Scanning image real-time processing method and system based on multi-source error analysis

Abstract

The invention discloses a scanned image real-time processing method and system based on multi-source error analysis, and relates to the technical field of image processing. Error generation reasons are classified into equipment factors, environment factors and operation factors, influence factors of all the factors are obtained, calibration characteristic parameters of grid areas of calibration images are extracted, scanning images of scanning equipment on the calibration images under the influence of any factor are obtained through a control variable method, and the calibration accuracy is improved. Scanning feature parameters at the corresponding grid areas in the scanned image are obtained, the scanning feature parameters under the influence of only any factor are used as input, the calibration feature parameters are used as output to construct a BP neural network, the to-be-scanned image is segmented according to the size proportion k of the to-be-scanned image and the calibration image, and the to-be-scanned image is obtained; and obtaining a correction value of the scanning feature parameter by using the constructed BP neural network, and performing weighting processing on the correction value to obtain a corrected output value of the scanning feature parameter of the to-be-scanned picture.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a real-time processing method and system for scanning images based on multi-source error analysis. Background Art

[0002] A flatbed scanner is a commonly used scanning device. During the scanning process of a document, the transmission mechanism of the scanner drives the scanning unit to slide along the vertically distributed X-axis and Y-axis guide axes, thereby scanning the object.

[0003] During the dynamic scanning process, the scanning unit will be affected by the vibration factors generated by the mechanical movement, resulting in blurred and distorted scanned images and insufficient clarity. In addition, after long-term use, the mechanical friction parts of the transmission structure of the scanning mechanism will wear out, causing the scanner to shake, resulting in skewed scanned images, further reducing the quality of the scanned images. At present, in order to solve the above-mentioned technical problems, the quality of the scanned images can only be improved to the greatest extent by regular inspection and maintenance of the scanning equipment, but the effect is not obvious. There is no very effective solution to the above-mentioned technical problems in the prior art. To this end, we propose a real-time processing method and system for scanned images based on multi-source error analysis. Summary of the invention

[0004] The main purpose of the present invention is to provide a real-time processing method and system for scanned images based on multi-source error analysis, which can effectively solve the problems in the background technology.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is:

[0006] The real-time processing method of scanning images based on multi-source error analysis includes:

[0007] Analyze the causes of scan image errors, classify the causes of errors into equipment factors, environmental factors, and operational factors based on the analysis results, and obtain the influencing factors k of each factor respectively. 1u , k 2v , k 3w , k 1u Expressed as the u-th influencing factor of the equipment factor; k 2v Expressed as the vth influencing factor of environmental factors; k 3w It is expressed as the w-th influencing factor of environmental factors; u, v, and w are all positive integers;

[0008] The influencing factors of the equipment factors include at least one of the wear amount of the guide shaft in the X-axis direction, the wear amount of the guide shaft in the Y-axis direction, the scanning movement speed in the X-axis direction, the scanning movement speed in the Y-axis direction, the vibration amplitude of the guide shaft in the X-axis direction during scanning, and the vibration amplitude of the guide shaft in the Y-axis direction during scanning;

[0009] The influencing factors of the environmental factors include at least one of the equipment operating temperature, environmental humidity, background light source brightness, and ambient light source brightness;

[0010] The operation factor includes at least one of the surface flatness of the object to be scanned, the overall size of the object to be scanned, the time interval since the last maintenance, the resolution setting value, and the equipment operation time;

[0011] Set a calibration image, divide the calibration image into several grids with a step size of λ, and extract the calibration feature parameter Cp of the grid area of ​​the calibration image ij , using the control variable method to obtain the scanned image of the calibration image by the scanning device under the influence of any of the factors, and obtain the scanning feature parameters of the corresponding grid area in the scanned image Cp ij It is expressed as the jth calibration feature parameter of the i-th grid area; It is represented as the j-th scanning characteristic parameter of the i-th grid area under the influence of the device factor only; It is represented by the j-th scanning characteristic parameter of the i-th grid area under the influence of the environmental factors only; It is represented by the j-th scanning characteristic parameter of the i-th grid area under the influence of the operating factor only; i=1,2,...,n; n is the number of grid areas; j=1,2,...,m; m is the type of characteristic parameter;

[0012] The characteristic parameters include at least one of geometric characteristic parameters, color characteristic parameters, texture characteristic parameters, resolution, and number of pixels; wherein the geometric characteristic parameters include at least one of perimeter, direction, aspect ratio, and eccentricity;

[0013] The scanning characteristic parameters only exist under the influence of the device factors As input, the calibration characteristic parameter Cp ij The first BP neural network is constructed for output; the scanning characteristic parameters only exist under the influence of the environmental factors. As input, the calibration characteristic parameter Cp ij Construct a second BP neural network for output; use the scanning characteristic parameters under the influence of the operating factors only As input, the calibration characteristic parameter Cp ij Constructing a third BP neural network for output; training the established neural network, and adjusting the number of hidden layers of the neural network model to an accuracy rate not lower than an expected value;

[0014] The calculation formula of the expected value is:

[0015]

[0016] Where E(Y) represents the expected value; N is the input sample size of the neural network; f(X i ) is represented as the output function of the neural network; X i Represented as an output sample of the neural network;

[0017] According to the size ratio k between the image to be scanned and the calibration image, the image to be scanned is divided into a number of grids with a step length of ε, ε=k×λ, and the step λ and its size satisfy the following relationship, specifically: In the formula, l x It is expressed as the length of the calibration image in the X-axis direction; l y It is expressed as the length of the calibration image along the Y axis; min[l x ,l y ] means taking l x and l y The minimum value in x ,l y ] means taking l x and l y The maximum value in ; θ is a constant coefficient, and θ∈(0,1);

[0018] Get the jth scanning feature parameter of the i-th grid area of ​​the image to be scanned during the scanning process Using the first BP neural network to obtain scanning feature parameters The first correction value Using the second BP neural network to obtain scanning feature parameters The second corrected value Using the third BP neural network to obtain scanning feature parameters The third corrected value

[0019] According to the scanning characteristic parameters The correction value is used to calculate the output value The calculation formula is: Among them, α, β, γ are weight coefficients of the first correction value, the second correction value and the third correction value respectively, and α, β, γ∈(0,1); 0<α<1; 0<β<1; 0<γ<1.

[0020] The process of determining the weight coefficients α, β, and γ includes the following steps:

[0021] The calibration image is scanned by a scanning device, a scanned image of the calibration image by the scanning device under the influence of the factors is obtained, and the j-th scanning feature parameter of the i-th grid area in the scanned image is extracted

[0022] Calculate the scanning feature parameters separately and the calibration characteristic parameter Cp ij The difference ΔCp ij , in,

[0023] Difference of scan feature parameters and ΔCp ij Perform correlation analysis and calculate the scanning feature parameters respectively and ΔCp ij The correlation coefficient r u 、r v 、r w ; The calculation formulas are:

[0024]

[0025] In the formula, is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area when only the equipment factor exists; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of environmental factors only; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of only the operation factor; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of all factors;

[0026] Using the correlation coefficient r u 、r v 、r w The calculation results are used to determine the weight coefficients α, β, and γ respectively according to the following formulas:

[0027]

[0028] A real-time scanning image processing system based on multi-source error analysis, including an error analysis module, a feature parameter acquisition module, a neural network construction module, an image processing module, an image correction module, and an image output module;

[0029] The error analysis module is used to analyze the causes of scanned image errors, and classify the causes of errors into equipment factors, environmental factors, and operational factors according to the analysis results, and obtain the influencing factors k of each of the factors respectively. 1u , k 2v , k 3w ;

[0030] The characteristic parameter acquisition module is used to set a calibration image, divide the calibration image into a number of grids with a step size of λ, and extract the calibration characteristic parameter Cp of the grid area of ​​the calibration image. ij , using the control variable method to obtain the scanned image of the calibration image by the scanning device under the influence of any of the factors, and obtain the scanning feature parameters of the corresponding grid area in the scanned image

[0031] The neural network building module is used to scan feature parameters only under the influence of the device factor. As input, the calibration characteristic parameter Cp ij The first BP neural network is constructed for output; the scanning characteristic parameters only exist under the influence of the environmental factors. As input, the calibration characteristic parameter Cp ij Construct a second BP neural network for output; use the scanning characteristic parameters under the influence of the operating factors only As input, the calibration characteristic parameter Cp ij Constructing a third BP neural network for output; training the established neural network, and adjusting the number of hidden layers of the neural network model to an accuracy rate not lower than an expected value;

[0032] The image processing module is used to divide the image to be scanned into a plurality of grids with a step length of ε according to a size ratio k between the image to be scanned and the calibration image, ε=k×λ;

[0033] The image correction module is used to obtain the jth scanning feature parameter of the i-th grid area of ​​the image to be scanned during the scanning process. Using the first BP neural network to obtain scanning feature parameters The first correction value Using the second BP neural network to obtain scanning feature parameters The second corrected value Using the third BP neural network to obtain scanning feature parameters The third corrected value

[0034] The image output module is used to calculate according to the formula: Calculate its output value

[0035] The system also includes a memory, a processor, and a computer program stored on the memory and executable on the processor.

[0036] The present invention has the following beneficial effects:

[0037] Compared with the prior art, the causes of errors are classified into equipment factors, environmental factors, and operating factors, the influencing factors of each of the factors are obtained, the calibration feature parameters of the grid area of ​​the calibration image are extracted, and the control variable method is used to respectively obtain the scanned image of the calibration image by the scanning device under the influence of any of the factors, and the scanning feature parameters at the corresponding grid area in the scanned image are obtained. The scanning feature parameters under the influence of any of the factors are used as input, and a BP neural network is constructed with the calibration feature parameters as output. According to the size ratio k of the image to be scanned and the calibration image, the image to be scanned is segmented, and the constructed BP neural network is used to obtain the correction value of the scanning feature parameter, and the correction value is weighted to obtain the output value of the scanning feature parameter of the image to be scanned. The invention can effectively solve the technical problems that the scanning unit is affected by the vibration factors generated by the mechanical movement during the dynamic scanning process, resulting in blurred and distorted scanned images and insufficient clarity, and the scanning image is skewed due to the wear of the mechanical friction parts of the transmission structure of the scanning mechanism after long-term use, thereby improving the quality of the scanned image. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the real-time processing method of scanned images based on multi-source error analysis of the present invention;

[0039] Figure 2 It is a structural block diagram of the scanning image real-time processing system based on multi-source error analysis of the present invention;

[0040] Figure 3 This is a structural block diagram of the neural network constructed in the solution of the present invention. DETAILED DESCRIPTION

[0041] The present invention will be further described below in conjunction with specific implementation methods, wherein the accompanying drawings are only used for exemplary descriptions and represent only schematic diagrams rather than actual drawings, and should not be understood as limiting the present invention. In order to better illustrate the specific implementation methods of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product.

[0042] The specific implementation process of the technical solution of the present invention includes the following steps:

[0043] Step 1: Analyze the causes of scan image errors, and classify the causes of errors into equipment factors, environmental factors, and operational factors according to the analysis results, and obtain the influencing factors k of each factor respectively. 1u , k 2v , k 3w , k 1u Expressed as the u-th influencing factor of the equipment factor; k 2v Expressed as the vth influencing factor of environmental factors; k 3wIt is expressed as the w-th influencing factor of environmental factors; u, v, and w are all positive integers;

[0044] The influencing factors of the equipment factors include at least one of the wear amount of the guide shaft in the X-axis direction, the wear amount of the guide shaft in the Y-axis direction, the scanning movement speed in the X-axis direction, the scanning movement speed in the Y-axis direction, the vibration amplitude of the guide shaft in the X-axis direction during scanning, and the vibration amplitude of the guide shaft in the Y-axis direction during scanning;

[0045] The influencing factors of the environmental factors include at least one of the equipment operating temperature, environmental humidity, background light source brightness, and ambient light source brightness;

[0046] The operational factors include at least one of the surface flatness of the object to be scanned, the overall size of the object to be scanned, the time interval since the last maintenance, the resolution setting value, and the operating time of the equipment;

[0047] Step 2: Set the calibration image, divide the calibration image into several grids with a step size of λ, and extract the calibration feature parameter Cp of the grid area of ​​the calibration image ij ;

[0048] Step 3: Use the control variable method to obtain the scanned images of the calibration image under the influence of any factor, and obtain the scanning feature parameters of the corresponding grid area in the scanned image. Cp ij It is expressed as the jth calibration feature parameter of the i-th grid area; It is represented as the jth scanning characteristic parameter of the i-th grid area under the influence of equipment factors only; It is represented as the jth scanning characteristic parameter of the i-th grid area under the influence of environmental factors only; It is represented as the j-th scanning characteristic parameter of the i-th grid area under the influence of only the operation factor; i = 1, 2, ..., n; n is the number of grid areas; j = 1, 2, ..., m; m is the type of characteristic parameter;

[0049] The feature parameters include at least one of geometric feature parameters, color feature parameters, texture feature parameters, resolution, and number of pixels; wherein the geometric feature parameters include at least one of perimeter, direction, aspect ratio, and eccentricity;

[0050] Step 4: Scan feature parameters with only device factors As input, to calibrate the characteristic parameter Cp ij The first BP neural network is constructed for output; the scanning characteristic parameters under the influence of environmental factors only exist As input, to calibrate the characteristic parameter Cp ijConstruct a second BP neural network for output; use the scanning characteristic parameters under the influence of only operating factors As input, to calibrate the characteristic parameter Cp ij Constructing a third BP neural network for output; training the established neural network, and adjusting the number of hidden layers of the neural network model until the accuracy is not less than the expected value;

[0051] Among them, the number of neurons in the input layer of the BP neural network is equal to the type of input feature parameters, the number of neurons in the output layer is equal to the output data dimension, that is, 1, and the initial value of the number of neurons in the intermediate hidden layer can be based on the empirical formula: Determine, where s is the number of neurons in the intermediate hidden layer, p and q are the number of neurons in the input layer and the number of neurons in the output layer respectively, q = 1; then the structure of the constructed neural network is as follows Figure 3 As shown;

[0052] Step 5: According to the size ratio k between the image to be scanned and the calibration image, the image to be scanned is divided into several grids with a step length of ε, ε = k × λ, and the step λ and its size satisfy the following relationship, specifically: In the formula, l x It is expressed as the length of the calibration image in the X-axis direction; l y It is expressed as the length of the calibration image along the Y axis; min[l x ,l y ] means taking l x and l y The minimum value in x ,l y ] means taking l x and l y The maximum value in ; θ is a constant coefficient, and θ∈(0,1);

[0053] Step 6: Obtain the jth scanning feature parameter of the i-th grid area of ​​the image to be scanned during the scanning process Using the first BP neural network to obtain scanning feature parameters The first correction value Using the second BP neural network to obtain scanning feature parameters The second corrected value Using the third BP neural network to obtain scanning feature parameters The third corrected value

[0054] Step 7: Based on the scanning feature parameters The output value is calculated by the correction value The calculation formula is: Among them, α, β, γ are weight coefficients of the first correction value, the second correction value and the third correction value respectively, and α, β, γ∈(0,1); 0<α<1; 0<β<1; 0<γ<1.

[0055] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A real-time scanning image processing method based on multi-source error analysis, characterized in that: include: Analyze the causes of scan image errors, classify the causes of errors into equipment factors, environmental factors, and operational factors based on the analysis results, and obtain the influencing factors k of each factor respectively. 1u , k 2v , k 3w , k 1u Expressed as the u-th influencing factor of the equipment factor; k 2v Expressed as the vth influencing factor of environmental factors; k 3w It is expressed as the w-th influencing factor of environmental factors; u, v, and w are all positive integers; Set a calibration image, divide the calibration image into several grids with a step size of λ, and extract the calibration feature parameter Cp of the grid area of ​​the calibration image ij , using the control variable method to obtain the scanned image of the calibration image by the scanning device under the influence of any of the factors, and obtain the scanning feature parameters of the corresponding grid area in the scanned image Cp ij It is expressed as the jth calibration feature parameter of the i-th grid area; It is represented as the j-th scanning characteristic parameter of the i-th grid area under the influence of the device factor only; It is represented by the j-th scanning characteristic parameter of the i-th grid area under the influence of the environmental factors only; It is represented by the j-th scanning characteristic parameter of the i-th grid area under the influence of the operating factor only; i=1,2,...,n; n is the number of grid areas; j=1,2,...,m; m is the type of characteristic parameter; The scanning characteristic parameters only exist under the influence of the device factors As input, the calibration characteristic parameter Cp ij The first BP neural network is constructed for output; the scanning characteristic parameters only exist under the influence of the environmental factors. As input, the calibration characteristic parameter Cp ij Construct a second BP neural network for output; The scanning characteristic parameters under the influence of the operating factors are As input, the calibration characteristic parameter Cp ij Constructing a third BP neural network for output; training the established neural network, and adjusting the number of hidden layers of the neural network model to an accuracy rate not lower than an expected value; According to the size ratio k between the image to be scanned and the calibration image, the image to be scanned is divided into a number of grids with a step size of ε, ε = k × λ, and the jth scanning feature parameter of the i-th grid area of ​​the image to be scanned during the scanning process is obtained. Using the first BP neural network to obtain scanning feature parameters The first correction value Using the second BP neural network to obtain scanning feature parameters The second corrected value Using the third BP neural network to obtain scanning feature parameters The third revised value According to the scanning characteristic parameters The correction value is used to calculate the output value The calculation formula is: Wherein, α, β, γ are weight coefficients of the first correction value, the second correction value and the third correction value respectively, and α, β, γ∈(0,1); 0<α<1; 0<β<1; 0<γ<1.

2. The real-time processing method for scanning images based on multi-source error analysis according to claim 1 is characterized in that: The influencing factors of the equipment factors include at least one of the wear amount of the guide shaft in the X-axis direction, the wear amount of the guide shaft in the Y-axis direction, the scanning movement speed in the X-axis direction, the scanning movement speed in the Y-axis direction, the vibration amplitude of the guide shaft in the X-axis direction during scanning, and the vibration amplitude of the guide shaft in the Y-axis direction during scanning; The influencing factors of the environmental factors include at least one of the equipment operating temperature, environmental humidity, background light source brightness, and ambient light source brightness; The operation factors include at least one of the surface flatness of the object to be scanned, the overall size of the object to be scanned, the time interval from the last maintenance, the resolution setting value, and the equipment operation time.

3. The real-time processing method for scanning images based on multi-source error analysis according to claim 1 is characterized in that: The feature parameters include at least one of geometric feature parameters, color feature parameters, texture feature parameters, resolution, and number of pixels; wherein the geometric feature parameters include at least one of circumference, direction, aspect ratio, and eccentricity.

4. The real-time processing method for scanning images based on multi-source error analysis according to claim 1 is characterized in that: The calculation formula of the expected value is: Where E(Y) represents the expected value; N is the input sample size of the neural network; f(X i ) is represented as the output function of the neural network; X i Represented as the output sample of the neural network.

5. The real-time processing method for scanning images based on multi-source error analysis according to claim 1 is characterized in that: The process of determining the weight coefficients α, β, and γ includes the following steps: The calibration image is scanned by a scanning device, a scanned image of the calibration image by the scanning device under the influence of the factors is obtained, and the j-th scanning feature parameter of the i-th grid area in the scanned image is extracted Calculate the scanning feature parameters separately and the calibration characteristic parameter Cp ij The difference ΔCp ij , in, Difference of scanning feature parameters and ΔCp ij Perform correlation analysis and calculate the scanning feature parameters respectively and ΔCp ij The correlation coefficient r u 、r v 、r w ; The calculation formulas are: In the formula, is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area when only the equipment factor exists; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of environmental factors only; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of only the operation factor; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of all factors; Using the correlation coefficient r u 、r v 、r w The calculation results are used to determine the weight coefficients α, β, and γ respectively according to the following formulas:

6. The real-time processing method for scanning images based on multi-source error analysis according to claim 1 is characterized in that: The segmentation step λ of the calibration image and its size satisfy the following relationship, specifically: In the formula, l x It is expressed as the length of the calibration image in the X-axis direction; l y It is expressed as the length of the calibration image along the Y axis; Expressed as taking l x and l y The minimum value in ; Expressed as taking l x and l y The maximum value among ; θ is a constant coefficient, and θ∈(0,1).

7. A real-time scanning image processing system based on multi-source error analysis, characterized in that: It includes error analysis module, feature parameter acquisition module, neural network construction module, image processing module, image correction module and image output module; The error analysis module is used to analyze the causes of scanned image errors, and classify the causes of errors into equipment factors, environmental factors, and operating factors according to the analysis results, and obtain the influencing factors k of each of the factors respectively. 1u , k 2v , k 3w , k 1u Expressed as the u-th influencing factor of the equipment factor; k 2v Expressed as the vth influencing factor of environmental factors; k 3w It is expressed as the w-th influencing factor of environmental factors; u, v, and w are all positive integers; The influencing factors of the equipment factors include at least one of the wear amount of the guide shaft in the X-axis direction, the wear amount of the guide shaft in the Y-axis direction, the scanning movement speed in the X-axis direction, the scanning movement speed in the Y-axis direction, the vibration amplitude of the guide shaft in the X-axis direction during scanning, and the vibration amplitude of the guide shaft in the Y-axis direction during scanning; The influencing factors of the environmental factors include at least one of the equipment operating temperature, environmental humidity, background light source brightness, and ambient light source brightness; The operation factor includes at least one of the surface flatness of the object to be scanned, the overall size of the object to be scanned, the time interval since the last maintenance, the resolution setting value, and the equipment operation time; The characteristic parameter acquisition module is used to set a calibration image, divide the calibration image into a number of grids with a step size of λ, and extract the calibration characteristic parameter Cp of the grid area of ​​the calibration image. ij , using the control variable method to obtain the scanned image of the calibration image by the scanning device under the influence of any of the factors, and obtain the scanning feature parameters of the corresponding grid area in the scanned image Cp ij It is expressed as the jth calibration feature parameter of the i-th grid area; It is represented as the j-th scanning characteristic parameter of the i-th grid area under the influence of the device factor only; It is represented by the j-th scanning characteristic parameter of the i-th grid area under the influence of the environmental factors only; It is represented by the j-th scanning characteristic parameter of the i-th grid area under the influence of the operating factor only; i=1,2,...,n; n is the number of grid areas; j=1,2,...,m; m is the type of characteristic parameter; The characteristic parameters include at least one of geometric characteristic parameters, color characteristic parameters, texture characteristic parameters, resolution, and number of pixels; wherein the geometric characteristic parameters include at least one of perimeter, direction, aspect ratio, and eccentricity; The neural network building module is used to scan feature parameters only under the influence of the device factor. As input, the calibration characteristic parameter Cp ij The first BP neural network is constructed for output; the scanning characteristic parameters only exist under the influence of the environmental factors. As input, the calibration characteristic parameter Cp ij Construct a second BP neural network for output; use the scanning characteristic parameters under the influence of the operating factors only As input, the calibration characteristic parameter Cp ij Constructing a third BP neural network for output; training the established neural network, adjusting the number of hidden layers of the neural network model to an accuracy rate not less than an expected value; wherein the calculation formula of the expected value is, Where E(Y) represents the expected value; N is the input sample size of the neural network; f(X i ) is represented as the output function of the neural network; X i Represented as an output sample of the neural network; The image processing module is used to divide the image to be scanned into a plurality of grids with a step length of ε according to the size ratio k between the image to be scanned and the calibration image, ε=k×λ; wherein the step length λ of the calibration image and its size satisfy the following relationship, specifically: In the formula, l x It is expressed as the length of the calibration image in the X-axis direction; l y It is expressed as the length of the calibration image along the Y axis; Expressed as taking l x and l y The minimum value in ; Expressed as taking l x and l y The maximum value in ; θ is a constant coefficient, and θ∈(0,1); The image correction module is used to obtain the jth scanning feature parameter of the i-th grid area of ​​the image to be scanned during the scanning process. Using the first BP neural network to obtain scanning feature parameters The first correction value Using the second BP neural network to obtain scanning feature parameters The second corrected value Using the third BP neural network to obtain scanning feature parameters The third revised value The image output module is used to scan the characteristic parameters The correction value is used to calculate the output value The calculation formula is: Wherein, α, β, γ are weight coefficients of the first correction value, the second correction value, and the third correction value, respectively, and α, β, γ∈(0,1); 0<α<1; 0<β<1; 0<γ<1; Wherein, the determination process of the weight coefficients α, β, γ includes the following steps: The calibration image is scanned by a scanning device, a scanned image of the calibration image by the scanning device under the influence of the factors is obtained, and the j-th scanning feature parameter of the i-th grid area in the scanned image is extracted Calculate the scanning feature parameters separately and the calibration characteristic parameter Cp ij The difference ΔCp ij , in, Difference of scanning feature parameters and ΔCp ij Perform correlation analysis and calculate the scanning feature parameters respectively and ΔCp ij The correlation coefficient r u 、r v 、r w ; The calculation formulas are: In the formula, is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area when only the equipment factor exists; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of environmental factors only; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of only the operation factor; is the mean value of the difference of the jth scanning characteristic parameter in the i-th grid area under the influence of all factors; Using the correlation coefficient r u 、r v 、r w The calculation results are used to determine the weight coefficients α, β, and γ respectively according to the following formulas:

8. The real-time scanning image processing system based on multi-source error analysis according to claim 7 is characterized in that: The system also includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, is capable of implementing the steps of the real-time processing method for scanned images based on multi-source error analysis as described in any one of claims 1-6.

Images