A method for processing flame images of combustion chamber of solid oxide fuel cell system

Abstract

The present application provides a method for processing flame images of a combustion chamber of a solid oxide fuel cell system, which is used to perform a series of preprocessing on the images collected from the combustion chamber of the SOFC system, so that the images have finer image details, which is convenient for subsequent high-precision and efficient combustion state analysis. The method includes: a processing device acquires an image to be processed of the combustion chamber of the SOFC system; the processing device performs image enhancement on the image to be processed through PIVF processing to obtain a first image; the processing device performs image enhancement on the first image through MSRCP processing to obtain a second image; the processing device performs grayscale processing on the first image to obtain a third image; the processing device fuses the first image, the second image and the third image to obtain a fourth image; the processing device fuses the first image, the second image and the fourth image to obtain a fifth image.

Description

Technical Field

[0001] The present application relates to the field of image processing, and in particular to a method for processing flame images of a combustion chamber of a solid oxide fuel cell system. Background Art

[0002] Solid oxide fuel cell (SOFC), as the third generation fuel cell, is currently a hot topic in research and commercial promotion. It is a new type of energy conversion device with the advantages of strong fuel adaptability, high efficiency, low cost and low pollution.

[0003] SOFC can emit very little or even no gas, and is a green energy source. After SOFC is added to a cogeneration technology system, its efficiency can be increased to 90%. In the SOFC system, the combustion chamber is mainly responsible for exhaust gas treatment and system heat supply. It is an important component to improve energy efficiency and reduce environmental pollution, and the safety and reliability of the SOFC system are important goals of its power generation system. If the system fault is not detected and resolved in time, the fault will gradually affect the stable and efficient operation of the entire system. Therefore, it is of great significance to monitor the combustion state of the combustion chamber during operation. Monitoring the combustion state of the SOFC combustion chamber requires identification and analysis of the combustion chamber flame to control the combustion process of the combustion chamber.

[0004] However, in the course of studying the existing related technologies, the inventors found that the images collected from the combustion chamber of the SOFC system generally have high brightness, low contrast, and color mixing. As a result, when analyzing the combustion state in the combustion chamber based on the images, it will lead to problems of low analysis accuracy and efficiency. Summary of the invention

[0005] The present application provides a method for processing flame images of a solid oxide fuel cell system combustion chamber, which is used to perform a series of preprocessing on images collected from the combustion chamber of the SOFC system, so that the images have finer image details, facilitating subsequent high-precision and efficient combustion state analysis.

[0006] In a first aspect, the present application provides a method for processing a flame image of a combustion chamber of a solid oxide fuel cell system, the method comprising:

[0007] The processing device obtains an image to be processed of a combustion chamber of the SOFC system, wherein the initial image is acquired by a camera disposed in the combustion chamber of the SOFC system;

[0008] The processing device performs image enhancement on the image to be processed through a perceptually inspired variational framework (PIVF) to obtain a first image to extract main colors;

[0009] The processing device performs image enhancement on the first image through Multi-Scale Retinex With Chromaticity Preservation (MSRCP) processing to obtain a second image to improve color contrast;

[0010] The processing device performs grayscale processing on the first image to obtain a third image to extract brightness information;

[0011] The processing device fuses the first image, the second image and the third image to obtain a fourth image to combine the grayscale information and the color information;

[0012] The processing device fuses the first image, the second image, and the fourth image to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image.

[0013] In conjunction with the first aspect of the present application, in a first possible implementation manner of the first aspect of the present application, the PIVF processing includes:

[0014] Obtain a weight matrix of the image size of the image to be processed, wherein the points closer to the center point in the weight matrix are given greater weights;

[0015] Normalize the image to be processed, obtain the image polynomial, and perform Fourier transform on the image polynomial;

[0016] Element-wise multiplication using the weight matrix and the image polynomial;

[0017] Perform inverse Fourier transform on the image polynomial;

[0018] Perform image calculations using pre-calculated polynomial coefficients to obtain a first image.

[0019] In combination with the first aspect of the present application, in a second possible implementation of the first aspect of the present application, the PIVF processing performs image enhancement in a gradient descent manner, and the PIVF processing is defined as:

[0020]

[0021] Among them, I k is the result of each iteration, k is the number of iterations, Δt is the descending step size, α and β are balance terms, I0 is the original image, μ is a constant of 0.5, represents the processing of the image, Δt is the step size of the gradient descent,

[0022] Defined as:

[0023]

[0024] Among them, x and y represent two pixels of the image, w(x,y) is the Gaussian function that defines the distance between x and y, and w(x,y) decreases as the distance increases. is a monotonically increasing and differentiable function,

[0025] In each descent process, I(k)-I(k-1) is calculated, and it is determined whether the difference is less than the image enhancement threshold. If it is, the image enhancement processing is completed, wherein the interval of the image enhancement threshold is 0.025 to 0.03.

[0026] In combination with the first aspect of the present application, in a third possible implementation manner of the first aspect of the present application, the MSRCP processing is defined as:

[0027]

[0028] Among them, I is the image, R, G, B are the RGB channels of the image, Msr is the multi-scale homomorphic filtering MSR algorithm, a and b are two hyperparameters, a and b are both 0.01, and Sp is the simplest color balance algorithm.

[0029] In combination with the first aspect of the present application, in a fourth possible implementation manner of the first aspect of the present application, a processing device fuses the first image, the second image, and the third image to obtain a fourth image, including:

[0030] The processing device fuses the first image, the second image, and the third image based on a hue preservation model to obtain a fourth image. The hue preservation model is defined as:

[0031]

[0032] Among them, I is the first image, L is the second image, g is the third image, and c is the image color channel.

[0033] In combination with the first aspect of the present application, in a fifth possible implementation manner of the first aspect of the present application, a processing device fuses the first image, the second image, and the fourth image to obtain a fifth image, including:

[0034] The processing device fuses the first image, the second image and the fourth image based on wavelet fusion processing to obtain a fifth image. The wavelet fusion is defined as:

[0035]

[0036] Where V∈{LH,HL,HH}, LH, HL, HH are the vertical, horizontal, and diagonal high-frequency details of the image after wavelet decomposition, c is the image channel, n is the number of images, S is the image detail component to be fused, and W is the weight of the image to be fused.

[0037] Wavelet fusion processing uses the Laplace operator to maintain the contrast weight of the image, and uses the Gaussian curve to give different brightness weights to different images. The contrast weight is defined as:

[0038] C c,t =A*I c,t ,

[0039] Among them, C is the contrast weight, A is the Laplacian operator, I is the image, * is the convolution operation,

[0040] The brightness weight is defined as:

[0041]

[0042] Among them, N is the brightness weight, I is the image, μ is the average brightness, and σ is 0.5.

[0043] In combination with the first aspect of the present application, in a sixth possible implementation manner of the first aspect of the present application, after the processing device fuses the first image, the second image, and the fourth image to obtain the fifth image, the method further includes:

[0044] The processing device performs recognition processing on the flame combustion state in the combustion chamber of the SOFC system based on the fifth image, and the recognition processing includes image feature recognition processing and image segmentation processing.

[0045] In a second aspect, the present application provides a device for processing flame images of a combustion chamber of a solid oxide fuel cell system, the device comprising:

[0046] An acquisition unit, used for acquiring an image to be processed of a combustion chamber of the SOFC system, wherein the initial image is acquired by a camera disposed in the combustion chamber of the SOFC system;

[0047] A first enhancement unit is used to perform image enhancement on the image to be processed by using a perception-based variational framework PIVF to obtain a first image so as to extract main colors;

[0048] A second enhancement unit is used to perform image enhancement on the first image through a multi-scale homomorphic filtering MSRCP process with chroma preservation to obtain a second image so as to improve color contrast;

[0049] A third enhancement unit, configured to perform grayscale processing on the first image to obtain a third image, so as to extract brightness information;

[0050] A first fusion unit is used to fuse the first image, the second image and the third image to obtain a fourth image so as to combine grayscale information and color information;

[0051] The second fusion unit is used to fuse the first image, the second image and the fourth image to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image.

[0052] In conjunction with the second aspect of the present application, in a first possible implementation manner of the second aspect of the present application, the PIVF processing includes:

[0053] Obtain a weight matrix of the image size of the image to be processed, wherein the points closer to the center point in the weight matrix are given greater weights;

[0054] Normalize the image to be processed, obtain the image polynomial, and perform Fourier transform on the image polynomial;

[0055] Element-wise multiplication using the weight matrix and the image polynomial;

[0056] Perform inverse Fourier transform on the image polynomial;

[0057] Perform image calculations using pre-calculated polynomial coefficients to obtain a first image.

[0058] In conjunction with the second aspect of the present application, in a second possible implementation of the second aspect of the present application, the PIVF process performs image enhancement in a gradient descent manner, and the PIVF process is defined as:

[0059]

[0060] Among them, I k is the result of each iteration, k is the number of iterations, Δt is the descending step size, α and β are balance terms, I0 is the original image, μ is a constant of 0.5, represents the processing of the image, Δt is the step size of the gradient descent,

[0061] Defined as:

[0062]

[0063] Among them, x and y represent two pixels of the image, w(x,y) is the Gaussian function that defines the distance between x and y, and w(x,y) decreases as the distance increases. is a monotonically increasing and differentiable function,

[0064] In each descent process, I(k)-I(k-1) is calculated, and it is determined whether the difference is less than the image enhancement threshold. If it is, the image enhancement processing is completed, wherein the interval of the image enhancement threshold is 0.025 to 0.03.

[0065] In conjunction with the second aspect of the present application, in a third possible implementation of the second aspect of the present application, the MSRCP process is defined as:

[0066]

[0067] Among them, I is the image, R, G, B are the RGB channels of the image, Msr is the multi-scale homomorphic filtering MSR algorithm, a and b are two hyperparameters, a and b are both 0.01, and Sp is the simplest color balance algorithm.

[0068] In combination with the second aspect of the present application, in a fourth possible implementation manner of the second aspect of the present application, the first fusion unit is specifically configured to:

[0069] Based on the hue preservation model, the first image, the second image and the third image are fused to obtain a fourth image. The hue preservation model is defined as:

[0070]

[0071] Among them, I is the first image, L is the second image, g is the third image, and c is the image color channel.

[0072] In combination with the second aspect of the present application, in a fifth possible implementation manner of the second aspect of the present application, the second fusion unit is specifically configured to:

[0073] Based on wavelet fusion processing, the first image, the second image and the fourth image are fused to obtain the fifth image. The wavelet fusion is defined as:

[0074]

[0075] Where V∈{LH,HL,HH}, LH, HL, HH are the vertical, horizontal, and diagonal high-frequency details of the image after wavelet decomposition, c is the image channel, n is the number of images, S is the image detail component to be fused, and W is the weight of the image to be fused.

[0076] Wavelet fusion processing uses the Laplace operator to maintain the contrast weight of the image, and uses the Gaussian curve to give different brightness weights to different images. The contrast weight is defined as:

[0077] C c,t =A*I c,t ,

[0078] Among them, C is the contrast weight, A is the Laplacian operator, I is the image, * is the convolution operation,

[0079] The brightness weight is defined as:

[0080]

[0081] Among them, N is the brightness weight, I is the image, μ is the average brightness, and σ is 0.5.

[0082] In combination with the second aspect of the present application, in a sixth possible implementation manner of the second aspect of the present application, the apparatus further includes an identification unit, which is configured to:

[0083] Based on the fifth image, the flame combustion state in the combustion chamber of the SOFC system is identified and processed, and the identification process includes image feature recognition processing and image segmentation processing.

[0084] In a third aspect, the present application provides a processing device, including a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the method provided in the first aspect of the present application or any possible implementation method of the first aspect of the present application is executed.

[0085] In a fourth aspect, the present application provides a computer-readable storage medium, which stores multiple instructions, and the instructions are suitable for a processor to load to execute the method provided by the first aspect of the present application or any possible implementation of the first aspect of the present application.

[0086] From the above content, it can be concluded that the present application has the following beneficial effects:

[0087] In order to serve the image recognition processing of the combustion state of the combustion chamber of the SOFC system, after obtaining the image to be processed collected from the combustion chamber, the present application applies PIVF processing to the image to be processed to obtain a first image after image enhancement, and on the other hand, continues to process the first image through MSRCP to obtain a second image after image enhancement, and on the other hand, continues to grayscale the first image to obtain a third image. At this time, the first, second and third images are fused to obtain a fourth image having the image information of the three images, and then the first, second and third images are fused to obtain a fourth image having the image information of the three images. Figure 2 The fourth image is fused to obtain a fifth image obtained by correcting the fourth image through the first image and the second image. At this time, the fifth image has finer image details, thereby facilitating subsequent high-precision and efficient combustion state analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0089] Figure 1 A schematic flow chart of a method for processing flame images of a combustion chamber of a solid oxide fuel cell system of the present application;

[0090] Figure 2 A schematic diagram of a scenario of a method for processing flame images of a combustion chamber of a solid oxide fuel cell system of the present application;

[0091] Figure 3 A comparative schematic diagram of the image enhancement effect of this application;

[0092] Figure 4 A schematic diagram of the structure of a device for processing flame images of a combustion chamber of a solid oxide fuel cell system of the present application;

[0093] Figure 5 A schematic diagram of the structure of the processing equipment of this application. DETAILED DESCRIPTION

[0094] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0095] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described here can be implemented in an order other than that illustrated or described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. The naming or numbering of steps in this application does not mean that the steps in the method flow must be executed in the time / logical sequence indicated by the naming or numbering. The process steps that have been named or numbered can change the execution order according to the technical purpose to be achieved, as long as the same or similar technical effects can be achieved.

[0096] The division of modules in this application is a logical division. There may be other division methods when it is implemented in actual applications. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection between modules can be electrical or other similar forms, which are not limited in this application. In addition, the modules or submodules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed in multiple circuit modules, and some or all of the modules may be selected according to actual needs to achieve the purpose of the present application.

[0097] Before introducing the method for processing flame images of a combustion chamber of a solid oxide fuel cell system provided in the present application, the background content involved in the present application is first introduced.

[0098] The method, device and computer-readable storage medium for processing flame images of the combustion chamber of a solid oxide fuel cell system provided in the present application can be applied to a processing device for performing a series of preprocessing on images collected from the combustion chamber of the SOFC system, so that the images have finer image details, facilitating subsequent high-precision and efficient combustion state analysis.

[0099] The method for processing flame images of a combustion chamber of a solid oxide fuel cell system mentioned in the present application may be executed by a processing device for the flame images of a combustion chamber of a solid oxide fuel cell system, or a server, a physical host, or a user equipment (UE) or other different types of processing equipment that integrates the processing device for the flame images of a combustion chamber of a solid oxide fuel cell system. Among them, the processing device for the flame images of a combustion chamber of a solid oxide fuel cell system may be implemented in hardware or software, and the UE may be a terminal device such as a smart phone, a tablet computer, a laptop computer, a desktop computer, or a personal digital assistant (PDA), and the processing device may be set in a device cluster.

[0100] Next, the method for processing flame images of a combustion chamber of a solid oxide fuel cell system provided by the present application is introduced.

[0101] First, see Figure 1 , Figure 1 A flow chart of a method for processing a flame image of a combustion chamber of a solid oxide fuel cell system of the present application is shown. The method for processing a flame image of a combustion chamber of a solid oxide fuel cell system provided by the present application may specifically include the following steps S101 to S106:

[0102] Step S101, a processing device acquires an image to be processed of a combustion chamber of a SOFC system, wherein the initial image is acquired by a camera disposed in the combustion chamber of the SOFC system;

[0103] It can be understood that the image to be processed, in layman's terms, is obtained by photographing the flame in the combustion chamber of the SOFC system with a camera. In the SOFC system, the combustion chamber is mainly responsible for exhaust gas treatment and system heat supply. It is in a combustion state, that is, there is a flame. Therefore, the flame can be used to judge the specific combustion state in the combustion chamber.

[0104] The camera, which can also be called a camera head, is an image acquisition device that can be used normally in the high temperature environment of the combustion chamber. Generally, an industrial high temperature resistant camera is used.

[0105] The camera can collect images at regular intervals or at regular intervals (in video format). It can work by default or under an external trigger signal. The specific working mode can be adjusted according to actual conditions and is not specifically limited here.

[0106] After the camera captures the image, it can be directly output or output through an intermediate device to a processing device for the processing device to perform image processing on it.

[0107] Of course, in some cases, the camera may also be a part of the processing device, that is, the camera is included in the processing device.

[0108] Step S102, the processing device performs image enhancement on the image to be processed through PIVF processing to obtain a first image to extract main colors;

[0109] Regarding the PIVF processing introduced in this application, that is, the perception-based variational framework processing, it can be understood that the flame image of the combustion chamber usually has the characteristics of high brightness, low contrast, color mixing, etc., which is not conducive to flame image recognition and analysis. Therefore, this application believes that the first thing to do is the color separation task to extract the color of the color mixed area, and the PIVF processing can extract the main color of the local area of ​​the image from the image to be processed, realize the separation of the main colors, and achieve image enhancement effect.

[0110] At the same time, in actual application, the present application found that although PIVF processing improves the color separation effect of the flame image, that is, the details of the flame image are clear, there are also cases where the color of some flames is slightly offset and the image contrast is insufficient. Therefore, this problem can be overcome through subsequent image processing.

[0111] In the specific application of PIVF processing, as a practical implementation method, it mainly includes the following processing contents:

[0112] 1. Obtain a weight matrix of the image size of the image to be processed, wherein the points closer to the center point in the weight matrix are given greater weights;

[0113] The weight matrix here can be regarded as a filter with a size of w*h, where the weight at the center point is the largest and the weight becomes smaller as the distance from the center point increases.

[0114] 2. Normalize the image to be processed, obtain the image polynomial, and perform Fourier transform on the image polynomial;

[0115] The image here is one of the RGB channels of the image, defined as x, and the image polynomial is F = ax + bx 2 +cx 3 +…zx 9 , for x, x 2 , x 3 ...perform Fourier transform and obtain x',x' 2 ,x' 3 ….

[0116] 3. Use the weight matrix and image polynomial to perform element-wise multiplication;

[0117] It can be understood that here is to change x', x' 2 ,x'3 The constructed image polynomials are multiplied by the initial weight matrix.

[0118] 4. Perform inverse Fourier transform on the image polynomial.

[0119] It can be understood that after the multiplication process, the image format before Fourier transformation can be restored through inverse Fourier transformation.

[0120] 5. Use the pre-calculated polynomial coefficients to perform image calculation to obtain the first image.

[0121] The inverse Fourier transform result is calculated using the pre-calculated a, b, c ... z coefficients corresponding to the polynomial, so that an enhanced image can be obtained.

[0122] As another practical implementation, on the other hand, PIVF processing can be understood as image enhancement in a gradient descent manner, and PIVF processing itself, as a specific implementation scheme, can also be defined in the form of a formula as follows:

[0123]

[0124] Among them, I k is the result of each iteration, k is the number of iterations, Δt is the descending step size, α and β are balance terms, I0 is the original image, μ is a constant of 0.5, represents the processing of the image, Δt is the step size of the gradient descent,

[0125] Defined as:

[0126]

[0127] Among them, x and y represent two pixels of the image, w(x,y) is the Gaussian function that defines the distance between x and y, and w(x,y) decreases as the distance increases. is a monotonically increasing and differentiable function,

[0128] In each descent process, I(k)-I(k-1) is calculated, and it is determined whether the difference is less than the image enhancement threshold. If it is, the image enhancement processing is completed, wherein the interval of the image enhancement threshold is 0.025 to 0.03.

[0129] Step S103, the processing device performs image enhancement on the first image through MSRCP processing to obtain a second image to improve color contrast;

[0130] For the first image obtained by PIVF processing, the present application continues to introduce MSRCP processing, that is, multi-scale homomorphic filtering processing with chroma preservation, which is used to maintain the color of the image while enhancing the contrast between the image colors.

[0131] At the same time, the present application found in actual application that, for the second image obtained by MSRCP processing, the contrast of the details and the main flame can achieve a good state, but the contrast of some flames is insufficient.

[0132] Similar to the PIVF process, as another exemplary implementation, the MSRCP process itself can be defined in the form of a formula as follows:

[0133]

[0134] Among them, I is the image, R, G, B are the RGB channels of the image, Msr is the multi-scale homomorphic filtering MSR algorithm, a and b are two hyperparameters, a and b are both 0.01, and Sp is the simplest color balance algorithm.

[0135] Step S104, the processing device performs grayscale processing on the first image to obtain a third image to extract brightness information;

[0136] In addition to the MSRCP processing, the present application also introduces a grayscale processing for the first image obtained by the PIVF processing, so as to convert the first image into a grayscale image, thereby reflecting the obvious brightness information in the first image.

[0137] Step S105, the processing device fuses the first image, the second image and the third image to obtain a fourth image to combine grayscale information and color information;

[0138] From the processing of the first, second and third images and their effects, it can be found that the three have their own advantages in image enhancement, namely, the advantages of color separation, contrast enhancement and brightness enhancement, which can achieve the effect of maintaining image details and enhancing the color contrast and brightness of the image. At this time, the present application can integrate the three. In this way, while retaining the advantages of the three, it can also overcome the shortcomings of the three to a certain extent to a large extent.

[0139] Specifically, the fourth image is processed for the purpose of maintaining the hue, and the hue-maintaining processing, that is, the specific execution subject of the fusion processing, can be executed by the hue-maintaining model, which is configured with the image processing logic involved to complete the fusion of the first, second and third images.

[0140] That is, the processing device may specifically fuse the first image, the second image, and the third image based on the tone preservation model to obtain the fourth image.

[0141] As another practical implementation, similar to the above, the hue preservation model itself can be defined as follows in the form of a formula:

[0142]

[0143] Among them, I is the first image, L is the second image, g is the third image, and c is the image color channel.

[0144] Step S106: The processing device fuses the first image, the second image, and the fourth image to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image.

[0145] Furthermore, in the process of fusing the first, second and third images in step S105, the present application believes that, although the disadvantages of the three images have been largely overcome while retaining their advantages, the problem can still be overcome to further enhance the image quality of the fourth image with respect to flame recognition.

[0146] Specifically, the present application may further continue to fuse the fourth image with the previous first and second images, so as to correct the fourth image through both the first image and the second image, balance the image details, promote a better balance between contrast and brightness, and eliminate the flame boundary fusion problem in small areas that may still exist in the image, avoiding over-enhancement and under-enhancement.

[0147] In order to serve the image recognition processing of the combustion state of the combustion chamber of the SOFC system, after obtaining the image to be processed collected from the combustion chamber, the present application applies PIVF processing to the image to be processed to obtain a first image after image enhancement, and on the other hand, continues to process the first image through MSRCP to obtain a second image after image enhancement, and on the other hand, continues to grayscale the first image to obtain a third image. At this time, the first, second and third images are fused to obtain a fourth image having the image information of the three images, and then the first, second and third images are fused to obtain a fourth image having the image information of the three images. Figure 2 The fourth image is fused to obtain a fifth image obtained by correcting the fourth image through the first image and the second image. At this time, the fifth image has finer image details, thereby facilitating subsequent high-precision and efficient combustion state analysis.

[0148] As another practical implementation, in the process of processing the fifth image, that is, fusing the first, second and fourth images, it can be implemented by wavelet fusion processing, that is:

[0149] The processing device may specifically fuse the first image, the second image and the fourth image based on wavelet fusion processing to obtain the fifth image.

[0150] It can be understood that wavelet fusion can decompose images at multiple scales, and its fusion process mainly includes:

[0151] 1. Wavelet transform to decompose the image;

[0152] 2. Use appropriate fusion rules to fuse coefficients;

[0153] 3. Reconstruct the image by inverse wavelet transform.

[0154] The image sequence is first decomposed by wavelet transform to obtain the vertical, horizontal, diagonal details and low-frequency components of the image. Then, the corresponding details of the image sequence are fused, and finally the different detail components are fused into the final image through inverse wavelet transform.

[0155] At the same time, wavelet fusion can be defined in the form of a formula as follows:

[0156]

[0157] Where V∈{LH,HL,HH}, LH, HL, HH are the vertical, horizontal, and diagonal high-frequency details of the image after wavelet decomposition, c is the image channel, n is the number of images, S is the image detail component to be fused, and W is the weight of the image to be fused.

[0158] In addition, the present application believes that image details also need to be maintained in the wavelet high-frequency components, and the Laplace operator can perform image sharpening and has the ability to maintain details, so the present application can specifically select the Laplace operator to maintain image detail information.

[0159] At the same time, the image also needs to maintain moderate brightness, and the Gaussian curve has nonlinear characteristics, which can give higher weights to images close to the average brightness. Therefore, the Gaussian curve is used to give different brightness weights to different images.

[0160] Specifically, the wavelet fusion process can maintain the contrast weight of the image through the Laplace operator, and use the Gaussian curve to give different brightness weights to different images, and the contrast weight can be defined as the following in the form of a formula:

[0161] C c,t =A*I c,t ,

[0162] Among them, C is the contrast weight, A is the Laplacian operator, I is the image, * is the convolution operation,

[0163] The brightness weight can be defined as the following in the form of a formula:

[0164]

[0165] Among them, N is the brightness weight, I is the image, μ is the average brightness, and σ is 0.5.

[0166] In the subsequent joint application of contrast weight and brightness weight, it can be done through the following formula:

[0167] W c,t =C c,t N c,t .

[0168] For the above aspects of image enhancement processing and image fusion processing, you can also refer to Figure 2 A scenario schematic diagram of the method for processing flame images of a combustion chamber of a solid oxide fuel cell system of the present application is shown for understanding.

[0169] In addition, as an example, see Figure 3 A comparative schematic diagram of the image enhancement effect of the present application is shown in FIG. Figure 3 On the left, there is an original image on the top and bottom, and on Figure 3 On the right, the upper and lower images are images that have undergone image enhancement (including image enhancement processing and image fusion processing), referred to as enhanced images. It can be seen that after processing by this application, the image on the right presents finer image details, and in the case of this finer image detail, in actual applications, it can also reflect more delicate flame color content.

[0170] Therefore, it is possible to provide more detailed image support for the subsequent flame state identification of the combustion chamber of the SOFC system, facilitating the subsequent high-precision and efficient combustion state analysis.

[0171] Correspondingly, after step S106, the application content of the image may also be included, namely:

[0172] The processing device performs recognition processing on the flame combustion state in the combustion chamber of the SOFC system based on the fifth image, and the recognition processing includes image feature recognition processing and image segmentation processing.

[0173] It can be understood that in the process of applying the fifth image to the flame combustion state, image feature recognition processing and image segmentation processing may be involved, and these specific image processing are proposed to identify the specific combustion state of the flame in the image content, that is, it can specifically involve image feature recognition processing and image segmentation processing of the flame image, so as to accurately identify its specific flame combustion state.

[0174] Of course, during the recognition process, image feature recognition processing and image segmentation processing of image contents other than flames may also be involved in order to eliminate background images or interfering images.

[0175] The recognition processing here can be specifically performed by different types of neural network models such as perceptron, convolutional neural network, residual shrinkage network, etc. For the model, it itself is trained by a sample flame image labeled with the flame state recognition result. Therefore, it is targeted at the flame image content in the input image and can be recognized. Since the input image has been enhanced for this application, it can better ensure that the model can recognize the specific and accurate flame combustion state.

[0176] From the above content, it can be seen that in order to serve the image recognition processing of the combustion state of the combustion chamber of the SOFC system, after obtaining the image to be processed collected from the combustion chamber, the present application applies PIVF processing to the image to be processed to obtain a first image after image enhancement, and on the other hand, continues to process the first image through MSRCP to obtain a second image after image enhancement, and on the other hand, continues to grayscale the first image to obtain a third image. At this time, the first, second and third images are fused to obtain a fourth image having the image information of the three images, and then the first, second and third images are fused to obtain a fourth image having the image information of the three images. Figure 2 The fourth image is fused to obtain a fifth image obtained by correcting the fourth image through the first image and the second image. At this time, the fifth image has finer image details, thereby facilitating subsequent high-precision and efficient combustion state analysis.

[0177] The above is an introduction to the method for processing flame images of the combustion chamber of a solid oxide fuel cell system provided in the present application. In order to facilitate better implementation of the method for processing flame images of the combustion chamber of a solid oxide fuel cell system provided in the present application, the present application also provides a device for processing flame images of the combustion chamber of a solid oxide fuel cell system from the perspective of functional modules.

[0178] See also Figure 4 , Figure 4 This is a structural schematic diagram of a device for processing flame images of a combustion chamber of a solid oxide fuel cell system of the present application. In the present application, the device 400 for processing flame images of a combustion chamber of a solid oxide fuel cell system may specifically include the following structure:

[0179] An acquisition unit 401 is used to acquire an image to be processed of a combustion chamber of the SOFC system, wherein the initial image is acquired by a camera disposed in the combustion chamber of the SOFC system;

[0180] A first enhancement unit 402 is used to perform image enhancement on the image to be processed by using a perception-based variational framework (PIVF) to obtain a first image, so as to extract main colors;

[0181] The second enhancement unit 403 is used to perform image enhancement on the first image by multi-scale homomorphic filtering with chroma preservation MSRCP processing to obtain a second image to improve color contrast;

[0182] The third enhancement unit 404 is used to perform grayscale processing on the first image to obtain a third image to extract brightness information;

[0183] A first fusion unit 405 is used to fuse the first image, the second image and the third image to obtain a fourth image to combine grayscale information and color information;

[0184] The second fusion unit 406 is used to fuse the first image, the second image and the fourth image to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image.

[0185] In an exemplary implementation, the PIVF process includes:

[0186] Obtain a weight matrix of the image size of the image to be processed, wherein the points closer to the center point in the weight matrix are given greater weights;

[0187] Normalize the image to be processed, obtain the image polynomial, and perform Fourier transform on the image polynomial;

[0188] Element-wise multiplication using the weight matrix and the image polynomial;

[0189] Perform inverse Fourier transform on the image polynomial;

[0190] Perform image calculations using pre-calculated polynomial coefficients to obtain a first image.

[0191] In another exemplary implementation, the PIVF process performs image enhancement in a gradient descent manner, and the PIVF process is defined as:

[0192]

[0193] Among them, I k is the result of each iteration, k is the number of iterations, Δt is the descending step size, α and β are balance terms, I0 is the original image, μ is a constant of 0.5, represents the processing of the image, Δt is the step size of the gradient descent,

[0194] Defined as:

[0195]

[0196] Among them, x and y represent two pixels of the image, w(x,y) is the Gaussian function that defines the distance between x and y, and w(x,y) decreases as the distance increases. is a monotonically increasing and differentiable function,

[0197] In each descent process, I(k)-I(k-1) is calculated, and it is determined whether the difference is less than the image enhancement threshold. If it is, the image enhancement processing is completed, wherein the interval of the image enhancement threshold is 0.025 to 0.03.

[0198] In yet another exemplary implementation, the MSRCP process is defined as:

[0199]

[0200] Among them, I is the image, R, G, B are the RGB channels of the image, Msr is the multi-scale homomorphic filtering MSR algorithm, a and b are two hyperparameters, a and b are both 0.01, and Sp is the simplest color balance algorithm.

[0201] In another exemplary implementation, the first fusion unit 405 is specifically configured to:

[0202] Based on the hue preservation model, the first image, the second image and the third image are fused to obtain a fourth image. The hue preservation model is defined as:

[0203]

[0204] Among them, I is the first image, L is the second image, g is the third image, and c is the image color channel.

[0205] In yet another exemplary implementation, the second fusion unit 406 is specifically configured to:

[0206] Based on wavelet fusion processing, the first image, the second image and the fourth image are fused to obtain the fifth image. The wavelet fusion is defined as:

[0207]

[0208] Where V∈{LH,HL,HH}, LH, HL, HH are the vertical, horizontal, and diagonal high-frequency details of the image after wavelet decomposition, c is the image channel, n is the number of images, S is the image detail component to be fused, and W is the weight of the image to be fused.

[0209] Wavelet fusion processing uses the Laplace operator to maintain the contrast weight of the image, and uses the Gaussian curve to give different brightness weights to different images. The contrast weight is defined as:

[0210] C c,t =A*I c,t ,

[0211] Among them, C is the contrast weight, A is the Laplacian operator, I is the image, * is the convolution operation,

[0212] The brightness weight is defined as:

[0213]

[0214] Among them, N is the brightness weight, I is the image, μ is the average brightness, and σ is 0.5.

[0215] In yet another exemplary implementation, the apparatus further includes an identification unit 407, configured to:

[0216] Based on the fifth image, the flame combustion state in the combustion chamber of the SOFC system is identified and processed, and the identification process includes image feature recognition processing and image segmentation processing.

[0217] This application also provides a processing device from the perspective of hardware structure, see Figure 5 , Figure 5 501, a memory 502, and an input / output device 503. The processor 501 is used to execute the computer program stored in the memory 502 to implement the following Figure 1 The steps of the method for processing a flame image of a combustion chamber of a solid oxide fuel cell system in the corresponding embodiment; or, the processor 501 is used to execute the computer program stored in the memory 502 to implement the following Figure 4 Corresponding to the functions of each unit in the embodiment, the memory 502 is used to store the processor 501 executing the above Figure 1 The computer program required for the method of processing flame images of the combustion chamber of the solid oxide fuel cell system in the corresponding embodiment.

[0218] Exemplarily, the computer program may be divided into one or more modules / units, one or more modules / units are stored in the memory 502, and executed by the processor 501 to complete the present application. One or more modules / units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the computer device.

[0219] The processing device may include, but is not limited to, a processor 501, a memory 502, and an input / output device 503. Those skilled in the art will appreciate that the illustration is merely an example of a processing device and does not constitute a limitation on the processing device, and may include more or fewer components than shown in the illustration, or a combination of certain components, or different components. For example, the processing device may also include a network access device, a bus, etc., and the processor 501, the memory 502, the input / output device 503, etc. are connected via a bus.

[0220] The processor 501 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the processing device, and uses various interfaces and lines to connect various parts of the entire device.

[0221] The memory 502 can be used to store computer programs and / or modules. The processor 501 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 502 and calling the data stored in the memory 502. The memory 502 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created according to the use of the processing device, etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0222] When the processor 501 is used to execute the computer program stored in the memory 502, the following functions can be implemented:

[0223] Acquire an image to be processed of a combustion chamber of the SOFC system, wherein the initial image is acquired by a camera disposed in the combustion chamber of the SOFC system;

[0224] The image to be processed is enhanced by using a perception-based variational framework PIVF to obtain a first image to extract the main colors;

[0225] Performing image enhancement on the first image by multi-scale homomorphic filtering with chroma preservation MSRCP processing to obtain a second image so as to improve color contrast;

[0226] Performing grayscale processing on the first image to obtain a third image to extract brightness information;

[0227] fusing the first image, the second image and the third image to obtain a fourth image to combine grayscale information and color information;

[0228] The first image, the second image and the fourth image are fused to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image.

[0229] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the processing device, processing equipment and corresponding units of the solid oxide fuel cell system combustion chamber flame image described above can refer to the following Figure 1 The description of the method for processing the flame image of the combustion chamber of the solid oxide fuel cell system in the corresponding embodiment will not be repeated here.

[0230] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0231] To this end, the present application provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the present application as follows: Figure 1 The steps of the method for processing the flame image of the combustion chamber of the solid oxide fuel cell system in the corresponding embodiment, the specific operation can refer to the following Figure 1 The description of the method for processing the flame image of the combustion chamber of the solid oxide fuel cell system in the corresponding embodiment will not be repeated here.

[0232] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0233] Due to the instructions stored in the computer-readable storage medium, the present application can be executed. Figure 1The steps of the method for processing the flame image of the combustion chamber of the solid oxide fuel cell system in the corresponding embodiment, therefore, the present application can be implemented as follows Figure 1 The beneficial effects that can be achieved by the method for processing flame images of a combustion chamber of a solid oxide fuel cell system in the corresponding embodiment are detailed in the previous description and will not be repeated here.

[0234] The above is a detailed introduction to the processing method, device, processing equipment and computer-readable storage medium for the flame image of the combustion chamber of the solid oxide fuel cell system provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for processing flame images of a combustion chamber of a solid oxide fuel cell system, characterized in that: The method comprises: The processing device obtains an image to be processed of a combustion chamber of a solid oxide fuel cell (SOFC) system, wherein the image to be processed is acquired by a camera disposed in the combustion chamber of the SOFC system; The processing device performs image enhancement on the image to be processed through a perception-based variational framework PIVF process to obtain a first image to extract main colors; The processing device performs image enhancement on the first image through multi-scale homomorphic filtering with chroma preservation MSRCP processing to obtain a second image to improve color contrast; The processing device performs grayscale processing on the first image to obtain a third image to extract brightness information; The processing device fuses the first image, the second image and the third image to obtain a fourth image to combine grayscale information and color information; The processing device fuses the first image, the second image and the fourth image to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image; The PIVF treatment includes: Obtaining a weight matrix of the image size of the image to be processed, wherein in the weight matrix, points closer to the center point are given greater weights; Normalizing the image to be processed, obtaining an image polynomial, and performing Fourier transform on the image polynomial; Perform element-wise multiplication using the weight matrix and the image polynomial; Performing an inverse Fourier transform on the image polynomial; Image calculation is performed using pre-calculated polynomial coefficients to obtain the first image.

2. The method according to claim 1, characterized in that The PIVF process performs image enhancement in a gradient descent manner, and the PIVF process is defined as: Among them, I k is the result of each iteration, k is the number of iterations, Δt is the descending step size, α and β are balance terms, I0 is the original image, μ is a constant of 0.5, represents the processing of the image, Δt is the step size of the gradient descent, Defined as: Among them, x and y represent two pixels of the image. w ( x,y ) is the Gaussian function that defines the distance between x and y. w ( x,y ) decreases with increasing distance, is a monotonically increasing and differentiable function, During each descent process, I(k)-I(k-1) is calculated, and it is determined whether the difference is less than the image enhancement threshold. If so, the image enhancement process is completed, wherein the interval of the image enhancement threshold is 0.025 to 0.

03.

3. The method according to claim 1, characterized in that The MSRCP process is defined as: Among them, I is the image, R, G, B are the RGB channels of the image, Msr is the multi-scale homomorphic filtering MSR algorithm, a and b are two hyperparameters, a and b are both 0.01, and Sp is the simplest color balance algorithm.

4. The method according to claim 1, characterized in that: The processing device fuses the first image, the second image, and the third image to obtain a fourth image, including: The processing device fuses the first image, the second image, and the third image based on a hue preservation model to obtain the fourth image, wherein the hue preservation model is defined as: Wherein, I is the first image, L is the second image, g is the third image, and c is the image color channel.

5. The method according to claim 1, characterized in that The processing device fuses the first image, the second image, and the fourth image to obtain a fifth image, including: The processing device fuses the first image, the second image and the fourth image based on wavelet fusion processing to obtain the fifth image, and the wavelet fusion is defined as: Wherein, V∈{LH,HL,HH}, LH, HL, HH are the vertical, horizontal, and diagonal high-frequency details of the image after wavelet decomposition, c is the image channel, n is the number of images, S is the image detail component to be fused, and W is the weight of the image to be fused. The wavelet fusion process maintains the contrast weight of the image through the Laplace operator, and uses the Gaussian curve to give different brightness weights to different images. The contrast weight is defined as: C c,t =A*I c,t , Among them, C is the contrast weight, A is the Laplacian operator, I is the image, * is the convolution operation, The brightness weight is defined as: Among them, N is the brightness weight, I is the image, μ is the average brightness, and σ is 0.

5.

6. The method according to claim 1, characterized in that After the processing device fuses the first image, the second image, and the fourth image to obtain a fifth image, the method further includes: The processing device performs recognition processing on the flame combustion state in the combustion chamber of the SOFC system based on the fifth image, wherein the recognition processing includes image feature recognition processing and image segmentation processing.

7. A device for processing flame images of a combustion chamber of a solid oxide fuel cell system, characterized in that: The device comprises: An acquisition unit, used for acquiring an image to be processed of a combustion chamber of a solid oxide fuel cell (SOFC) system, wherein the image to be processed is acquired by a camera disposed in the combustion chamber of the SOFC system; A first enhancement unit is used to perform image enhancement on the image to be processed by using a perception-based variational framework (PIVF) to obtain a first image, so as to extract main colors; A second enhancement unit is used to perform image enhancement on the first image through a multi-scale homomorphic filtering MSRCP process with chroma preservation to obtain a second image so as to improve color contrast; a third enhancement unit, configured to perform grayscale processing on the first image to obtain a third image so as to extract brightness information; a first fusion unit, configured to fuse the first image, the second image and the third image to obtain a fourth image so as to combine grayscale information and color information; a second fusion unit, configured to fuse the first image, the second image, and the fourth image to obtain a fifth image, so as to correct the fourth image by using both the first image and the second image; The PIVF treatment includes: Obtaining a weight matrix of the image size of the image to be processed, wherein in the weight matrix, points closer to the center point are given greater weights; Normalizing the image to be processed, obtaining an image polynomial, and performing Fourier transform on the image polynomial; Perform element-wise multiplication using the weight matrix and the image polynomial; Performing an inverse Fourier transform on the image polynomial; Image calculation is performed using pre-calculated polynomial coefficients to obtain the first image.

8. A processing device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and the processor executes the method according to any one of claims 1 to 6 when calling the computer program in the memory.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the method according to any one of claims 1 to 6.

Images