A method for flame propagation image recognition on ignition chamber jet ignition

Abstract

The application relates to the technical field of ignition chamber jet ignition research, in particular to a flame propagation picture recognition processing method for ignition chamber jet ignition. The method comprises the following steps: acquiring a background picture and extracting a to-be-calculated shadow picture; extracting the to-be-calculated shadow picture to obtain a rough extraction picture; deleting pixel points outside a visual range to obtain a fine extraction picture; determining a flame connected region used for calculating the flame speed and the flame filling area of the flame at a moment before the flame touches the bottom of a view window; and calculating the flame speed and the flame filling area of the flame at the moment before the flame touches the bottom of the view window and calculating the flame speed and the flame filling area of the flame at a moment after the flame touches the bottom of the view window. The application solves the problem that in the middle and late stages of jet ignition flame combustion, a particularly bright area appears at the bottom of the view window, the pixels of the bright area are lost, the calculated flame filling area is smaller than the actual value, and the flame area cannot be calculated.

Description

Technical Field

[0001] This application relates to the field of ignition chamber jet ignition technology, and in particular to a method for identifying and processing flame propagation images related to ignition chamber jet ignition. Background Technology

[0002] Turbulent high-temperature jet ignition has the potential to solve the problems of ignition difficulties and slow flame propagation speed faced by lean-burn ignition, and can achieve active control of ignition phase, making it an important research topic for the foreseeable future. To deepen the understanding of the ignition mechanism and influencing factors of turbulent high-temperature jets, many scholars and research institutions have conducted studies using shadow images captured by constant-volume combustion bombs, providing necessary and strong support for its application and optimization in internal combustion engines. Combustion in the engine's main combustion chamber needs to be determined from aspects such as ignition location, ignition distribution, and flame propagation. Therefore, analyzing changes in flame ignition location, flame penetration distance, and flame propagation speed through shadow images is an important method for studying the ignition of turbulent high-temperature jets.

[0003] However, current technologies have the following shortcomings in calculating penetration distance and flame area:

[0004] 1. The method of manually identifying and measuring the flame penetration distance is subject to significant errors due to interference from factors such as the thickness of the flame surface, and is also inefficient. When studying the flame filling area of ​​the ignition flame, manual intervention is ineffective. Furthermore, the capture of key information such as the jet ignition time is slow and prone to errors due to human factors.

[0005] 2. Traditionally, when using MATLAB programs to process the flame, a particularly bright area appears at the bottom of the viewport in the middle and later stages of the flame combustion ignited by the jet. This results in the loss of pixels in the bright area, which in turn causes the calculated flame filling area to be smaller than the actual value. Moreover, the error increases as the process progresses.

[0006] 3. It is impossible to use the same program to process images of a complete jet ignition and combustion process. The development of a jet ignition flame involves several stages, making it difficult to calculate accurate flame penetration velocity and flame filling area at different times using the same program. Summary of the Invention

[0007] This application provides a method for identifying and processing images of flame propagation ignited by jet ignition in an ignition chamber, which can solve the problems of low efficiency, large measurement error, and inconvenience in existing technologies for calculating penetration distance and flame area.

[0008] The technical solution of this application is a method for identifying and processing flame propagation images of ignition chamber jet ignition, implemented based on a MATLAB program. The method includes:

[0009] S1: Obtain the background image, and obtain several original shadow images related to the background image and arranged in sequence, and extract the first original shadow image in the sequence as the shadow image to be calculated;

[0010] S2: Extract the shadow image to be calculated based on the background image to obtain a coarse extracted image;

[0011] S3: Determine the visualization range in the coarsely extracted image, delete pixels outside the visualization range in the coarsely extracted image to obtain the finely extracted image;

[0012] S4: Based on the refined image, determine the flame connectivity region used to calculate the flame velocity and flame filling area before the flame reaches the bottom of the window after ignition;

[0013] S5: Based on the flame connection area, calculate the flame velocity and flame filling area of ​​the flame before it touches the bottom of the window after ignition, and calculate the flame velocity and flame filling area of ​​the flame after it touches the bottom of the window after ignition.

[0014] S6: Sequentially extract the next original shadow image in the sequential arrangement as the shadow image to be calculated, and repeat steps S2 to S6 until all original shadow images have been processed.

[0015] Optionally, step S1 includes:

[0016] S11: Obtain a folder, the folder including: a background image, and several original shadow images arranged sequentially after the background image and related to the background image;

[0017] S12: Use the imread function in the MATLAB program to read the folder and extract the first original shadow image in the sequential arrangement as the shadow image to be calculated.

[0018] Optionally, step S2 includes:

[0019] S21: Use the mat2gray function to perform grayscale processing on the background image and the image to be calculated for the shadow, and obtain a grayscale background image and a grayscale shadow image;

[0020] S22: Based on the grayscale background image, for the grayscale shadow image, the following preliminary processing is performed in sequence: background removal using the imsubtract function, brightness adjustment using the imadjust function, binarization using the im2bw function, island removal using the bwareaopen function, bright spot removal using the bwmorph function, edge object removal using the imclearboder function, and flame boundary detection using the edge function based on the Canny operator, to obtain the coarsely extracted image.

[0021] Optionally, step S3 includes:

[0022] S31: Determine the actual size of the coarsely extracted image;

[0023] S32: Establish a coordinate system for the coarsely extracted image and determine the coordinate information of the coarsely extracted image, as well as determine the scale based on the actual size and coordinate information;

[0024] S33: Determine the visualization range in the coarsely extracted image, and delete pixels outside the visualization range in the coarsely extracted image based on coordinate information and scale to obtain the finely extracted image.

[0025] Optionally, step S32 includes:

[0026] S321: The upper left corner of the coarsely extracted image is taken as the origin of the coordinate system, the direction of the jet flame penetration is taken as the direction of the vertical axis extension, and the direction perpendicular to the direction of the vertical axis extension is taken as the direction of the horizontal axis extension.

[0027] S322: Determine the coordinate information of the coarsely extracted image based on the coordinate system;

[0028] S323: Determine the scale based on the actual size and coordinate information of the roughly extracted image.

[0029] Optionally, step S4 includes:

[0030] S41: For the refined image, perform the following operations in sequence: dilation operation using the strel and imdilate functions, gap filling using the imfill function, smoothing using the strel function combined with the diamond function, and value filtering using the medfilt function to obtain the refined image;

[0031] S42: For finely processed images, the bwlabel function is used to determine the connected regions, and the regionprops function is used to determine the area of ​​each connected region based on the connected regions.

[0032] S43: The find function is used to determine the confirmation area in the finely processed image that is smaller than a preset specified value, and the ismember function is used to determine the flame connection area based on the connection area and the confirmation area, which is used to calculate the flame velocity and flame filling area of ​​the flame before it touches the bottom of the window after ignition.

[0033] Optionally, step S5 includes:

[0034] S51: Perform background boundary circle processing on the background image to obtain a background circle image;

[0035] S52: Based on the coordinate system, the flame connected region and the background circle image are superimposed, and the gap is filled by the imfill function to obtain the preliminary superimposed image;

[0036] S53: Extract the preliminary overlay image based on the background circle image to obtain the overlay extracted image;

[0037] S54: Based on the superimposed and extracted images, calculate the flame velocity and flame-filled area of ​​the flame before it touches the bottom of the viewport after ignition, and calculate the flame velocity and flame-filled area of ​​the flame after it touches the bottom of the viewport after ignition.

[0038] Optionally, step S51 includes:

[0039] S511: For a grayscale background image, perform the following steps in sequence: brightness adjustment and black-and-white swapping using the imadjust function; brightness adjustment using the imadjust function; binarization using the im2bw function; island removal using the bwareaopen function; bright spot removal using the bwmorph function; and boundary circle processing based on edge detection using the edge function using the Canny operator, to obtain a background circle image.

[0040] Beneficial effects:

[0041] (1) This application uses a MATLAB program to process the shadow image of the jet ignition flame in the ignition chamber. Compared with the traditional manual identification method, it has the characteristics of fast processing speed and high data recognition accuracy. It also provides a MATLAB processing method and detailed process for batch processing of jet ignition flame shadow images, which can provide some guidance or reference for image processing methods.

[0042] (2) This application solves the problem that in the middle and late stages of flame combustion ignited by jet, a particularly bright area appears at the bottom of the window, which leads to the loss of pixels in the bright area, resulting in the calculated flame filling area being smaller than the actual value, or even the flame area being unable to be calculated.

[0043] In summary, this application can solve the problems of low work efficiency, large measurement error and inconvenience in the existing technology for calculating penetration distance and flame area. It enables batch processing of shadow images of different stages of the jet ignition process using the same program, and obtains more accurate flame propagation speed and flame filling area of ​​the jet flame in the main combustion chamber. At the same time, it provides certain technical support for the study of the change law of the ignition flame in the main combustion chamber. Attached Figure Description

[0044] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 shows a set of shadow images obtained by extracting flames using existing techniques;

[0046] Figure 2 This is a flowchart illustrating the image recognition and processing method for flame propagation ignition via jet ignition in Example 1.

[0047] Figure 3 shows an example of the background image and the original shadow image in Example 1;

[0048] Figure 4 This is an example image of a grayscale background image from Example 1;

[0049] Figure 5 This is an example image of a grayscale shadow image from Example 1;

[0050] Figure 6 This is an example image obtained after background removal from a grayscale shadow image in Example 1;

[0051] Figure 7 This is an example image obtained by adjusting the brightness of an image after background removal in Example 1;

[0052] Figure 8 This is an example image obtained by binarizing the image after brightness adjustment in Example 1;

[0053] Figure 9 This is an example image obtained by removing isolated islands from a binarized image in Example 1.

[0054] Figure 10 This is an example image obtained by removing bright spots from an image that has undergone island removal processing in Example 1;

[0055] Figure 11 This is an example image obtained by removing edge objects from an image after removing bright spots, as shown in Example 1.

[0056] Figure 12 This is an example image of the coarsely extracted image in Example 1;

[0057] Figure 13 This is an example diagram of establishing a coordinate system based on the coarsely extracted image in Example 1;

[0058] Figure 14 This is a schematic diagram showing the scale confirmed in Example 1;

[0059] Figure 15 This is an example image of the refined image extracted in Example 1;

[0060] Figure 16 This is an example image obtained after performing a dilation operation on the finely extracted image in Example 1;

[0061] Figure 17 This is an example image obtained by filling gaps in an image that has undergone dilation in Example 1.

[0062] Figure 18 This is an example image obtained after smoothing the image after gap filling in Example 1;

[0063] Figure 19 This is an example image of the finely processed image in Example 1;

[0064] Figure 20 This is an example diagram of the flame connection area in Embodiment 1;

[0065] Figure 21 This is an example image of the background circle image in Embodiment 1;

[0066] Figure 22 This is an example image obtained by overlaying the flame connectivity region and the background circular image in Example 1;

[0067] Figure 23 This is an example image of the initial overlay of images in Example 1;

[0068] Figure 24 This is an example image of the overlay extraction of images in Example 1;

[0069] Figure 25 This is an example image set illustrating the image recognition and processing process before the flame touches the viewing window in Example 2;

[0070] Figure 26 This is an example image set showing the image recognition and processing process in the later stage of combustion, after the flame touches the viewing window and a bright light appears at the bottom, as described in Example 3.

[0071] Figure 27 This is a schematic diagram showing the comparison of pixel penetration distance and pixel flame area of ​​the jet flame igniting the flame over time in the embodiments of this application;

[0072] Figure 28 for Figure 27 A partial schematic diagram. Detailed Implementation

[0073] The embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described below do not represent all embodiments consistent with this application. They are merely examples of systems and methods consistent with some aspects of this application as detailed in the claims.

[0074] The jet ignition process in the ignition chamber can be divided into three stages: the squeezing stage, the jet stage, and the combustion process in the main combustion chamber. Among these, the second and third stages, based on constant-volume incendiary bombs, are the key stages for jet ignition research.

[0075] Some problems arise when using MATLAB to process shadow images to extract flame information, as shown in Figure 1. Figure 1 shows a group of shadow images obtained by extracting flames using existing techniques. Figures 1(a) and 1(b) are the shadow images at the ignition moment of the main combustion chamber and the images after background removal, respectively. It can be seen from the figures that due to the influence of the first and second stages, many bright squeezed gases or combustion products from the pre-combustion chamber are gathered at the bottom of the window. Figures 1(c) and 1(d) are the flame filling area processing results in the middle and late stages of combustion in the main combustion chamber. It can be seen that there are different degrees of missing areas at the bottom of the window. This is because bright areas of different sizes appear at the bottom of the window, resulting in the loss of pixel recognition.

[0076] In summary, it can be seen that due to the different characteristics of flame images in the jet ignition process, it is currently impossible to use a unified program to batch process the shadow images of the entire process. The problem of pixel loss caused by window brightness in the middle and later stages of combustion urgently needs to be solved.

[0077] To address the aforementioned issues, this application provides a method for identifying and processing images of flame propagation during jet ignition in an ignition chamber. This method enables batch processing of shadow images from different stages of the jet ignition process using the same procedure, resulting in more accurate flame propagation speed and flame filling area of ​​the jet flame in the main combustion chamber. It also provides technical support for studying the variation patterns of the ignition flame in the main combustion chamber.

[0078] Example 1

[0079] Example 1 is used for image recognition processing after a flame touches the bottom of the viewport.

[0080] This application provides a method for identifying and processing flame propagation images related to jet ignition in an ignition chamber, such as... Figure 2 As shown, Figure 2 This is a flowchart illustrating the flame propagation image recognition and processing method for ignition chamber jet ignition in Example 1, including:

[0081] S1: Obtain the background image, and obtain several original shadow images related to the background image and arranged in sequence, and extract the first original shadow image in the sequence as the shadow image to be calculated.

[0082] Specifically, taking the shadow image of the main combustion chamber flame after it touches the bottom of the viewing window in the third stage as an example, the flame filling area is calculated. The penetration distance after the flame touches the bottom of the viewing window remains unchanged. The same procedure is used to calculate the penetration distance and flame filling area before the flame touches the bottom of the viewing window.

[0083] Step S1 includes:

[0084] S11: Get the folder, which includes: the background image, and several original shadow images arranged in sequence after the background image and related to the background image.

[0085] Specifically, prepare a folder of images showing the shadows ignited by the jet flame for batch calculation, where the first image in the folder should be the background image. Use the `imread` function in MATLAB to read the prepared folder; then use a `for` loop to start the calculations sequentially from the second image, i.e., the original shadow image.

[0086] As shown in Figure 3, Figure 3 is an example of the background image and the original shadow image in Example 1. Figure 3(a) is the background image and Figure 3(b) is the original shadow image.

[0087] S12: Use the imread function in the MATLAB program to read the folder and extract the first original shadow image in the sequential arrangement as the shadow image to be calculated.

[0088] S2: Extract the shadow image to be calculated based on the background image to obtain a coarse extracted image.

[0089] Specifically, the outline information of the ignition flame is initially extracted.

[0090] Step S2 includes:

[0091] S21: Use the mat2gray function to perform grayscale processing on the background image and the image to be calculated for the shadow, resulting in a grayscale background image and a grayscale shadow image.

[0092] Specifically, the mat2gray function is used to perform grayscale processing on the background image and the image to be calculated for shadow, thereby achieving the normalization operation of the image matrix. For example... Figure 4 and Figure 5 As shown, Figure 4 5 is an example image of a grayscale background image in Example 1, and 6 is an example image of a grayscale shadow image in Example 1.

[0093] S22: Based on the grayscale background image, for the grayscale shadow image, the following preliminary processing is performed in sequence: background removal using the imsubtract function, brightness adjustment using the imadjust function, binarization using the im2bw function, island removal using the bwareaopen function, bright spot removal using the bwmorph function, edge object removal using the imclearboder function, and flame boundary detection using the edge function based on the Canny operator, to obtain the coarsely extracted image.

[0094] Specifically, the `imsubtract` function is used to perform image segmentation on the grayscale shadow image, i.e., to remove the background. This involves subtracting the grayscale background image from the grayscale shadow image. The result is as follows: Figure 6 As shown, Figure 6 This is an example image obtained after background removal from a grayscale shadow image in Example 1.

[0095] The `imadjust` function is used to adjust the brightness of an image obtained after background removal, typically to enhance brightness. Figure 7 As shown, Figure 7 This is an example image obtained by adjusting the brightness of the image after background removal in Example 1.

[0096] The `im2bw` function is used to binarize the image obtained after brightness adjustment. This involves a grayscale threshold value, which needs to be adjusted according to the image detail requirements throughout the calculation process. Figure 8 As shown, Figure 8 This is an example image obtained after binarizing the image after brightness adjustment in Example 1.

[0097] The `bwareaopen` function is used to remove islands from the binarized image. A value needs to be set so that bright areas smaller than this value are removed. This value may need to be adjusted based on the specific image processing requirements. In this case, the value is set to 5. Figure 9 As shown, Figure 9 This is an example image obtained by removing isolated islands from a binarized image in Example 1.

[0098] The bwmorph function is used to remove bright spots from the image obtained after island removal processing, such as... Figure 10 As shown, Figure 10 This is an example image obtained by removing bright spots from an image that has undergone island removal processing in Example 1.

[0099] The `imclearboder` function is used to remove edge objects from the image obtained after removing highlights, such as... Figure 11 As shown, Figure 11 This is an example image obtained by removing edge objects from an image after removing highlights, as shown in Example 1.

[0100] Flame boundary detection is performed using the edge function based on the Canny operator. The Canny operator has stronger filtering capabilities compared to other operators, enabling it to detect more edge details. Users can select different operators according to their needs, such as... Figure 12 As shown, Figure 12 This is an example image of the coarse image extraction in Example 1.

[0101] S3: Determine the visualization range in the coarsely extracted image, delete pixels outside the visualization range in the coarsely extracted image, and obtain the finely extracted image.

[0102] Specifically, since each pixel in the coarsely extracted image has its own coordinates, in order to facilitate further image processing, it is necessary to determine relevant information about the pixel coordinate system, such as the resolution of the shadow image and the pixel coordinate range of the circular viewport.

[0103] Step S3 includes:

[0104] S31: Determine the actual size of the coarsely extracted image.

[0105] S32: Establish a coordinate system for the coarsely extracted image and determine the coordinate information of the coarsely extracted image, as well as determine the scale based on the actual size and coordinate information.

[0106] Step S32 includes:

[0107] S321: The top left corner of the coarsely extracted image is taken as the origin of the coordinate system, the direction of the jet flame penetration is taken as the direction of the vertical axis extension, and the direction perpendicular to the vertical axis extension is taken as the direction of the horizontal axis extension.

[0108] Specifically, because the pixel coordinate system remains the same throughout the entire image processing process, taking a background grayscale image as an example, the relevant information about the pixel coordinate system is determined, such as... Figure 13 As shown, Figure 13As an example of establishing a coordinate system based on the coarsely extracted image in Embodiment 1, the range of the pixel coordinate axis can be obtained as 836×488 pixels. Taking the upper left corner of the image as the origin O of the planar coordinate system, the direction of the jet flame penetration is the vertical coordinate axis y, that is, the vertical pixel direction, and the other vertical axis is the horizontal coordinate axis x, that is, the horizontal pixel direction.

[0109] S322: Determine the coordinate information of the coarsely extracted image based on the coordinate system.

[0110] S323: Determine the scale based on the actual size and coordinate information of the roughly extracted image.

[0111] Specifically, the scale K represents the ratio between the actual size captured by the camera and the actual pixel count of the shadow image. The method involves placing a known-sized scale on a circular glass window and capturing the shadow image using a camera. Figure 14 As shown, Figure 14 This is a schematic diagram of the scale confirmation in Example 1. The scale measures the distance TB between points T and B in the vertical direction, and simultaneously calculates the pixel positions T1 and B1 corresponding to the pixel images T and B, and calculates the pixel distance T1B1.

[0112] Determine the scale K: K = S / S0, where S is the scale diameter TB and S0 is the scale image diameter T1B1 in the shaded image.

[0113] S33: Determine the visualization range in the coarsely extracted image, and delete pixels outside the visualization range in the coarsely extracted image based on coordinate information and scale to obtain the finely extracted image.

[0114] Specifically, based on the coordinate information provided in step S32, pixels outside the visualization range, i.e., outside the circular viewport, such as some text annotations, are deleted. Figure 15 As shown, Figure 15 This is an example image of the refined image extracted in Example 1.

[0115] S4: Based on the refined image, determine the flame connectivity region used to calculate the flame velocity and flame filling area before the flame reaches the bottom of the window after ignition.

[0116] Step S4 includes:

[0117] S41: For the refined image, perform the following operations in sequence: dilation using the strel and imdilate functions, gap filling using the imfill function, smoothing using the strel function combined with the diamond function, and value filtering using the medfilt function to obtain the refined image.

[0118] Specifically, the `strel` and `imdilate` functions are used to dilate the refined image, filling in the gaps at the edges, such as... Figure 16 As shown, Figure 16 This is an example image obtained after performing a dilation operation on the finely extracted image in Example 1.

[0119] The `imfill` function is used to fill gaps in an image obtained after dilation, such as... Figure 17 As shown, Figure 17 This is an example image obtained by filling gaps in an image after the dilation operation in Example 1.

[0120] The `strel` function, combined with a diamond shape, is used to smooth the image after gap filling, such as... Figure 18 As shown, Figure 18 This is an example image obtained after smoothing the image after gap filling in Example 1.

[0121] The medfilt function is used to perform value filtering on the smoothed image to obtain a refined image, such as... Figure 19 As shown, Figure 19 This is an example image of the finely processed image in Example 1.

[0122] S42: For finely processed images, the bwlabel function is used to determine the connected regions, and the regionprops function is used to determine the area of ​​each connected region based on the connected regions.

[0123] Specifically, the bwlabel function is used to find connected components, and the regionprops function is used to locate and display the area of ​​each connected component.

[0124] S43: The find function determines the confirmation area in the finely processed image that is smaller than a preset specified value, and the ismember function determines the flame connection area based on the connection area and the confirmation area, which is used to calculate the flame velocity and flame filling area of ​​the flame before it reaches the bottom of the window after ignition.

[0125] Specifically, the `find` function is used to find regions smaller than a specified value, and the `ismember` function is used to determine if a connected element is in the set. This step is used to delete isolated areas smaller than the specified value, ultimately obtaining the connected flame area regions, such as... Figure 20 As shown, Figure 20 This is an example diagram of the flame connection area in Embodiment 1.

[0126] The purpose of setting a specific value is to provide more specific guidance for deleting isolated areas, such as... Figure 19 As shown, Figure 19The middle section has a larger connected area located at the top, designated as A1. Figure 19 It also has smaller areas located on the bottom, designated as A2, A3, etc., so the size of the specified value needs to be smaller than A1, but larger than A2, A3 and other smaller areas.

[0127] S5: Based on the flame connection area, calculate the flame velocity and flame filling area of ​​the flame before it touches the bottom of the window after ignition, and calculate the flame velocity and flame filling area of ​​the flame after it touches the bottom of the window after ignition.

[0128] Specifically, the flame speed is calculated by multiplying the penetration distance pixels at different times by the scale K, and the flame filling area is calculated by multiplying the sum of the pixels in the connected flame area region at different times by the square of the scale K.

[0129] After the flame touches the bottom of the window after ignition, the flame propagation distance no longer changes, and the flame filling area continues to increase because the pixels in the brighter area at the bottom of the window cannot be identified in the later stages of combustion in the main combustion chamber.

[0130] Step S5 includes:

[0131] S51: Perform background boundary circle processing on the background image to obtain a background circle image.

[0132] Step S51 includes:

[0133] S511: For a grayscale background image, perform the following steps in sequence: brightness adjustment and black-and-white swapping using the imadjust function; brightness adjustment using the imadjust function; binarization using the im2bw function; island removal using the bwareaopen function; bright spot removal using the bwmorph function; and boundary circle processing based on edge detection using the edge function using the Canny operator, to obtain a background circle image.

[0134] Specifically, the brightness of the resulting image is adjusted using imadjust, and the function is also used to swap the black and white areas in the image.

[0135] By binarizing the resulting image, removing isolated islands, removing bright spots, and performing edge detection using the Canny operator, a circular image, namely the viewport boundary circle, can be obtained.

[0136] The image obtained after edge detection is processed by deleting pixels at a certain distance from the origin towards the jet penetration direction (the y-axis), leaving only a 4 / 5 circle of the boundary circle. This results in the processed boundary circle. Figure 21As shown, Figure 21 This is an example image of the background circle image in Embodiment 1.

[0137] S52: Based on the coordinate system, the flame connected region and the background circle image are superimposed, and the gap is filled by the imfill function to obtain the preliminary superimposed image.

[0138] Specifically, such as Figure 22 and Figure 23 As shown, Figure 22 This is an example image obtained by overlaying the flame connectivity region and the background circular image in Example 1. Figure 23 This is an example image of the initial overlay of images in Example 1.

[0139] S53: Extract the initial overlay image based on the background circular image to obtain the overlay extracted image.

[0140] Specifically, by subtracting the background circle image from the initial overlay image, the flame-filled area can be obtained, such as... Figure 24 As shown, Figure 24 This is an example image of the overlay extraction of images in Example 1.

[0141] S54: Based on the superimposed and extracted images, calculate the flame velocity and flame-filled area of ​​the flame before it touches the bottom of the viewport after ignition, and calculate the flame velocity and flame-filled area of ​​the flame after it touches the bottom of the viewport after ignition.

[0142] Specifically, the flame speed is as described in the aforementioned scheme, and will not be repeated here.

[0143] Summing the pixels in the flame-filled area and then multiplying by the square of the scale K, we obtain the flame-filled area after the flame touches the bottom of the viewport.

[0144] S6: Sequentially extract the next original shadow image in the sequential arrangement as the shadow image to be calculated, and repeat steps S2 to S6 until all original shadow images have been processed.

[0145] Specifically, by using a for loop, steps S2 to S6 can be used to calculate the flame penetration distance and flame filling speed during the entire flame propagation and combustion process, thus solving the problem of bright light appearing at the bottom of the window in the later stages of combustion, making it impossible to identify pixels.

[0146] Using a fixed threshold set in the preliminary tests, the `xlswrite` function was used to write the flame penetration distance and flame-filled area at different times into an Excel spreadsheet. Since the penetration speed cannot be directly obtained and requires further processing of the penetration distance, only the penetration distance was recorded. Throughout the calculation process, many function statements have threshold settings. These can be continuously adjusted from the default values ​​to meet the user's image quality requirements, and then the calculations are iteratively performed based on the set values.

[0147] In summary, this application solves the problem of pixel recognition failure due to bright light appearing at the bottom of the viewing window during the later stages of combustion by utilizing a background circle with a notch. It also addresses the inability to process a complete jet ignition and combustion process using the same procedure. Key points are as follows:

[0148] 1. The boundary circle is obtained from the background image; 2. The boundary circle of the background image is processed by using the imadjust function to swap the black and white in the image; 3. An opening is made from the top of the background boundary circle, leaving only a 4 / 5 circle; 4. The pixels of the processed boundary circle and the image to be processed are added together, and the imfill function is used to fill the gap; 5. The area of ​​the boundary circle pixels is deleted.

[0149] Example 2

[0150] Example 2 is used for image recognition processing before the flame touches the viewing window.

[0151] like Figure 25 As shown, Figure 25 This is a set of example images showing the image recognition and processing process before the flame touches the viewing window in Example 2.

[0152] Example 3

[0153] Example 3 is used for image recognition and processing of the later stage of combustion, where the flame touches the viewing window and a bright light appears at the bottom.

[0154] like Figure 26 As shown, Figure 26 This is an example image set illustrating the image recognition and processing process in Example 3, showing the later stage of combustion after the flame touches the viewing window and a noticeable bright light appears at the bottom.

[0155] The above three examples analyze the processing capability of using the same program to process shadow images of the jet ignition process across the entire operating range.

[0156] like Figure 27 As shown, Figure 27 This is a schematic diagram comparing the pixel penetration distance and pixel flame area of ​​the jet flame igniting the flame over time in an embodiment of this application. Figure 27As can be seen, the horizontal axis represents the camera's shooting time, the black square scatter plot represents the penetration distance, and the penetration distance remains unchanged when the flame reaches the bottom of the window; the black circle scatter plot represents the flame area, and the curve represents the reference curve of the penetration distance curve before processing, that is, the step of processing the flame filling area after the flame touches the bottom of the window after ignition without measuring and calculating. It can be seen that the development of the flame can be seen. The penetration distance value of this curve decreases, indicating that the light at the bottom of the window is detected.

[0157] like Figure 28 As shown, Figure 28 for Figure 27 A partial schematic diagram, from Figure 28 As can be seen from this, several cases of flame propagation and changes in flame filling area are consistent with actual flame propagation, which fully demonstrates the scientific nature and feasibility of the embodiments of this application.

[0158] The embodiments of this application have been described in detail above, but the content is only a preferred embodiment of this application and should not be considered as limiting the scope of this application. All equivalent changes and improvements made within the scope of this application should still fall within the patent coverage of this application.

Claims

1. A method for identifying and processing images of flame propagation during jet ignition in an ignition chamber, characterized in that, The method, implemented according to a MATLAB program, includes: S1: Obtain the background image, and obtain several original shadow images related to the background image and arranged in sequence, and extract the first original shadow image in the sequence as the shadow image to be calculated; S2: Extract the shadow image to be calculated based on the background image to obtain a coarse extracted image; S3: Determine the visualization range in the coarsely extracted image, delete pixels outside the visualization range in the coarsely extracted image to obtain the finely extracted image; S4: Based on the refined image, determine the flame connectivity region used to calculate the flame velocity and flame filling area before the flame reaches the bottom of the window after ignition; S5: Based on the flame connection area, calculate the flame velocity and flame filling area of ​​the flame before it touches the bottom of the window after ignition, and calculate the flame velocity and flame filling area of ​​the flame after it touches the bottom of the window after ignition. S6: Sequentially extract the next original shadow image in the sequential arrangement as the shadow image to be calculated, and repeat steps S2 to S6 until all original shadow images have been processed. Step S4 includes: S41: For the refined image, perform the following operations in sequence: dilation operation using the strel and imdilate functions, gap filling using the imfill function, smoothing using the strel function combined with the diamond function, and value filtering using the medfilt function to obtain the refined image; S42: For finely processed images, the bwlabel function is used to determine the connected regions, and the regionprops function is used to determine the area of ​​each connected region based on the connected regions. S43: The find function is used to determine the confirmation area in the finely processed image that is smaller than a preset specified value, and the ismember function is used to determine the flame connection area based on the connection area and the confirmation area, which is used to calculate the flame velocity and flame filling area of ​​the flame before it touches the bottom of the window after ignition. Step S5 includes: S51: Perform background boundary circle processing on the background image to obtain a background circle image; S52: Based on the coordinate system, the flame connected region and the background circle image are superimposed, and the gap is filled by the imfill function to obtain the preliminary superimposed image; S53: Extract the preliminary overlay image based on the background circle image to obtain the overlay extracted image; S54: Based on the superimposed and extracted images, calculate the flame velocity and flame-filled area of ​​the flame before it touches the bottom of the viewport after ignition, and calculate the flame velocity and flame-filled area of ​​the flame after it touches the bottom of the viewport after ignition.

2. The method for identifying and processing flame propagation images of ignition chamber jet ignition according to claim 1, characterized in that, Step S1 includes: S11: Obtain a folder, the folder including: a background image, and several original shadow images arranged sequentially after the background image and related to the background image; S12: Use the imread function in the MATLAB program to read the folder and extract the first original shadow image in the sequential arrangement as the shadow image to be calculated.

3. The method for identifying and processing flame propagation images of ignition chamber jet ignition according to claim 1, characterized in that, Step S2 includes: S21: Use the mat2gray function to perform grayscale processing on the background image and the image to be calculated for the shadow, and obtain a grayscale background image and a grayscale shadow image; S22: Based on the grayscale background image, for the grayscale shadow image, the following preliminary processing is performed in sequence: background removal using the imsubtract function, brightness adjustment using the imadjust function, binarization using the im2bw function, island removal using the bwareaopen function, bright spot removal using the bwmorph function, edge object removal using the imclearboder function, and flame boundary detection using the edge function based on the Canny operator, to obtain the coarsely extracted image.

4. The method for identifying and processing flame propagation images of ignition chamber jet ignition according to claim 1, characterized in that, Step S3 includes: S31: Determine the actual size of the coarsely extracted image; S32: Establish a coordinate system for the coarsely extracted image and determine the coordinate information of the coarsely extracted image, as well as determine the scale based on the actual size and coordinate information; S33: Determine the visualization range in the coarsely extracted image, and delete pixels outside the visualization range in the coarsely extracted image based on coordinate information and scale to obtain the finely extracted image.

5. The method for identifying and processing flame propagation images of ignition chamber jet ignition according to claim 4, characterized in that, Step S32 includes: S321: The upper left corner of the coarsely extracted image is taken as the origin of the coordinate system, the direction of the jet flame penetration is taken as the direction of the vertical axis extension, and the direction perpendicular to the direction of the vertical axis extension is taken as the direction of the horizontal axis extension. S322: Determine the coordinate information of the coarsely extracted image based on the coordinate system; S323: Determine the scale based on the actual size and coordinate information of the roughly extracted image.

6. The method for identifying and processing flame propagation images of ignition chamber jet ignition according to claim 1, characterized in that, Step S51 includes: S511: For a grayscale background image, perform the following steps in sequence: brightness adjustment and black-and-white swapping using the imadjust function; brightness adjustment using the imadjust function; binarization using the im2bw function; island removal using the bwareaopen function; bright spot removal using the bwmorph function; and boundary circle processing based on edge detection using the edge function using the Canny operator, to obtain a background circle image.

Images