Chronic cough syndrome differentiation detection method in Traditional Chinese Medicine based on tongue diagnosis image enhancement

Abstract

The present application relates to the technical field of tongue diagnosis image enhancement processing, and specifically to a method for detecting chronic cough syndrome differentiation in traditional Chinese medicine based on tongue diagnosis image enhancement, comprising: obtaining a patient's tongue diagnosis grayscale image; analyzing the tongue coating roughness reflected in the grayscale co-occurrence matrix of each edge pixel point in each direction, and combining the degree of difference in the tongue coating roughness corresponding to each direction to obtain the significance of multi-directional cracks in the tongue coating of the local area of ​​each edge pixel point; by analyzing the average level and change of the multi-directional crack significance of edge pixels near each edge pixel point, obtaining the fluid loss attention of each pixel point; and using an adaptive contrast enhancement algorithm based on the fluid loss attention to perform contrast enhancement on the tongue diagnosis grayscale image, and using a convolutional neural network in combination with the enhanced tongue diagnosis image to detect chronic cough. The present application can significantly enhance the tongue diagnosis image and improve the contrast enhancement accuracy of the tongue diagnosis image.

Description

Technical Field

[0001] The present application relates to the technical field of tongue diagnosis image enhancement processing, and specifically to a method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement. Background Art

[0002] Currently, tongue diagnosis images captured by medical imaging equipment suffer from poor contrast due to limitations in imaging technology and imaging conditions, which impacts the accuracy of chronic cough syndrome differentiation in Traditional Chinese Medicine (TCM). Therefore, contrast enhancement is needed to improve the visual quality and color of the patient's tongue, thereby more effectively assisting physicians in the diagnosis and treatment of chronic cough.

[0003] Existing technologies often use the Adaptive Contrast Enhancement (ACE) algorithm to enhance the contrast of tongue diagnosis images. This enhances the contrast in tongue diagnosis images, improving the visual quality of tongue texture and color. However, due to the complexity of the tongue coating texture in tongue diagnosis images, existing technologies do not fully consider this complexity when specifically enhancing the contrast in tongue diagnosis images. This results in an inability to effectively enhance the contrast in tongue diagnosis images, affecting the accuracy of Traditional Chinese Medicine (TCM) differentiation of chronic cough syndromes. Summary of the Invention

[0004] In order to solve the above technical problems, the present application provides a traditional Chinese medicine chronic cough syndrome differentiation detection method based on tongue diagnosis image enhancement to solve the existing problems.

[0005] The present invention discloses a method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement, which adopts the following technical solutions:

[0006] One embodiment of the present application provides a method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement, comprising the following steps:

[0007] Obtaining a grayscale image of the patient's tongue diagnosis;

[0008] By analyzing the distribution of tongue coating texture near each edge pixel in the tongue diagnosis grayscale image, the tongue coating roughness reflected in the grayscale co-occurrence matrix of each edge pixel in each direction is obtained. Combined with the degree of difference in the tongue coating roughness corresponding to each direction, the significance of multi-directional cracks in the tongue coating in the local area of ​​each edge pixel is obtained;

[0009] By analyzing the average level and change of the multi-directional crack significance of edge pixels near each edge pixel in the tongue diagnosis grayscale image, the fluid loss attention degree of each pixel is obtained, and an adaptive contrast enhancement algorithm is used according to the fluid loss attention degree to enhance the contrast of the tongue diagnosis grayscale image;

[0010] Detect chronic cough using convolutional neural networks combined with enhanced tongue diagnosis images;

[0011] The calculation method of the body fluid loss attention level of each pixel is:

[0012] ;

[0013] In the formula, is the fluid loss attention of the t-th pixel, is an exponential function with a natural constant as base, is the mean of the multi-directional crack feature sequence of the t-th pixel point, is the mean of the absolute values ​​of all elements in the first-order difference sequence of the multi-directional crack feature sequence of the t-th pixel point, is the t-th pixel, It is the set of all edge pixels in the tongue diagnosis grayscale image.

[0014] Preferably, an edge detection algorithm is used to extract all edge pixel points in the tongue diagnosis grayscale image, and a window of a preset size is used as the local area of ​​each edge pixel point with each edge pixel point as the center.

[0015] Preferably, each edge pixel point and all pixels in its local area are used as inputs of the gray level co-occurrence matrix algorithm to obtain the gray level co-occurrence matrix of each edge pixel point in the four directions of 0°, 45°, 90° and 135° respectively.

[0016] Preferably, the method for calculating the tongue coating roughness reflected in the gray level co-occurrence matrix of each edge pixel point in each direction is:

[0017] ;

[0018] In the formula, is the tongue coating roughness reflected in the gray level co-occurrence matrix of the i-th edge pixel in the j-th direction, is the information entropy of all elements in the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, is the number of rows of the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, and are the means of the sth row and s-1th row in the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, respectively.

[0019] Preferably, the calculation method of the significance of multi-directional cracks of the tongue coating in the local area of ​​each edge pixel point is:

[0020] ;In the formula, is the significance of multi-directional cracks in the local area of ​​the i-th edge pixel point, is the mean value of the tongue coating roughness corresponding to the i-th edge pixel in the four directions, is the number of directions of the gray-level co-occurrence matrix, and are the tongue coating roughness reflected by the gray level co-occurrence matrix of the i-th edge pixel in the j-th and j-1-th directions, To avoid constants with denominators equal to 0.

[0021] Preferably, the multi-directional crack significances corresponding to all edge pixels in the local area of ​​each edge pixel in the tongue diagnosis grayscale image are arranged from small to large to form a multi-directional crack feature sequence for each edge pixel.

[0022] Preferably, the contrast enhancement of the tongue diagnosis grayscale image further includes:

[0023] Extract the low-frequency component image of the tongue diagnosis grayscale image, subtract the tongue diagnosis grayscale image from the low-frequency component image, and obtain the high-frequency component image of the tongue diagnosis grayscale image by combining the subtracted image with the fluid loss attention of each pixel point;

[0024] An adaptive contrast enhancement algorithm is used to enhance the contrast of the high-frequency component image. The contrast-enhanced high-frequency component image is recombined with the low-frequency component image to enhance the contrast of the tongue diagnosis grayscale image and obtain a tongue diagnosis enhanced image.

[0025] Preferably, the method for acquiring the high-frequency component image is further as follows:

[0026] The attention degree of body fluid loss at each pixel point is used as the high-frequency coefficient of the pixel point at each position. The grayscale value of each pixel point in the image obtained after the difference is multiplied by the high-frequency coefficient of the pixel point at the corresponding position to obtain the high-frequency component image of the tongue diagnosis grayscale image.

[0027] Preferably, the method of using a convolutional neural network in combination with the enhanced tongue diagnosis image to detect chronic cough further includes:

[0028] Tongue diagnosis images of chronic cough patients with different symptom types and healthy patients were pre-selected to train the convolutional neural network. The enhanced patient tongue diagnosis images were input into the trained convolutional neural network to obtain the symptom types corresponding to the patient tongue diagnosis images.

[0029] This application has at least the following beneficial effects:

[0030] This application fully considers the complexity of tongue coating texture, measures the fluid loss characteristics in tongue diagnosis grayscale images more accurately, and uses the fluid loss attention as a high-frequency coefficient. This can accurately extract the high-frequency features of fluid loss associated with chronic cough symptoms, which is beneficial for avoiding the subsequent impact on the accuracy of chronic cough syndrome differentiation in Traditional Chinese Medicine.

[0031] Furthermore, by extracting the high-frequency features of fluid loss associated with chronic cough symptoms, the present application enables the adaptive contrast enhancement algorithm to focus on the contrast within the chronic cough symptom area when performing contrast enhancement, thereby achieving a targeted contrast enhancement in tongue diagnosis images, thereby improving the accuracy of subsequent TCM chronic cough syndrome differentiation.

[0032] This application uses tongue diagnosis images after contrast enhancement and color conversion processing to further highlight the shape, color, and moistness of the tongue quality and tongue coating, improve the visual effect of the patient's tongue quality and tongue color, and more effectively assist physicians in the diagnosis and treatment of chronic cough. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0034] Figure 1 This is a flowchart of the steps of the Traditional Chinese Medicine chronic cough syndrome differentiation detection method based on tongue diagnosis image enhancement provided in this application. DETAILED DESCRIPTION

[0035] In order to further illustrate the technical means and effects adopted by this application to achieve the predetermined invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features and effects of the TCM chronic cough syndrome differentiation detection method based on tongue diagnosis image enhancement proposed in this application. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics of one or more embodiments may be combined in any suitable form.

[0036] Unless otherwise defined, terms such as "comprises," "comprising," or any other variants thereof are intended to encompass non-exclusive inclusion, such that a circuit structure, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such article or device. In the absence of further restrictions, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the article or device comprising the element. In addition, the term "and\or" as used herein includes any and all combinations of one or more related listed items. All technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application pertains.

[0037] The specific scheme of the method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement provided by this application is described in detail below with reference to the accompanying drawings.

[0038] One embodiment of the present application provides a method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement. For details, please refer to Figure 1 , including the following steps:

[0039] Step 1: Obtain the patient's tongue diagnosis image.

[0040] The patient's tongue diagnosis image is collected by medical imaging equipment. During the process, the patient is required to sit upright and face the light, then naturally extend the tongue out of the mouth, keep the tongue relaxed, the tongue surface flat and the tongue tip slightly downward, and open the mouth as much as possible to fully expose the tongue; at the same time, the lens of the medical imaging equipment is aimed at the tongue at a 10° depression angle, and flash photography is performed at a close distance of 40 cm to obtain the patient's tongue diagnosis image.

[0041] Furthermore, in this embodiment, the tongue diagnosis image is converted into grayscale by using the OpenCV computer vision library to obtain the patient's tongue diagnosis grayscale image, which is conducive to the subsequent accurate extraction of the high-frequency part of the tongue diagnosis image and improves the visual effect of the patient's tongue quality and tongue color.

[0042] Step 2: By analyzing the distribution of tongue coating texture near each edge pixel point in the tongue diagnosis grayscale image, the tongue coating roughness reflected in the grayscale co-occurrence matrix of each edge pixel point in each direction is obtained. Combined with the degree of difference in the tongue coating roughness corresponding to each direction, the significance of multi-directional cracks in the tongue coating in the local area of ​​each edge pixel point is obtained.

[0043] Generally, patients with chronic cough experience symptoms such as cough, itchy throat, scanty and sticky sputum, expectoration, and dry cough. Clinical trials have shown that cough, itchy throat, and scanty and sticky sputum are more common in patients with chronic cough. These symptoms can lead to excessive evaporation of moisture from the respiratory mucosa, causing loss of body fluids on the tongue coating. Furthermore, chronic cough can lead to insufficient lung yin, which can cause endogenous dryness and heat in patients, further damaging the body fluids on the tongue coating.

[0044] At the same time, the loss of body fluid on the tongue coating will cause the tongue coating to lose its normal moisture, resulting in the tongue diagnosis image showing complex tongue coating texture characteristics such as dry and cracked tongue coating, rough particles, and even cracks. The complex tongue coating texture characteristics often cause the tongue coating color in the tongue diagnosis image to change, thereby causing the tongue diagnosis image to reflect the tongue diagnosis characteristics of chronic cough.

[0045] Therefore, in order to accurately measure the complex texture features of the tongue coating, such as rough particles, dryness and even cracks, and to purposefully enhance the contrast in the tongue diagnosis image, the tongue diagnosis grayscale image is input as the input of the Canny edge detection algorithm to obtain each edge pixel point in the tongue diagnosis grayscale image.

[0046] Edge pixels in a tongue diagnosis grayscale image are generally located in areas with complex tongue coating texture. Measuring the coarseness of the tongue coating can be affected by interfering factors, such as insufficient water intake, excessive fatigue, or exposure to dry environments. However, since chronic cough is a long-term symptom, the coarseness of the tongue coating is often more pronounced in patients. Therefore, the more pronounced the dryness and coarseness of the tongue coating are in the local area of ​​the edge pixel in the tongue diagnosis grayscale image, the more likely the patient is to experience symptoms such as coughing, itchy throat, and less sticky sputum. This indicates that the location of this pixel is more representative of the tongue diagnosis characteristics of chronic cough, and the local area where this pixel is located should be given more attention when performing contrast enhancement.

[0047] Therefore, in order to more accurately measure the rough granular features on the tongue coating, a window of size K×K is set as the local area of ​​each edge pixel point in the tongue diagnosis grayscale image, with each edge pixel point in the center. In this embodiment, the value of K is 9, and the implementer can perform adaptive value selection.

[0048] In order to clearly show the characteristics of dry and cracked tongue coating and rough granules in the local area, each edge pixel and all pixels in its local area are used as the input of the gray-level co-occurrence matrix method, where the step size d is 1, and the gray-level co-occurrence matrices in the four directions of 0°, 45°, 90° and 135° are obtained.

[0049] Furthermore, according to the distribution of the gray level co-occurrence matrix in each direction of the local area of ​​the edge pixel point, the tongue coating roughness reflected in the gray level co-occurrence matrix of the edge pixel point in each direction is calculated:

[0050] ;

[0051] In the formula, is the tongue coating roughness reflected in the gray level co-occurrence matrix of the i-th edge pixel in the j-th direction, is the information entropy of all elements in the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, is the number of rows of the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, and are the means of the sth row and s-1th row in the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, respectively.

[0052] It is understandable that patients with chronic cough will cause fluid loss, which will cause the tongue coating to show more obvious granular roughness characteristics in different directions. This is used to reduce the influence of interference factors when measuring the granular roughness characteristics of the tongue coating. If the grayscale symbiosis pairs in a certain direction of the local area are more complex and the grayscale symbiosis difference is higher, the granular roughness characteristics of the texture change in this direction can be more significantly displayed, that is, the tongue coating roughness characteristics reflected by the grayscale symbiosis matrix in this direction are more significant.

[0053] Generally speaking, the cracked tongue coating caused by chronic cough is caused by overall fluid deficiency, not local damage. Therefore, the cracked tongue coating caused by fluid depletion will manifest in different directions, rather than cracks appearing on the tongue coating surface in a single direction. Therefore, if the tongue coating roughness characteristics reflected by the grayscale co-occurrence matrix in all four directions are high, and at the same time, the tongue coating roughness characteristics reflected in the four directions are relatively small, it can more clearly and intuitively reflect the significant characteristics of cracked tongue coating.

[0054] Therefore, the significance of multi-directional cracks of the tongue coating in the local area of ​​each edge pixel is calculated. In this embodiment, the specific calculation formula is:

[0055] ;

[0056] In the formula, is the significance of multi-directional cracks in the local area of ​​the i-th edge pixel point, is the mean value of the tongue coating roughness corresponding to the i-th edge pixel in the four directions, is the number of directions of the gray-level co-occurrence matrix, and are the tongue coating roughness reflected by the gray level co-occurrence matrix of the i-th edge pixel in the j-th and j-1-th directions, To avoid a constant with a denominator of 0, the value range is 0.1 to 0.5, and the value in this embodiment is 0.2.

[0057] Among them, the significance of multi-directional cracks reflects the multi-directional crack characteristics of the dry and cracked tongue coating reflected in the local area of ​​the edge pixel point. The larger the multi-directional crack characteristics of the dry and cracked tongue coating, the more it can reflect the overall deficiency of body fluid on the tongue coating. That is, the more the patient has manifestations of body fluid loss, the more likely the location of the pixel point is to reflect the symptoms of chronic cough. Therefore, more attention should be paid to the local area where the pixel point is located when performing contrast enhancement.

[0058] Step 3: Analyze the average level and changes of the multi-directional crack significance of edge pixels near each edge pixel in the tongue diagnosis grayscale image to obtain the fluid loss attention of each pixel, and use an adaptive contrast enhancement algorithm based on the fluid loss attention to enhance the contrast of the tongue diagnosis grayscale image.

[0059] Due to limitations in medical imaging equipment and imaging conditions, tongue diagnosis images can exhibit poor contrast in areas with chronic cough symptoms, impacting the accuracy of chronic cough syndrome differentiation in Traditional Chinese Medicine (TCM). Therefore, it is necessary to consider the complexity of the tongue coating texture in areas with chronic cough symptoms and purposefully enhance the contrast in tongue diagnosis images. This allows for a more targeted focus on the contrast within the chronic cough symptom area, thereby improving the accuracy of subsequent TCM chronic cough syndrome differentiation.

[0060] Generally, overall fluid deficiency causes multi-directional cracks at various locations on the tongue coating. Furthermore, due to the global nature of fluid loss, the multi-directional cracks manifested in localized areas of chronic cough are similar. Therefore, the multi-directional crack significance of all edge pixels within the local area of ​​each edge pixel in the tongue diagnosis grayscale image is ranked from smallest to largest, forming the multi-directional crack feature sequence for each edge pixel.

[0061] Furthermore, in this embodiment, the fluid loss attention level of each pixel in the tongue diagnosis grayscale image is calculated:

[0062] ;

[0063] In the formula, is the fluid loss attention of the t-th pixel, is an exponential function with a natural constant as base, is the mean of the multi-directional crack feature sequence of the t-th pixel point, is the mean of the absolute values ​​of all elements in the first-order difference sequence of the multi-directional crack feature sequence of the t-th pixel point, is the t-th pixel, It is the set of all edge pixels in the tongue diagnosis grayscale image.

[0064] It should be noted that when performing contrast enhancement, it is hoped to focus on the contrast in the chronic cough symptom area, and fluid loss is an important manifestation of chronic cough. The higher the average level of the multi-directional crack characteristics of the dry tongue coating, and the smaller the change in the multi-directional crack characteristics of the dry tongue coating, that is, the more similar the multi-directional crack characteristics reflected by fluid loss in the local area, the more prominent the characteristics of fluid loss on the patient's tongue coating.

[0065] Generally speaking, the greater the attention paid to the fluid loss of each pixel in the tongue diagnosis grayscale image, the more it can reflect the high-frequency details of the fluid loss on the patient's tongue coating. When using the adaptive contrast enhancement algorithm for contrast enhancement, the high-frequency detail characteristics of the fluid loss should be enhanced.

[0066] In order to purposefully enhance the contrast in the tongue diagnosis image, the tongue diagnosis grayscale image is used as the input of an adaptive contrast enhancement algorithm (ACE), wherein the tongue diagnosis grayscale image is smoothed to obtain a low-frequency component image of the tongue diagnosis grayscale image; then, the tongue diagnosis grayscale image and the low-frequency component image are subtracted, and the fluid loss attention degree of each pixel point is used as the high-frequency coefficient of the pixel point at each position, and the grayscale value of each pixel point in the image obtained after the subtraction is multiplied by the high-frequency coefficient of the pixel point at the corresponding position to obtain a high-frequency component image of the tongue diagnosis grayscale image; the high-frequency component image is contrast enhanced by using the gain function of the adaptive contrast enhancement algorithm, and the contrast-enhanced high-frequency component image is recombined with the low-frequency component image to obtain a tongue diagnosis enhanced image output by the adaptive contrast enhancement algorithm, wherein the adaptive contrast enhancement algorithm and the specific process of combining the high-frequency image and the low-frequency image are well-known technologies to those skilled in the art and are not redundantly described in this embodiment.

[0067] Step 4: Use convolutional neural network combined with enhanced tongue diagnosis images to detect chronic cough.

[0068] Under normal circumstances, the tongue and tongue coating have vitality and color, and the tongue color is light red, which is what Chinese medicine calls "rongrunhonghuo". However, if the patient is infected with symptoms of chronic cough, he or she is prone to symptoms such as cough, itchy throat, sticky sputum, sensitivity to stimuli, sputum, dry mouth, etc., which causes the patient to have an abnormal condition of excessive loss of body fluid, resulting in changes in the color of the tongue coating and the manifestation of body fluid loss in the tongue diagnosis image.

[0069] In order to accurately perform TCM chronic cough syndrome differentiation detection on tongue diagnosis images, the color space conversion method is used to convert the tongue diagnosis enhanced image into a color image, and the tongue diagnosis image of the patient after color conversion is obtained.

[0070] Furthermore, tongue diagnosis images of different chronic cough patients and tongue diagnosis images of healthy patients are pre-selected to train the convolutional neural network. Preferably, in this embodiment, 1,000 standard tongue diagnosis images of each disease patient are selected and collected, including standard tongue diagnosis images of chronic cough patients, standard tongue diagnosis images of wind-heat cold patients, standard tongue diagnosis images of bronchial asthma patients and standard tongue diagnosis images of healthy patients. The CNN convolutional neural network is trained by using the collected standard tongue diagnosis images of patients with various diseases, wherein the pooling layer of the CNN convolutional neural network adopts maximum pooling and uses the Sigmoid function as the activation function to obtain a trained neural network model. The specific training process of the neural network is a well-known technology for those skilled in the art and will not be repeated in this embodiment.

[0071] The color-converted tongue diagnosis image of the patient is input into the trained neural network model to obtain the symptom type corresponding to the patient's tongue diagnosis image. For example, if the symptom type corresponding to the patient's tongue diagnosis image is chronic cough, then the patient may suffer from chronic cough. The method is used to assist doctors in completing the diagnosis and treatment of chronic cough. The training and classification of the CNN neural network model are well-known technologies and will not be elaborated on in detail.

[0072] It is understood that references to "one embodiment" or "some embodiments" in the present specification mean that one or more embodiments of the present application include a particular feature, structure, or characteristic described in conjunction with that embodiment. Thus, if "in one embodiment," "in some embodiments," "in other embodiments," or "in other embodiments" appear in different places in this specification, they do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and their variations all mean "including but not limited to," unless otherwise specifically emphasized.

[0073] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above description is of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-tasking and parallel processing are also possible or may be advantageous. At the same time, the size of the sequence number of each step in the embodiment does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments in this specification.

[0074] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement, characterized in that: The following steps are involved: Obtaining a grayscale image of the patient's tongue diagnosis; By analyzing the distribution of tongue coating texture near each edge pixel in the tongue diagnosis grayscale image, the tongue coating roughness reflected in the grayscale co-occurrence matrix of each edge pixel in each direction is obtained. Combined with the degree of difference in the tongue coating roughness corresponding to each direction, the significance of multi-directional cracks in the tongue coating in the local area of ​​each edge pixel is obtained; By analyzing the average level and change of the multi-directional crack significance of edge pixels near each edge pixel in the tongue diagnosis grayscale image, the fluid loss attention degree of each pixel is obtained, and an adaptive contrast enhancement algorithm is used according to the fluid loss attention degree to enhance the contrast of the tongue diagnosis grayscale image; Detect chronic cough using convolutional neural networks combined with enhanced tongue diagnosis images; The calculation method of the body fluid loss attention level of each pixel is: ； In the formula, is the fluid loss attention of the t-th pixel, is an exponential function with a natural constant as base, is the mean of the multi-directional crack feature sequence of the t-th pixel point, is the mean of the absolute values ​​of all elements in the first-order difference sequence of the multi-directional crack feature sequence of the t-th pixel point, is the t-th pixel, It is the set of all edge pixels in the tongue diagnosis grayscale image.

2. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, characterized in that: An edge detection algorithm is used to extract all edge pixels in the tongue diagnosis grayscale image, and a window of preset size is used as the local area of ​​each edge pixel with each edge pixel as the center.

3. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, wherein: Each edge pixel and all pixels in its local area are used as the input of the gray level co-occurrence matrix algorithm, and the gray level co-occurrence matrix of each edge pixel in the four directions of 0°, 45°, 90° and 135° is obtained respectively.

4. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, wherein: The calculation method of the tongue coating roughness reflected in the gray level co-occurrence matrix of each edge pixel in each direction is: ； In the formula, is the tongue coating roughness reflected in the gray level co-occurrence matrix of the i-th edge pixel in the j-th direction, is the information entropy of all elements in the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, is the number of rows of the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, and are the means of the sth row and s-1th row in the gray-level co-occurrence matrix of the i-th edge pixel in the j-th direction, respectively.

5. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, wherein: The calculation method of the significance of multi-directional cracks in the local area of ​​each edge pixel point is: ;In the formula, is the significance of multi-directional cracks in the local area of ​​the i-th edge pixel point, is the mean value of the tongue coating roughness corresponding to the i-th edge pixel in the four directions, is the number of directions of the gray-level co-occurrence matrix, and are the tongue coating roughness reflected by the gray level co-occurrence matrix of the i-th edge pixel in the j-th and j-1-th directions, To avoid constants with denominators equal to 0.

6. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, wherein: The multi-directional crack significances of all edge pixels in the local area of ​​each edge pixel in the tongue diagnosis grayscale image are arranged from small to large to form a multi-directional crack feature sequence for each edge pixel.

7. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, wherein: The contrast enhancement of the tongue diagnosis grayscale image further includes: Extract the low-frequency component image of the tongue diagnosis grayscale image, subtract the tongue diagnosis grayscale image from the low-frequency component image, and obtain the high-frequency component image of the tongue diagnosis grayscale image by combining the subtracted image with the fluid loss attention of each pixel point; An adaptive contrast enhancement algorithm is used to enhance the contrast of the high-frequency component image. The contrast-enhanced high-frequency component image is recombined with the low-frequency component image to enhance the contrast of the tongue diagnosis grayscale image and obtain a tongue diagnosis enhanced image.

8. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 7, wherein: The method for obtaining the high-frequency component image is further as follows: The attention degree of body fluid loss at each pixel point is used as the high-frequency coefficient of the pixel point at each position. The grayscale value of each pixel point in the image obtained after the difference is multiplied by the high-frequency coefficient of the pixel point at the corresponding position to obtain the high-frequency component image of the tongue diagnosis grayscale image.

9. The method for detecting chronic cough syndrome differentiation in Traditional Chinese Medicine based on tongue diagnosis image enhancement according to claim 1, wherein: The method of using a convolutional neural network in combination with the enhanced tongue diagnosis image to detect chronic cough further includes: Tongue diagnosis images of chronic cough patients with different symptom types and healthy patients were pre-selected to train the convolutional neural network. The enhanced patient tongue diagnosis images were input into the trained convolutional neural network to obtain the symptom types corresponding to the patient tongue diagnosis images.

Images