as follows:

[0084] ;

[0085] in, Indicates The position of the individual at the iteration, represents the lower bound of the search space, represents the upper bound of the search space, Represents a random number generated between 0 and 1.

[0086] The expression of the first position formula is as follows:

[0087] ;

[0088] in, Indicates The first position of the individual at the iteration, represents the position of the current optimal individual, Indicates the current iteration number, represents the total number of iterations, Indicates The average position of the current individual at the iteration, Represents the position update coefficient.

[0089] The expression of the second position formula is as follows:

[0090] ;

[0091] in, Indicates The second position of the individual at the iteration, represents the Levy flight distribution function, Indicates The worst position of the current individual at the iteration.

[0092] The expression of the third position formula is as follows:

[0093] ;

[0094] in, Indicates The third position of the individual at the iteration, and Represents a development adjustment parameter fixed between 0 and 1.

[0095] The expression of the fourth position formula is as follows:

[0096] ;

[0097] in, Indicates The fourth position of the individual at the iteration, represents the quality function of the balanced search strategy, represents the motion coefficient, represents the flight slope, and , Represents a random number, ,and .

[0098] The initial value of is 0.

[0099] In an optional embodiment, the ink color matching model includes an input layer, a hidden layer and an output layer, wherein the input layer is used to input C, M, Y, K and W channel color data, the number of neurons in the input layer is 5, the number of hidden layers is 1, the number of neurons in the hidden layer is 20, and the output layer is used to output the types and concentrations of each ink, the number of neurons in the output layer is 2.

[0100] In an optional embodiment, the output expression of the output layer is as follows:

[0101] ;

[0102] in, Indicates The output of a neuron, represents the connection weights between each neuron node in the hidden layer and each neuron node in the output layer, represents the output of the hidden layer, represents the bias of each neuron node in the output layer, represents the Purelin function, and , represents the tanh function, Represents the connection weights between each neuron node in the input layer and each neuron node in the hidden layer. represents the input of the input layer, Represents the bias of each neuron node in the hidden layer.

[0103] In an optional embodiment, the ink formula corresponding to each pixel in the image to be rendered is sorted according to the position of the pixel in the image to be rendered to obtain an ink formula sequence corresponding to the image to be rendered, including:

[0104] C1. Record the position of each pixel in the image to be rendered in the image to be rendered;

[0105] C2, based on the position of the pixel in the image to be rendered, arranging the ink formula corresponding to each pixel in position order to obtain an ink formula sorting sequence;

[0106] C3, sliding in the ink formula sorting sequence based on a preset sliding window, and determining whether there is an identical ink formula in the sliding window;

[0107] C4. If the same ink formula exists in the sliding window, the ink formula sorting sequence is updated to obtain an ink formula sequence;

[0108] C5. If the same ink formula does not exist in the sliding window, continue to slide the sliding window in the ink formula sorting sequence.

[0109] An image is composed of pixels, each of which has a specific coordinate. For example, if the resolution of an image is 1000x1000 pixels, each pixel will have its specific (x, y) coordinate. For each pixel, according to this position, the ink formula of each pixel (such as the value of C, M, Y, K, W) is arranged in the order of the image. Usually, the sorting method can be from top to bottom, from left to right, or other preset order, depending on the actual application requirements. For example, for the first pixel (0, 0) in the image, assuming its ink formula is (C1, M1, Y1, K1, W1), then the order of the ink formula is recorded. For the second pixel (0, 1), assuming its ink formula is (C2, M2, Y2, K2, W2), and then continue to arrange in this way until the ink formulas of all pixels are recorded; Sliding window is a common algorithmic technique used to dynamically view a fixed-size subset in a piece of data, define a fixed-size window (for example, the window size is 3), and slide this window in the ink formula sequence to check whether the same ink formula exists in the window; Suppose we have an ink formula sequence: [(C1, M1, Y1, K1, W1), (C2, M2,Y2, K2, W2), (C1, M1, Y1, K1, W1), (C3, M3, Y3, K3, W3)], we use a sliding window of size 2, starting from the first item of the sequence, to check whether there is a repeated ink formula in the window; When the window slides to (C1, M1, Y1, K1, W1) and (C2, M2, Y2, K2, W2), the two formulas are different, but when the window slides to (C1, M1, Y1, K1, W1) and (C1, M1, When the same ink formula is found in the sliding window, we need to update the ink formula sequence, which may be to remove the duplicate formula or rearrange it according to some rules. If (C1, M1, Y1, K1, W1) and (C1, M1, Y1, K1, W1) are found to be duplicates, then we can merge the second formula.

[0110] In an optional embodiment, if the same ink formula exists in the sliding window, the ink formula sorting sequence is updated to obtain an ink formula sequence, including:

[0111] D1. Overlap the same ink formula in the sliding window;

[0112] D2. Add corresponding pixel points in the image to be rendered to the same ink formula to obtain an ink formula sequence.

[0113] In an optional embodiment, controlling a color rendering device to render color based on an ink formula sequence corresponding to an image to be rendered includes:

[0114] E1, traverse the ink formula sequence corresponding to the image to be rendered;

[0115] E2. When the overlapping part in the ink formula sequence corresponding to the image to be rendered is not traversed, the rendering device is controlled pixel by pixel to render the color;

[0116] E3. When traversing to the overlapping part in the ink formula sequence corresponding to the image to be rendered, the pixel points between the overlapping parts are jumped to control the rendering device to render the color.

[0117] If the ink formula of the current pixel is different from that of the next pixel in the ink formula sequence, the color rendering device will perform color rendering operations pixel by pixel. If the ink formula of a pixel in the image is (C1, M1, Y1, K1, W1), and the ink formula of the next pixel is (C2, M2, Y2, K2, W2), the two are different, so the color rendering device will operate according to the ink formula of each pixel separately; if the ink formulas of some pixels in the ink formula sequence are the same (that is, two consecutive pixels have the same ink formula), then in this area, there is no need to perform color rendering operations pixel by pixel. Instead, these uncolored pixels can be skipped and the next position with the same ink formula can be directly jumped to continue controlling the color rendering device to work.

[0118] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A control method for controlling a color rendering device for digital printing to render color, characterized in that: include: Acquire an image to be rendered, acquire RGB color data of each pixel in the image to be rendered, and convert the RGB color data of each pixel in the image to be rendered into Lab color data; Convert the Lab color data of each pixel in the image to be rendered into CMYKW color data based on a pre-trained color conversion model; Matching the CMYKW color data of each pixel in the image to be rendered with a preset ink database, and if the CMYKW color data matches the preset ink database, obtaining a matching ink formula corresponding to each pixel in the image to be rendered; If the CMYKW color data does not match the preset ink database, ink color matching is performed on the CMYKW color data of each pixel in the unmatched image to be rendered based on a pre-trained ink color matching model to obtain an ink formula corresponding to each pixel in the unmatched image to be rendered; Sorting the ink formula corresponding to each pixel in the image to be rendered according to the position of the pixel in the image to be rendered, so as to obtain an ink formula sequence corresponding to the image to be rendered; The color rendering device is controlled to render color based on the ink formula sequence corresponding to the image to be rendered.

2. A control method for controlling a color rendering device for digital printing to render color according to claim 1, characterized in that: Converting the RGB color data of each pixel in the image to be rendered into Lab color data includes: The RGB color data of each pixel in the image to be rendered is linearized to obtain the linear RGB color data of each pixel in the image to be rendered, wherein the linearization formula of the RGB color data is as follows: ； in, Represents linear single-channel color data, Represents single-channel color data; Calculate the three-dimensional coordinates of the linear RGB color data of each pixel in the image to be rendered, wherein the calculation formula of the three-dimensional coordinates of the linear RGB color data is as follows: ； in, Three-dimensional coordinates representing linear RGB color data, , and Represents R, G and B channel color data; The linear RGB color data is converted into Lab color data based on the three-dimensional coordinates of the linear RGB color data of each pixel in the image to be rendered.

3. A control method for controlling a color rendering device for digital printing to render color according to claim 1, characterized in that: The color conversion model adopts an improved BP neural network, and the BP neural network is optimized based on an optimization algorithm to obtain an improved BP neural network. The R, G and B channel color data of the RGB color data of each pixel in the image to be rendered are used as input data, and the color conversion model outputs the C, M, Y, K and W channel color data of the CMYKW color data of each pixel in the image to be rendered.

4. A control method for controlling a color rendering device for digital printing to render color according to claim 3, characterized in that: The BP neural network is optimized based on an optimization algorithm to obtain an improved BP neural network, including: The individual population is randomly initialized, and the position of each individual represents the combination of each weight coefficient and bias coefficient of the BP neural network; For each combination of weights and bias values ​​represented by an individual, the output of the BP neural network is calculated by forward propagation, and the fitness of the individual is calculated according to the error between the output of the BP neural network and the actual target, wherein the fitness is the mean square error; Randomly generate a position update coefficient, where the position update coefficient is a constant ranging from 0 to 1, determine whether the current number of iterations is greater than a preset number of iterations threshold, and determine whether the position update coefficient is greater than 0.5, if the current number of iterations is greater than the preset number of iterations and the position update coefficient is greater than 0.5, then update the position of the individual using a first position formula, if the current number of iterations is greater than the preset number of iterations and the position update coefficient is not greater than 0.5, then update the position of the individual using a second position formula, if the current number of iterations is not greater than the preset number of iterations and the position update coefficient is greater than 0.5, then update the position of the individual using a third position formula, if the current number of iterations is not greater than the preset number of iterations and the position update coefficient is not greater than 0.5, then update the position of the individual using a fourth position formula; The optimal individual is obtained based on the updated positions of all individuals, and it is determined whether the training error reaches the convergence value or the number of iterations reaches the maximum value. If the training error reaches the convergence value or the number of iterations reaches the maximum value, the iteration ends and the weight and bias value combination of the neural network is output, otherwise the number of iterations is increased by 1.

5. A control method for controlling a color rendering device for digital printing to render color according to claim 4, characterized in that: The initialization formula of the individual is as follows: ； in, Indicates The position of the individual at the iteration, represents the lower bound of the search space, represents the upper bound of the search space, Indicates the generated random number between 0 and 1; The expression of the first position formula is as follows: ； in, Indicates The first position of the individual at the iteration, represents the position of the current optimal individual, Indicates the current iteration number, represents the total number of iterations, Indicates The average position of the current individual at the iteration, represents the position update coefficient; The expression of the second position formula is as follows: ； in, Indicates The second position of the individual at the iteration, represents the Levy flight distribution function, Indicates The worst position of the current individual at the iteration; The expression of the third position formula is as follows: ； in, Indicates The third position of the individual at the iteration, and represents a development adjustment parameter fixed between 0 and 1; The expression of the fourth position formula is as follows: ； in, Indicates The fourth position of the individual at the iteration, represents the quality function of the balanced search strategy, represents the motion coefficient, represents the flight slope, and , Represents a random number, ,and .

6. A control method for controlling a color rendering device for digital printing to render color according to claim 1, characterized in that: The ink color matching model includes an input layer, a hidden layer and an output layer, wherein the input layer is used to input C, M, Y, K and W channel color data, the number of neurons in the input layer is 5, the number of layers in the hidden layer is 1, the number of neurons in the hidden layer is 20, and the output layer is used to output the types and concentrations of various inks, and the number of neurons in the output layer is 2.

7. A control method for controlling a color rendering device for digital printing to render color according to claim 6, characterized in that: The output expression of the output layer is as follows: ； in, Indicates The output of a neuron, represents the connection weights between each neuron node in the hidden layer and each neuron node in the output layer, represents the output of the hidden layer, represents the bias of each neuron node in the output layer, represents the Purelin function, and , represents the tanh function, Represents the connection weights between each neuron node in the input layer and each neuron node in the hidden layer. represents the input of the input layer, Represents the bias of each neuron node in the hidden layer.

8. A control method for controlling a color rendering device for digital printing to render color according to claim 1, characterized in that: Sorting the ink formula corresponding to each pixel in the image to be rendered according to the position of the pixel in the image to be rendered to obtain an ink formula sequence corresponding to the image to be rendered, including: Recording the position of each pixel in the image to be rendered in color; Based on the position of the pixel point in the image to be rendered, the ink formula corresponding to each pixel point is arranged in position order to obtain an ink formula sorting sequence; Sliding a preset sliding window in the ink formula sorting sequence to determine whether the same ink formula exists in the sliding window; If the same ink formula exists in the sliding window, the ink formula sorting sequence is updated to obtain an ink formula sequence; If the same ink formula does not exist in the sliding window, the sliding window continues to slide in the ink formula sorting sequence.

9. A control method for controlling a color rendering device for digital printing to render color according to claim 8, characterized in that: If the same ink formula exists in the sliding window, the ink formula sorting sequence is updated to obtain an ink formula sequence, including: Overlapping the same ink formula in the sliding window; The corresponding pixel points in the image to be rendered are added to the same ink formula to obtain an ink formula sequence.

10. A control method for controlling a color rendering device for digital printing to render color according to claim 9, characterized in that: Controlling the color rendering device to render color based on the ink formula sequence corresponding to the image to be rendered includes: Traversing the ink formula sequence corresponding to the image to be rendered; When the overlapping part in the ink formula sequence corresponding to the image to be rendered is not traversed, the rendering device is controlled pixel by pixel to render the color; When the overlapping part in the ink formula sequence corresponding to the image to be rendered is traversed, the pixel points between the overlapping parts are jumped to control the rendering device to render the color.

Images