Combustion efficiency detection method of wood-fired energy-saving stove based on image processing

Abstract

The present invention relates to the technical field of image data processing, and more specifically, to a method for detecting the combustion efficiency of a wood-fired energy-saving stove based on image processing. The method comprises: obtaining similar pixel points of a pixel point in a flame grayscale image in eight neighborhood directions, and obtaining a similarity range value of the pixel point based on the Euclidean distance between the pixel point and each corresponding similar pixel point; calculating the probability that the pixel point in the flame grayscale image is an inner flame area; determining the inner flame area of the flame grayscale image by the probability that the pixel point in the flame grayscale image is an inner flame area, and extracting the outer flame area of the flame grayscale image by constructing a grayscale co-occurrence matrix to obtain the flame area; obtaining the combustion efficiency of the wood-fired energy-saving stove by the rectangularity of the flame area of the flame grayscale image and the heat utilization efficiency of the stove body heat map, thereby effectively improving the accuracy of the obtained combustion efficiency of the wood-fired energy-saving stove.

Description

Technical Field

[0001] The present invention relates to the technical field of image data processing, and more particularly to a method for detecting the combustion efficiency of a wood-fired energy-saving stove based on image processing. Background Art

[0002] A wood-fired energy-saving stove is an energy-saving device whose core goal is to improve energy efficiency and reduce fuel consumption. By measuring the stove's combustion efficiency during the combustion process, we can ensure that the fuel is fully burned, avoiding waste. This also effectively reduces pollutant emissions (such as carbon monoxide and particulate matter). Therefore, during the research and testing phase of wood-fired stoves, combustion efficiency testing can accurately measure the stove's actual energy savings compared to traditional wood-fired equipment, allowing for subsequent research and improvement.

[0003] In the existing solution, the patent application document with publication number CN119131042A discloses an online monitoring and analysis method for the thermal efficiency of an industrial furnace based on infrared images. The method obtains a thermal image of the industrial furnace; obtains the segmented grayscale value of any pixel point in the thermal image on the three channels of R, G, and B; constructs a new grayscale image with the new grayscale values of all pixels; and segments the new grayscale image to obtain a segmentation image to monitor and analyze the thermal efficiency of the industrial furnace.

[0004] Existing solutions analyze thermal efficiency by acquiring a single thermal image. This method fails to accurately extract the dynamic characteristics of a wood-fired energy-saving stove's flame during combustion. The shape, size, intensity, and position of the flame vary significantly over time, resulting in low segmentation accuracy. Therefore, there is an urgent need to accurately extract the characteristics of a dynamically changing flame to accurately assess combustion efficiency. Summary of the Invention

[0005] To solve the above-mentioned technical problem of how to accurately extract the characteristics of a dynamically changing flame and thus accurately evaluate its combustion efficiency, the present invention proposes a method for detecting the combustion efficiency of a wood-fired energy-saving stove based on image processing, which includes the following steps:

[0006] Determine the grayscale peak value in the flame grayscale image; obtain similar pixel points in the flame grayscale image in the eight neighborhood directions, and obtain the similarity range value of the pixel point based on the Euclidean distance between the pixel point and the corresponding similar pixel points; calculate the The probability that the i-th pixel in the flame grayscale image is the inner flame area :

[0007] ;

[0008] For the Grayscale peak value in the frame flame grayscale image, 、 Respectively The grayscale value and similarity range value of the i-th pixel in the frame flame grayscale image, For the The variance of the similarity range value of the i-th pixel in the flame grayscale image of the frame and its two previous and next frames, is an exponential function with base e, is the linear normalization function, is the absolute value symbol; the inner flame area of the flame grayscale image is determined by the probability that the pixel point in the flame grayscale image is the inner flame area, and the outer flame area of the flame grayscale image is extracted by constructing the gray level co-occurrence matrix to obtain the flame area; the combustion efficiency of the wood-fired energy-saving stove is obtained by the flame area rectangularity of the flame grayscale image and the heat utilization efficiency of the stove body heat map.

[0009] The present invention accurately calculates the combustion efficiency of a wood-fired energy-saving stove by combining changes in flame shape with heat utilization efficiency. In acquiring the changes in flame shape, the present invention takes into account that flames typically exhibit dynamic changes and consist of an outer flame region and an inner flame region. A single-frame flame grayscale image alone cannot accurately capture the dynamic changes of the flame. Therefore, the present invention captures the characteristics of the large and high-brightness inner flame region and combines this with dynamic analysis of the area changes in successive image frames. This reduces interference from reflective areas on the furnace wall while accurately calculating the probability that each pixel in the flame grayscale image is in the inner flame region. This allows accurate determination of the inner flame region in the flame grayscale image, effectively improving the accuracy of the calculation of the combustion efficiency of a wood-fired energy-saving stove. Furthermore, the present invention constructs a gray-level co-occurrence matrix to extract texture homogeneity features from the flame grayscale image, accurately determining the outer flame region in the flame grayscale image. Furthermore, based on the inner and outer flame regions, the flame region in the flame grayscale image is accurately determined, effectively improving the accuracy of the calculation of the combustion efficiency of a wood-fired energy-saving stove.

[0010] According to the method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing provided by the present invention, the method of determining the grayscale peak value in the flame grayscale image also includes: obtaining the flame grayscale image and the furnace body heat map during the combustion process of the wood-fired energy-saving stove at each acquisition moment, and obtaining a flame image set consisting of multiple frames of continuous flame grayscale images.

[0011] The present invention takes into account that a single-frame image cannot accurately reflect the dynamic changes of the flame area, and therefore prepares for the subsequent dynamic change feature extraction of the flame area by acquiring continuous image frames.

[0012] According to the image processing-based combustion efficiency detection method for a wood-fired energy-saving stove provided by the present invention, the grayscale peak value in the flame grayscale image is determined, including: constructing a grayscale histogram of the flame grayscale image with the grayscale value of the pixel point in the flame grayscale image as the horizontal axis and the number of pixel points corresponding to the grayscale value as the vertical axis, with the lower left corner of the grayscale histogram being the coordinate origin; starting from the rightmost side of the grayscale histogram, traversing to determine the first local maximum value of the number of pixel points, and taking the grayscale value corresponding to the local maximum value as the grayscale peak value in the flame grayscale image; wherein the local maximum value is greater than the number of adjacent pixel points on the left and right sides in the grayscale histogram.

[0013] According to the image processing-based combustion efficiency detection method for wood-fired energy-saving stoves provided by the present invention, the obtaining of similar pixel points of a pixel point in the flame grayscale image in the eight-neighborhood directions includes: continuously obtaining adjacent pixel points of the pixel point by diffusing outward in the eight-neighborhood directions of the pixel point until the grayscale difference between the adjacent pixel points in the direction and the pixel point is greater than a first threshold, and the obtaining is stopped, and the last adjacent pixel point whose grayscale difference is not greater than the first threshold is used as the similar pixel point of the pixel point in the direction.

[0014] The present invention takes into account that the inner flame area of the flame area is located in the central area of the flame area, has high brightness and changes smoothly, and is usually large in area. Therefore, by analyzing the area of similar areas formed by the pixel point and the surrounding pixels with small grayscale differences, the probability of the pixel point being the inner flame area is accurately evaluated.

[0015] According to the image processing-based combustion efficiency detection method for wood-fired energy-saving stoves provided by the present invention, the similarity range value of the pixel point is obtained based on the Euclidean distance between the pixel point and the corresponding similar pixel points, including: taking the normalized mean of the Euclidean distances between the pixel point and the similar pixel points in the eight neighborhood directions as the similarity range value of the pixel point.

[0016] According to the image processing-based combustion efficiency detection method for wood-fired energy-saving stoves provided by the present invention, the inner flame area of the flame grayscale image is determined by the probability that the pixel point in the flame grayscale image is the inner flame area, including: taking the median of the probability that the pixel point in the flame grayscale image is the inner flame area as a comparison value, if the probability that the pixel point is the inner flame area is greater than or equal to the comparison value, then the pixel point is a pixel point in the inner flame area; using a region growing algorithm to process the pixel points in the inner flame area in the flame grayscale image to obtain the inner flame area in the flame grayscale image.

[0017] According to the image processing-based combustion efficiency detection method for wood-fired energy-saving stoves provided by the present invention, the outer flame area of the flame grayscale image is extracted by constructing a grayscale co-occurrence matrix, including: obtaining the grayscale mean of the adjacent pixel points along the gradient change direction of the edge pixel points of the inner flame area in the flame grayscale image as the reference value of the outer flame area; taking the adjacent pixel points whose grayscale values among the remaining pixel points except the inner flame area in the flame grayscale image have a difference with the reference value less than a second threshold as analysis pixel points, and constructing a grayscale co-occurrence matrix based on the analysis pixel points to obtain the outer flame area of the flame grayscale image.

[0018] The present invention takes into account that although the grayscale of the outer flame area of the flame grayscale image is mixed with the grayscale of the surrounding ash area, the texture of the outer flame area is uniform, while the texture of the ash area is more complex. Therefore, the homogeneity features are extracted by constructing a grayscale co-occurrence matrix, so as to accurately identify the outer flame area with uniform texture.

[0019] According to the image processing-based combustion efficiency detection method for wood-fired energy-saving stoves provided by the present invention, the outer flame area of the flame grayscale image is obtained by constructing a grayscale co-occurrence matrix based on the analysis pixels, including: setting distance and direction parameters to construct the grayscale co-occurrence matrix of each analysis pixel point, and obtaining the homogeneity value of each analysis pixel point; recording the analysis pixel points whose homogeneity values are greater than a third threshold as high-homogeneity pixels; and using a region growing algorithm to process the high-homogeneity pixels in the flame grayscale image to obtain the outer flame area in the flame grayscale image.

[0020] According to the image processing-based combustion efficiency detection method for a wood-fired energy-saving stove provided by the present invention, the method for obtaining the rectangularity of the flame area of a flame grayscale image and the heat utilization efficiency of a furnace body heat map includes: obtaining a flame image set in which the flame grayscale image is located; taking the ratio of the number of pixel points in the flame area to the area of the minimum circumscribed rectangle of the flame area as the rectangularity of the flame area; integrating the calorific value of the furnace body heat map corresponding to each frame of the flame grayscale image in the flame image set to obtain the heat output value of the wood-fired energy-saving stove; obtaining the theoretical calorific value of the wood-fired energy-saving stove based on the mass of the combustion material used in the process of acquiring the flame image set; and obtaining the heat utilization efficiency of the wood-fired energy-saving stove based on the difference between the heat output value of the wood-fired energy-saving stove and the theoretical calorific value.

[0021] According to the method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing provided by the present invention, the combustion efficiency of the wood-fired energy-saving stove is obtained by measuring the rectangularity of the flame area of the flame grayscale image and the heat utilization efficiency of the stove body heat map, including: calculating the combustion efficiency of the wood-fired energy-saving stove:

[0022] ; To improve the combustion efficiency of wood-fired energy-saving stoves, For the The rectangularity of the flame area of the frame flame grayscale image, is the number of flame grayscale image frames in the flame image set, To improve the heat utilization efficiency of wood-fired energy-saving stoves, is a linear normalization function.

[0023] The present invention provides an accurate method for calculating the combustion efficiency of a wood-fired energy-saving stove. Based on the characteristics that a flame with a higher rectangularity burns more stably and the fuel can more fully contact and burn with oxygen, combined with the heat utilization efficiency of the wood-fired energy-saving stove, the interference of external factors can be effectively reduced, thereby accurately obtaining the combustion efficiency value of the wood-fired energy-saving stove.

[0024] The present invention has the following beneficial effects:

[0025] Based on the above technical solution, the present invention provides an image processing-based method for detecting the combustion efficiency of a wood-fired energy-saving stove. When detecting the combustion efficiency of a wood-fired energy-saving stove, the present invention combines changes in the flame's appearance with heat utilization efficiency to accurately determine the combustion efficiency of the wood-fired energy-saving stove. In the process of obtaining changes in the flame's appearance, the present invention obtains the characteristics of the large and high brightness inner flame region and dynamically analyzes the area changes of consecutive image frames. This reduces interference from reflective areas on the stove wall while accurately calculating the probability that each pixel in the flame grayscale image is the inner flame region. This allows the inner flame region of the flame grayscale image to be accurately determined, effectively improving the accuracy of the calculation of the combustion efficiency of the wood-fired energy-saving stove. Furthermore, the present invention extracts texture homogeneity features from the flame grayscale image by constructing a gray-level co-occurrence matrix, thereby accurately determining the outer flame region of the flame grayscale image. Furthermore, based on the inner and outer flame regions, the flame region in the flame grayscale image is accurately determined, effectively improving the accuracy of the calculation results of the combustion efficiency of the wood-fired energy-saving stove. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flowchart of the steps of a method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing provided by an embodiment of the present invention;

[0027] Figure 2 This is an example of a flame grayscale image during the combustion process of a wood-fired energy-saving stove provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0029] In order to extract the dynamically changing flame features in the flame grayscale image and thus accurately obtain the combustion efficiency detection results of the wood-fired energy-saving stove, an embodiment of the present invention discloses a combustion efficiency detection method for a wood-fired energy-saving stove based on image processing. This method obtains the flame areas consisting of the inner flame area and the outer flame area by analyzing the characteristic changes of continuous flame grayscale images, thereby effectively improving the accuracy of the combustion efficiency detection results of the wood-fired energy-saving stove.

[0030] See also Figure 1 As shown, Figure 1 A flowchart of a method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing provided by an embodiment of the present invention, the method comprising the following steps:

[0031] S1: Obtain a flame grayscale image from a flame image set.

[0032] For example, in an embodiment of the present invention, obtaining a flame grayscale image in a flame image set includes: obtaining a flame grayscale image during the combustion of a wood-fired energy-saving stove at each acquisition moment, and obtaining a flame image set consisting of multiple frames of continuous flame grayscale images.

[0033] Specifically, based on the preset acquisition frequency, a high-definition camera is used to capture the combustion image of the wood-fired energy-saving stove body during the test phase. After preprocessing, the flame grayscale image is obtained. The multiple frames of continuous flame grayscale images collected within the preset period are combined into a flame image set, and the pixel points in different flame grayscale images correspond one to one.

[0034] Preprocessing can include grayscale processing, image denoising, image quality enhancement, highlighting regions of interest, and the like. These can be configured based on actual needs and are not limited in this embodiment of the present invention. The acquisition frequency can be set to 30 frames per second, and the preset period can be within 30 minutes after the wood-fired energy-saving stove begins burning. These can be configured based on actual needs.

[0035] For example, grayscale processing can be achieved through weighted averaging; adaptive filters can be used to remove random noise in the image to improve the signal-to-noise ratio; adaptive histogram equalization can be performed on the image to improve the contrast in the image; and sharpening technology can be used to enhance the details of the image.

[0036] For ease of processing, the sizes of the flame grayscale images at all acquisition moments can be adjusted to the same size, which can be set specifically according to actual needs.

[0037] It should be noted that, based on the above steps, the flame grayscale image and flame image set corresponding to each acquisition moment can be obtained. Figure 2 As shown, Figure 2 This is an example of a flame grayscale image during the combustion process of a wood-fired energy-saving stove provided by an embodiment of the present invention, combined with Figure 2 It can be seen that the flame area is composed of an inner flame area with higher brightness and an outer flame area with lower brightness. The inner flame area usually appears as a larger sheet area, while the outer flame area completely wraps the inner flame area. Based on this, the embodiment of the present invention can accurately extract the inner flame area and the outer flame area from the flame grayscale image according to their respective characteristics, that is, perform the following steps.

[0038] S2: Determine the grayscale peak value in the flame grayscale image; obtain the similarity range value of the pixel point based on the Euclidean distance between the pixel point and the corresponding similar pixels; calculate the probability that the pixel point in the flame grayscale image is the inner flame area based on the grayscale change and similarity range value of the pixel point between each frame of the flame grayscale image and the two frames before and after it.

[0039] It should be noted that if a pixel in the flame grayscale image is a pixel in the inner flame area, then in the continuous multi-frame image, the grayscale value of the pixel will be very close to the grayscale peak of the current flame grayscale image, that is, it has a higher brightness. Based on this feature, the pixel in the inner flame area can be preliminarily screened out. Figure 2 As can be seen in the flame grayscale image, in addition to the inner flame, there are also brighter, irregularly shaped reflective areas on the furnace wall. These reflective areas are impurities produced during the combustion process, adhering to the furnace wall and intermingling with the brightness characteristics of the inner flame. Furthermore, in flame grayscale images, due to the dynamic nature of the flame, the shape and position of the inner flame may vary across multiple frames, and thus its area. However, reflective areas are generally more stable, with minimal changes in area between consecutive frames.

[0040] Therefore, in order to further distinguish the inner flame area from the furnace wall reflective area, the embodiment of the present invention can also combine the area between the pixel point and the surrounding pixels with smaller grayscale differences and the area fluctuation to jointly determine the probability that the pixel point is the inner flame area.

[0041] For example, in an embodiment of the present invention, determining the grayscale peak in the flame grayscale image includes: constructing a grayscale histogram of the flame grayscale image with the grayscale value of the pixel in the flame grayscale image as the horizontal axis and the number of pixels corresponding to the grayscale value as the vertical axis, with the lower left corner of the grayscale histogram being the coordinate origin; traversing from the rightmost side of the grayscale histogram to determine the first local maximum of the number of pixels, and taking the grayscale value corresponding to the local maximum as the grayscale peak in the flame grayscale image; wherein the local maximum is greater than the number of adjacent pixels on the left and right sides in the grayscale histogram.

[0042] For example, when obtaining the area between a pixel point and surrounding pixels with smaller grayscale differences, similar pixels of the pixel point in the flame grayscale image in eight neighborhood directions can be obtained respectively. Similar pixels are boundary points of the similar range of the pixel point in the neighborhood direction.

[0043] For example, in an embodiment of the present invention, obtaining similar pixel points of a pixel point in a flame grayscale image in eight neighborhood directions includes: continuously obtaining adjacent pixel points of the pixel point by diffusing outward in the eight neighborhood directions of the pixel point until the grayscale difference between the adjacent pixel point in that direction and the pixel point is greater than a first threshold, and stopping obtaining, and taking the last adjacent pixel point whose grayscale difference is not greater than the first threshold as the similar pixel point of the pixel point in that direction.

[0044] The first threshold may be set to 30, and the first threshold may be set according to actual needs.

[0045] For example, in an embodiment of the present invention, the similarity range value of the pixel point is obtained based on the Euclidean distance between the pixel point and the corresponding similar pixel points, including: taking the normalized mean of the Euclidean distances between the pixel point and the similar pixel points in the eight neighborhood directions as the similarity range value of the pixel point.

[0046] The normalization method may be the maximum and minimum normalization in the linear normalization method, and may be specifically set according to actual needs.

[0047] It's understood that the similarity range value of a pixel is used to represent the size of the similarity range for that pixel. The Euclidean distance between a pixel and each of the similar pixels in the eight neighborhood directions is the Euclidean distance between the pixel and the boundary points of its similarity range in those neighborhood directions. A smaller Euclidean distance indicates a smaller similarity range for the pixel, and the corresponding similarity range value is also smaller.

[0048] For example, in an embodiment of the present invention, the probability that a pixel in each flame grayscale image is in the inner flame region is calculated based on the grayscale change and similarity range between the pixel in each flame grayscale image and the two frames before and after it. For details, see the following relationship:

[0049] ;

[0050] For the The probability that the i-th pixel in the flame grayscale image is the inner flame area, For the Grayscale peak value in the frame flame grayscale image, For the The gray value of the i-th pixel in the frame flame gray image, For the The similarity range value of the i-th pixel in the frame flame grayscale image, For the The variance of the similarity range value of the i-th pixel in the flame grayscale image of the frame and its two previous and next frames, is an exponential function with base e, is the linear normalization function, is the absolute value symbol.

[0051] Among them, The two frames before and after the flame grayscale image are divided into the first Frame flame grayscale image and Frame flame grayscale image.

[0052] In the above formula, Indicates the The relationship between the i-th pixel and the i-th pixel in the flame grayscale image The degree of proximity between the grayscale peaks in the frame flame grayscale image. The smaller the value, the closer the brightness of the pixel is to the maximum brightness, and the higher the degree of conformity to the brightness characteristics of the inner flame area. The degree of proximity is converted into probability through the exponential function. The smaller the difference, the higher the degree of proximity, and the greater the probability of corresponding to the inner flame area.

[0053] Indicates the The area feature of the similar range of the i-th pixel in the frame flame grayscale image. The larger the value, the larger the area of the similar range of the i-th pixel, and the more likely it is the flame inner flame area.

[0054] This value reflects the stability of the similarity range for the i-th pixel. For the inner flame region, the variance is larger due to its dynamic changes, while the variance is smaller for the reflective region. Therefore, the larger the value, the greater the probability that the i-th pixel is in the inner flame region.

[0055] In summary, the closer the grayscale value of a pixel in the flame grayscale image and its two preceding and following images are to their grayscale peaks, the larger the similarity range between the flame grayscale image and its two preceding and following images, and the smaller the fluctuation in that range, the more closely that pixel conforms to the characteristics of the inner flame region, and the greater the probability that it is in the inner flame region. After determining the probability of each pixel in the flame grayscale image belonging to the inner flame region based on the above steps, proceed to the following steps.

[0056] S3: Determine the inner flame region of the flame grayscale image by the probability that the pixel point in the flame grayscale image is the inner flame region, and extract the outer flame region of the flame grayscale image by constructing a gray level co-occurrence matrix to obtain the flame region.

[0057] For example, in an embodiment of the present invention, the inner flame area of the flame grayscale image is determined by the probability that the pixel point in the flame grayscale image is the inner flame area, including: taking the median of the probability that the pixel point in the flame grayscale image is the inner flame area as a comparison value, if the probability that the pixel point is the inner flame area is greater than or equal to the comparison value, then the pixel point is a pixel point in the inner flame area; using a region growing algorithm to process the pixel points in the inner flame area in the flame grayscale image to obtain the inner flame area in the flame grayscale image.

[0058] The step of processing the inner flame region pixels in the flame grayscale image by using a region growing algorithm to obtain the inner flame region in the flame grayscale image can be obtained by existing technologies and will not be described in detail in the embodiment of the present invention.

[0059] It can be understood that by comparing the probabilities of pixels belonging to the inner flame region based on the above steps, the inner flame region pixels can be accurately obtained. By processing the inner flame region pixels in the flame grayscale image using the region growing algorithm, the inner flame region composed of the inner flame region pixels can be obtained. Then, the following steps are performed to extract the outer flame region of the flame grayscale image.

[0060] It should be noted that, combined with Figure 2 It can be seen that the edge of the inner flame region of the flame is connected to the outer flame region and has a lower brightness. The ash produced by burning wood is gray and also has a lower brightness. The outer flame region is connected to the ash region produced by burning wood, making it difficult to accurately identify the outer flame region's boundary. However, the grayscale value texture of the pixels in the ash region is chaotic and uneven, while the grayscale values of the pixels in the outer flame region are more uniform and have no obvious texture. In addition, the adjacent pixels along the gradient change direction of the edge pixels of the inner flame region usually belong to the outer flame region.

[0061] Based on this, the embodiment of the present invention can determine the adjacent pixels that may be the outer flame area based on the edge pixels of the inner flame area, and construct a grayscale co-occurrence matrix based on the adjacent pixels to extract the homogeneity features of the image except the inner flame area, thereby accurately identifying the outer flame area with uniform texture.

[0062] For example, in an embodiment of the present invention, the outer flame area of a flame grayscale image is extracted by constructing a grayscale co-occurrence matrix, including: obtaining the grayscale mean of the adjacent pixel points along the gradient change direction of the edge pixel points of the inner flame area in the flame grayscale image as a reference value of the outer flame area; taking the adjacent pixel points whose grayscale values ​​among the remaining pixel points except the inner flame area in the flame grayscale image have a difference with the reference value that is less than a second threshold as analysis pixel points, and constructing a grayscale co-occurrence matrix based on the analysis pixel points to obtain the outer flame area of the flame grayscale image.

[0063] The second threshold value may be set to 20. The second threshold value may be set according to actual needs, and the embodiment of the present invention does not impose any excessive restrictions on this.

[0064] For example, in an embodiment of the present invention, a grayscale co-occurrence matrix is constructed based on the analysis pixels to obtain the outer flame area of the flame grayscale image, including: setting distance and direction parameters to construct the grayscale co-occurrence matrix of each analysis pixel to obtain the homogeneity value of each analysis pixel; recording the analysis pixel points whose homogeneity values are greater than a third threshold as high homogeneity pixels; using a region growing algorithm to process the high homogeneity pixels in the flame grayscale image to obtain the outer flame area in the flame grayscale image.

[0065] The distance can be set to 1 pixel; the direction parameter can be set to four directions: 0 degrees, 45 degrees, 90 degrees, and 135 degrees; the distance and direction parameters can be set according to actual needs.

[0066] For example, when setting the third threshold, the homogeneity values of the analysis pixels may be sorted, and the top 75% of the analysis pixels may be recorded as high homogeneity pixels. The third threshold may be specifically set according to actual needs.

[0067] The step of using a region growing algorithm to process highly homogeneous pixels in the flame grayscale image to obtain the outer flame region in the flame grayscale image can be obtained through existing technologies and will not be described in detail in the embodiment of the present invention.

[0068] For example, after using the region growing algorithm to process the highly homogeneous pixels in the flame grayscale image to obtain the outer flame area in the flame grayscale image, morphological operations can also be used to process the outer flame area. The specific steps can be implemented through existing technologies and will not be described in detail in the embodiments of the present invention.

[0069] After obtaining the inner flame area and outer flame area of each frame of the flame grayscale image based on the above steps, the combustion efficiency of the wood-fired energy-saving stove can be evaluated according to the characteristics of the flame area composed of the inner flame area and the outer flame area.

[0070] S4: The combustion efficiency of the wood-fired energy-saving stove is obtained by the flame area rectangularity of the flame grayscale image and the heat utilization efficiency of the stove body heat map.

[0071] It should be noted that a rectangular flame region in a flame grayscale image indicates uniform combustion and stable heat output. Therefore, the closer the flame region's shape is to a rectangle, the more stable the combustion and the higher the combustion efficiency. The difference between a wood-fired energy-saving stove's actual and theoretical heat output can directly reflect its heat utilization efficiency. Higher heat utilization efficiency corresponds to higher combustion efficiency.

[0072] Based on this, the embodiment of the present invention can accurately obtain the combustion efficiency of the wood-fired energy-saving stove based on the flame area rectangularity of the flame grayscale image and the heat utilization efficiency of the stove body heat map.

[0073] For example, in an embodiment of the present invention, before obtaining the heat utilization efficiency of the furnace body heat map, the method further includes: obtaining the furnace body heat map during the combustion process of the wood-fired energy-saving stove at each collection moment, and the flame grayscale image at each collection moment corresponds to the furnace body heat map one by one.

[0074] The furnace body heat map may be captured by an infrared thermal imager. The acquisition frequency and cycle of the furnace body heat map may be specifically referred to the above-mentioned method for acquiring the flame grayscale image, which will not be described in detail in the embodiment of the present invention.

[0075] For example, in an embodiment of the present invention, a method for obtaining the rectangularity of the flame area of a flame grayscale image and the heat utilization efficiency of a furnace body heat map includes: obtaining a flame image set in which the flame grayscale image is located; taking the ratio of the number of pixel points in the flame area to the area of the minimum circumscribed rectangle of the flame area as the rectangularity of the flame area; integrating the calorific values of all furnace body heat maps corresponding to each frame of the flame grayscale image in the flame image set to obtain the heat output value of the wood-fired energy-saving stove; obtaining the theoretical calorific value of the wood-fired energy-saving stove based on the mass of the combustion material used in the process of acquiring the flame image set; and obtaining the heat utilization efficiency of the wood-fired energy-saving stove based on the difference between the heat output value of the wood-fired energy-saving stove and the theoretical calorific value.

[0076] The number of pixels in the flame area is the area of the flame area.

[0077] Specifically, when obtaining the heat utilization efficiency of the wood-fired energy-saving stove based on the difference between the heat output value and the theoretical calorific value of the wood-fired energy-saving stove, the ratio of the absolute value of the difference between the heat output value and the theoretical calorific value to the theoretical calorific value can be used as the heat utilization efficiency of the wood-fired energy-saving stove.

[0078] Among them, when obtaining the theoretical calorific value of the wood-fired energy-saving stove based on the mass of the combustion material used in the flame image set acquisition process, the calorific value of the combustion material can be obtained, and the product of the calorific value of the combustion material and the total mass of the combustion can be used as the theoretical calorific value. The specific steps can be obtained through existing technology, and the embodiments of the present invention will not be repeated here.

[0079] It is understandable that the higher the heat utilization efficiency of the wood-fired energy-saving stove, the more complete the combustion of the wood-fired energy-saving stove, the less heat loss caused, and the corresponding higher the combustion efficiency.

[0080] For example, in the embodiment of the present invention, the combustion efficiency of the wood-fired energy-saving stove is calculated, and the specific relationship can be referred to as follows:

[0081] ;

[0082] To improve the combustion efficiency of wood-fired energy-saving stoves, For the The rectangularity of the flame area of the frame flame grayscale image, is the number of flame grayscale image frames in the flame image set, To improve the heat utilization efficiency of wood-fired energy-saving stoves, is a linear normalization function.

[0083] In the above formula, The closer the rectangularity of the flame area of the flame grayscale image is to 1, the The closer the flame area of the frame flame grayscale image is to a rectangle, the more stable the combustion is. Indicates flame image concentration The mean rectangularity of the flame area in the frame flame grayscale image is used to characterize the overall stability of the flame during the combustion process. The stability value is quantified between 0 and 1 by using the maximum and minimum normalization in the linear normalization function. The larger the value, the more stable the shape change state of the flame area, the more stable the combustion state, and the corresponding higher the combustion efficiency.

[0084] In summary, the higher the heat utilization efficiency of the wood-fired energy-saving stove and the more stable the combustion state, the higher its combustion efficiency.

[0085] It can be seen that in the embodiment of the present invention, when obtaining the combustion efficiency of the wood-fired energy-saving stove, the grayscale peak value in the flame grayscale image can be determined; similar pixel points of the pixel point in the flame grayscale image in the eight neighborhood directions are obtained, and the similarity range value of the pixel point is obtained according to the Euclidean distance between the pixel point and the corresponding similar pixel points; the first The probability that the i-th pixel in the flame grayscale image is the inner flame area :

[0086] ;

[0087] For the Grayscale peak value in the frame flame grayscale image, 、 Respectively The grayscale value and similarity range value of the i-th pixel in the frame flame grayscale image, For the The variance of the similarity range value of the i-th pixel in the flame grayscale image of the frame and its two previous and next frames, is an exponential function with base e, is the linear normalization function, is the absolute value symbol; the inner flame area of the flame grayscale image is determined by the probability that the pixel point in the flame grayscale image is the inner flame area, and the outer flame area of the flame grayscale image is extracted by constructing a gray level co-occurrence matrix to obtain the flame area; the combustion efficiency of the wood-fired energy-saving stove is obtained by the rectangularity of the flame area of the flame grayscale image and the heat utilization efficiency of the furnace body heat map, which effectively improves the accuracy of the combustion efficiency of the obtained wood-fired energy-saving stove.

[0088] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing, characterized in that: include: Determine the grayscale peak in the flame grayscale image; Obtain similar pixels of a pixel in the flame grayscale image in the eight neighborhood directions, and obtain the similarity range value of the pixel according to the Euclidean distance between the pixel and the corresponding similar pixels; Calculate the The probability that the i-th pixel in the flame grayscale image is the inner flame area : ； For the Grayscale peak value in the frame flame grayscale image, 、 Respectively The grayscale value and similarity range value of the i-th pixel in the frame flame grayscale image, For the The variance of the similarity range value of the i-th pixel in the flame grayscale image of the frame and its two previous and next frames, is an exponential function with base e, is the linear normalization function, is the absolute value symbol; the inner flame area of the flame grayscale image is determined by the probability that the pixel point in the flame grayscale image is the inner flame area, and the outer flame area of the flame grayscale image is extracted by constructing the gray level co-occurrence matrix to obtain the flame area; The combustion efficiency of the wood-fired energy-saving stove is obtained by the rectangularity of the flame area of the flame grayscale image and the heat utilization efficiency of the stove body heat map, including: obtaining the flame image set in which the flame grayscale image is located; taking the ratio of the number of pixels in the flame area to the area of the minimum circumscribed rectangle of the flame area as the rectangularity of the flame area; integrating the calorific value of the stove body heat map corresponding to each frame of the flame grayscale image in the flame image set to obtain the heat output value of the wood-fired energy-saving stove; obtaining the theoretical calorific value of the wood-fired energy-saving stove based on the mass of the combustion material used in the process of acquiring the flame image set; and obtaining the heat utilization efficiency of the wood-fired energy-saving stove based on the difference between the heat output value of the wood-fired energy-saving stove and the theoretical calorific value. Calculate the combustion efficiency of a wood-fired energy-saving stove: ； To improve the combustion efficiency of wood-fired energy-saving stoves, For the The rectangularity of the flame area of the frame flame grayscale image, is the number of flame grayscale image frames in the flame image set, To improve the heat utilization efficiency of wood-fired energy-saving stoves, is a linear normalization function.

2. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 1 is characterized in that: The step of determining the grayscale peak value in the flame grayscale image also includes: The flame grayscale image and furnace body heat map of the wood-fired energy-saving stove during the combustion process are acquired at each acquisition moment, and a flame image set consisting of multiple frames of continuous flame grayscale images is obtained.

3. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 1 is characterized in that: Determining the grayscale peak value in the flame grayscale image includes: The grayscale value of the pixel in the flame grayscale image is used as the horizontal axis, and the number of pixels corresponding to the grayscale value is used as the vertical axis to construct the grayscale histogram of the flame grayscale image, with the lower left corner of the grayscale histogram as the coordinate origin; starting from the rightmost side of the grayscale histogram, the first local maximum of the number of pixels is traversed to determine the first local maximum of the number of pixels, and the grayscale value corresponding to the local maximum is used as the grayscale peak in the flame grayscale image; wherein, the local maximum in the grayscale histogram is greater than the number of adjacent pixels on the left and right sides.

4. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 1 is characterized in that: The step of obtaining similar pixel points of a pixel point in the flame grayscale image in eight neighborhood directions includes: The adjacent pixel points of the pixel point are continuously acquired by diffusing outward in the eight neighborhood directions of the pixel point until the grayscale difference between the adjacent pixel points in the direction and the pixel point is greater than the first threshold. The acquisition is stopped and the last adjacent pixel point whose grayscale difference is not greater than the first threshold is taken as the similar pixel point of the pixel point in the direction.

5. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 1 is characterized in that: The obtaining of the similarity range value of the pixel point according to the Euclidean distance between the pixel point and the corresponding similar pixel points includes: The normalized mean of the Euclidean distances between a pixel and similar pixels in the eight neighborhood directions is taken as the similarity range value of the pixel.

6. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 1 is characterized in that: The determining of the inner flame area of the flame grayscale image by the probability that a pixel point in the flame grayscale image is an inner flame area includes: The median of the probability that a pixel point in the flame grayscale image is in the inner flame area is used as a comparison value. If the probability that a pixel point is in the inner flame area is greater than or equal to the comparison value, then the pixel point is a pixel point in the inner flame area. The region growing algorithm is used to process the pixels in the inner flame area in the flame grayscale image to obtain the inner flame area in the flame grayscale image.

7. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 1 is characterized in that: The step of extracting the outer flame region of the flame grayscale image by constructing a gray level co-occurrence matrix includes: Obtain the grayscale mean of the adjacent pixels along the gradient change direction of the edge pixels of the inner flame area in the flame grayscale image as the reference value of the outer flame area; The adjacent pixels whose grayscale values except the inner flame area in the flame grayscale image differ from the reference value by less than a second threshold are taken as analysis pixels. The grayscale co-occurrence matrix is constructed based on the analysis pixels to obtain the outer flame area of the flame grayscale image.

8. The method for detecting combustion efficiency of a wood-fired energy-saving stove based on image processing according to claim 7 is characterized in that: The step of constructing a gray level co-occurrence matrix based on the analyzed pixels to obtain the outer flame region of the flame grayscale image includes: The distance and direction parameters are set to construct the grayscale co-occurrence matrix of each analysis pixel point to obtain the homogeneity value of each analysis pixel point; the analysis pixels with homogeneity values greater than the third threshold are recorded as high-homogeneity pixels; the region growing algorithm is used to process the high-homogeneity pixels in the flame grayscale image to obtain the outer flame region in the flame grayscale image.

Images