Method for interpreting cat and dog allergen test strip results based on artificial intelligence visual recognition

Abstract

The present invention relates to the technical field of image processing, and specifically to a method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition, including the steps of obtaining a set of allergy line samples, collecting adsorption lines formed after the reaction between the pet to be tested and the test paper, obtaining a set of aligned adsorption lines after multiple cycles, color mean calculation, selecting the allergy line with the maximum similarity and outputting its corresponding allergen. The present invention significantly reduces single-test errors such as ambient lighting and shooting angle through multiple image alignment and averaging processing; at the same time, the similarity function is combined with weights, suppression values ​​and convex optimization correction factors to make the test results more accurate and reliable. When the similarity exceeds the threshold, the suspected allergen can be quickly locked, and the final verification is carried out in combination with an increased dose exposure test. This method is compatible with a variety of allergen test scenarios, and is easy to expand to cloud-based big data management, which can greatly improve the efficiency and accuracy of cat and dog allergy detection.

Description

Technical Field

[0001] The invention relates to the technical field of image processing, and in particular to a method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition. Background Art

[0002] As people's demand for companionship with pets such as cats and dogs increases, pets play an increasingly important role in family life. However, the number of people with pet allergies and allergic symptoms is increasing. Common allergens for cats and dogs include proteins or other allergenic substances in dander, hair, saliva and excrement. For people with allergies, it may cause allergic reactions in the respiratory tract, skin and immune system.

[0003] In order to quickly detect the sensitivity of humans or pets to cat and dog allergens, some portable test strips have come into being. These test strips usually display the test results through immunochromatography or chemical color reaction. However, in actual applications, this type of detection method still mainly relies on the user's observation and judgment with the naked eye. Once the color difference is not obvious enough or there is interference from ambient light, it will seriously affect the accuracy and repeatability of the results. In addition, pet owners may not have professional interpretation experience and lack unified testing conditions, resulting in subjective factors having a greater impact on the test results.

[0004] Currently, common cat and dog allergen test strips still face the following problems in actual use:

[0005] 1. Human interpretation is highly subjective: The color development of test strips generally relies on naked eye observation, which is greatly affected by ambient light, observation angle and individual experience, and is prone to errors. Especially when the color reaction is close to the threshold, it is often difficult for users to make an accurate judgment, resulting in a highly subjective result.

[0006] 2. High professional requirements for personnel: In some pet hospitals or professional testing institutions, although the staff have certain interpretation experience, they still need to repeatedly compare color cards or control testing conditions (such as time, temperature, and light) to obtain relatively accurate results. For ordinary pet owners, self-interpretation at home is often more prone to errors.

[0007] 3. Unable to achieve big data management and automated analysis: Traditional test strips are mostly one-time and offline. The test results cannot be effectively connected to the cloud database or subsequent diagnosis system, and it is impossible to build a continuously accumulated big data model to track individual differences or conduct population statistical analysis.

[0008] 4. Lack of intelligent identification and quantitative means: Although some testing manufacturers provide simple color comparison references, there is still a lack of real-time and objective analysis of test strip images using artificial intelligence visual recognition technology. Once the test strip has a slight color difference or local uneven color development, it is difficult to make an accurate distinction with the naked eye, and it is impossible to provide sufficient data support for subsequent disease screening or medication guidance.

[0009] Therefore, there is an urgent need for a method for interpreting cat and dog allergen test strip results based on artificial intelligence visual recognition. By automatically preprocessing the captured test strip images, intelligent feature extraction and interpretation algorithm analysis, an objective, rapid and visual judgment of the allergen detection results can be achieved. It can also be further linked to large-scale pet allergy data in the cloud for statistics and comparison, thereby providing pet owners or veterinarians with accurate and intelligent diagnostic references. Summary of the invention

[0010] In view of the above problems, the purpose of the present invention is to propose: a method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition, the method comprising the following steps:

[0011] S1. Take samples from the control line and test line of the cat and dog allergen test strips to obtain an allergy line sample set ,in, The serial number of the allergen tested;

[0012] S2. Test the cat and dog allergen test strips to obtain the adsorption line set ;

[0013] S3. Use image recognition method to test the position of the control line and the test line of the cat and dog allergen test paper in the same area, align their positions, and obtain the adsorption line set after position alignment. ;

[0014] S4, loop step S1 to step S3 Once the position alignment is achieved, the adsorption line set is obtained. ,satisfy: ,in, Indicates the serial number is The first Adsorption line data after secondary image recognition and alignment;

[0015] S5. Yes Post-adsorption line set Calculate the color mean and obtain the mean adsorption line set ;

[0016] S6. Collect the mean adsorption lines Allergy line sample set Perform color comparison to obtain the maximum similarity adsorption line, and output the allergen corresponding to the maximum similarity adsorption line.

[0017] Further, step S1 is specifically as follows: after a single allergen is in contact with the test line and the control line of the test paper for a period of time, an image acquisition device is used to capture the image of the test paper to obtain an allergy line sample. ;

[0018] in, Indicates For each single allergen allergy line sample of the serial number, Indicates the serial number of the allergen, and obtains the allergy line sample set of a single allergen through set operation .

[0019] Furthermore, the allergen is a commercially available standard product or a highly purified product obtained through extraction testing, and the allergy line sample The following conditions are met:

[0020] T1: The allergen appears on the test line of the test paper and combines with the gold-labeled antibody, and the color deepens; at this time, the allergy line sample Includes foreground colors of the area where the allergen appears and other locations on the test strip;

[0021] T2: The allergen does not appear on the test line of the test strip, but appears on the control line, and does not bind to the gold-labeled antibody. Its foreground color is lighter than the colors at other positions, so it should be ignored.

[0022] Further, step S2 is specifically as follows: after the test paper is in contact with the mixed allergens from the cat or dog to be tested, or any uncertain allergen for a period of time, an image acquisition device is used to capture the image of the test paper to obtain the adsorption line , and form a collection of adsorption lines ;

[0023] The above allergens or uncertain allergens come from the genetic information, physiological hormone information, excrement information, saliva information, sweat information, etc. of the cat or dog to be tested, including but not limited to at least one of DNA, RNA, P450 metabolic enzymes, urine, feces, saliva, and sweat.

[0024] Further, step S6 specifically includes the following steps:

[0025] S61. Collect allergy line samples of single allergen after gold labeling The average allergy line collection of the test strip after alignment with the position after multiple uses of the gold-labeled antibody Compare and calculate the similarity for the foreground color at the same position; for each allergy line sample, take the difference of the closest foreground color in the mean allergy line set as the similarity; the similarity is expressed as: , the difference between the foreground color of the allergy line of the single allergen and the foreground color of the mean allergy line is calculated and the interference elimination process is performed.

[0026] Further, S62, introduce similarity function and add weight The weight is determined by the number of times the single allergen has appeared on the test strip after position alignment. The more times it appears, the greater the weight.

[0027] Further, S63, determining the inhibition value according to the ratio of the number of occurrences of this single allergen to other single allergens at the same location ; When the number of occurrences of the single allergen is much smaller than other single allergens at the same position, the inhibition value is calculated and the color shift is suppressed.

[0028] Further, S64, introduces a correction factor , adjust the similarity function and suppression value after introducing weights; the correction factor is controlled by the convex optimization algorithm, so that the output results gradually approach the threshold 1.

[0029] Furthermore, in step S61, the similarity Calculated according to the following formula:

[0030]

[0031] in, is the allergy line sample; z is the distribution position mark of the allergen on the test paper;

[0032] Indicated in the allergy line sample The An allergy line sample area mark, which is used to indicate the color value of the complete area where the allergy is triggered;

[0033] Indicates that in the mean allergy line set, the allergy line sample For the overlapping parts, the specific color value is determined by arithmetic mean;

[0034] Indicates the color difference of the same allergen in different samples.

[0035] Further, in step S62, the weight is calculated by the following formula: And introduce the initial similarity:

[0036]

[0037]

[0038] in, is the secondary similarity, which means the similarity after adding weight;

[0039] represents the weight, which is positively correlated with the secondary similarity, and ;

[0040] Indicates allergy line sample The actual number of tests;

[0041] Indicates that in the mean allergic line set, the position judgment and allergic line sample The number of times the mean line coincides with the position of ;

[0042] When testing a certain allergy line sample When the number of times is greater than the number of times it appears in a mixed test, it satisfies ,and , indicating that the location has a stronger preference for this single allergen.

[0043] Further, step S63, introducing the suppression value Used to compensate for position bias; when the allergy line sample The more the position of is biased towards itself, the lower the probability of other lines appearing. Approaching 1, the suppression value Approaching 0; suppression value Calculate according to the following formula:

[0044]

[0045] in, Indicates the maximum number of tests for all allergy line samples at the same position;

[0046] Indicates that at the same position, The number of tests of the closest other allergy line samples, and this number is less than The number of tests;

[0047] Further, step S64, introducing a correction factor , merge the weighted similarity with the suppression value to satisfy:

[0048]

[0049] Correction Factor Controlled by a convex optimization algorithm, the secondary similarity distribution gradually approaches 1 to allow for slight misjudgments within a certain range.

[0050] Furthermore, when any calculated final similarity reaches above 0.95, the suspected allergen is output.

[0051] Furthermore, the output suspected allergens are sampled and used to conduct an increased dose exposure test on the cat or dog to be tested, and their physiological reactions are observed to verify whether they are the ultimate allergens.

[0052] The beneficial effects of the present invention are:

[0053] 1. Reduce manual interpretation errors: Through artificial intelligence-based visual recognition technology, the color information of the control line and the test line on the cat and dog allergen test strips is automatically collected, aligned, and similarity calculated, which significantly reduces the subjective errors caused by traditional naked eye observation and improves the objectivity and consistency of the test results.

[0054] 2. Enhanced reading accuracy: The present invention accurately aligns the test paper position, and then obtains the mean adsorption line set through multiple cycles of collected adsorption line data, and then comprehensively compares it with the allergic line sample set. It can effectively eliminate the errors caused by factors such as lighting, angle, and exposure differences in a single test, thereby obtaining higher reading accuracy.

[0055] 3. Implement dynamic weighting and suppression mechanism: Set weights for the number of occurrences of each allergen and introduce suppression values, so that when a specific allergen is identified multiple times in multiple tests, its similarity judgment will be more credible; otherwise, the suppression mechanism will reduce the risk of misjudgment. At the same time, combined with the correction factor under the control of the convex optimization algorithm, the result gradually approaches the threshold of 1, ensuring the stability and adaptability of the algorithm under different detection environments.

[0056] In summary, the present invention makes full use of artificial intelligence visual recognition and data fusion technology of multiple cycle detection to make the interpretation of cat and dog allergen test strips more accurate, controllable and traceable, providing an efficient and convenient technical means for pet allergy diagnosis, and greatly improving the accuracy and reliability of detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION

[0058] In order to deepen the understanding of the present invention, the present invention will be further described in detail below in conjunction with examples. The examples are only used to explain the present invention and do not constitute a limitation on the protection scope of the present invention.

[0059] Embodiment 1

[0060] according to Figure 1 As shown, this embodiment provides a method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition, and the method comprises the following steps:

[0061] S1. Take samples from the control line and test line of the cat and dog allergen test strips to obtain an allergy line sample set ,in, The serial number of the allergen tested;

[0062] S2. Test the cat and dog allergen test strips to obtain the adsorption line set ;

[0063] S3. Use image recognition method to test the position of the control line and the test line of the cat and dog allergen test paper in the same area, align their positions, and obtain the adsorption line set after position alignment. ;

[0064] S4, loop step S1 to step S3 Once the position alignment is achieved, the adsorption line set is obtained. ,satisfy: ,in, Indicates the serial number is The first Adsorption line data after secondary image recognition and alignment;

[0065] S5. Yes Post-adsorption line set Calculate the color mean and obtain the mean adsorption line set ;

[0066] S6. Collect the mean adsorption lines Allergy line sample set Perform color comparison to obtain the maximum similarity adsorption line, and output the allergen corresponding to the maximum similarity adsorption line.

[0067] Further, step S1 is specifically as follows: after a single allergen is in contact with the test line and the control line of the test paper for a period of time, an image acquisition device is used to capture the image of the test paper to obtain an allergy line sample. ;

[0068] in, Indicates For each single allergen allergy line sample of the serial number, Indicates the serial number of the allergen, and obtains the allergy line sample set of a single allergen through set operation .

[0069] Furthermore, the allergen is a commercially available standard product or a highly purified product obtained through extraction testing, and the allergy line sample The following conditions are met:

[0070] T1: The allergen appears on the test line of the test paper and combines with the gold-labeled antibody, and the color deepens; at this time, the allergy line sample Includes foreground colors of the area where the allergen appears and other locations on the test strip;

[0071] T2: The allergen does not appear on the test line of the test strip, but appears on the control line, and does not bind to the gold-labeled antibody. Its foreground color is lighter than the colors at other positions, so it should be ignored.

[0072] Further, step S2 is specifically as follows: after the test paper is in contact with the mixed allergens from the cat or dog to be tested, or any uncertain allergen for a period of time, an image acquisition device is used to capture the image of the test paper to obtain the adsorption line , and form a collection of adsorption lines ;

[0073] The above allergens or uncertain allergens come from the genetic information, physiological hormone information, excrement information, saliva information, sweat information, etc. of the cat or dog to be tested, including but not limited to at least one of DNA, RNA, P450 metabolic enzymes, urine, feces, saliva, and sweat.

[0074] Further, step S6 specifically includes the following steps:

[0075] S61. Collect allergy line samples of single allergen after gold labeling The average allergy line collection of the test strip after alignment with the position after multiple uses of the gold-labeled antibody Compare and calculate the similarity for the foreground color at the same position; for each allergy line sample, take the difference of the closest foreground color in the mean allergy line set as the similarity; the similarity is expressed as: , the difference between the foreground color of the allergy line of the single allergen and the foreground color of the mean allergy line is calculated and the interference elimination process is performed.

[0076] Further, S62, introduce similarity function and add weight The weight is determined by the number of times the single allergen has appeared on the test strip after position alignment. The more times it appears, the greater the weight.

[0077] Further, S63, determining the inhibition value according to the ratio of the number of occurrences of this single allergen to other single allergens at the same location ; When the number of occurrences of the single allergen is much smaller than other single allergens at the same position, the inhibition value is calculated and the color shift is suppressed.

[0078] Further, S64, introduces a correction factor , adjust the similarity function and suppression value after introducing weights; the correction factor is controlled by the convex optimization algorithm, so that the output results gradually approach the threshold 1.

[0079] Furthermore, in step S61, the similarity Calculated according to the following formula:

[0080]

[0081] in, is the allergy line sample; z is the distribution position mark of the allergen on the test paper;

[0082] Indicated in the allergy line sample The An allergy line sample area mark, which is used to indicate the color value of the complete area where the allergy is triggered;

[0083] Indicates that in the mean allergy line set, the allergy line sample For the overlapping parts, the specific color value is determined by arithmetic mean;

[0084] Indicates the color difference of the same allergen in different samples.

[0085] Further, in step S62, the weight is calculated by the following formula: And introduce the initial similarity:

[0086]

[0087]

[0088] in, is the secondary similarity, which means the similarity after adding weight;

[0089] represents the weight, which is positively correlated with the secondary similarity, and ;

[0090] Indicates allergy line sample The actual number of tests;

[0091] Indicates that in the mean allergic line set, the position judgment and allergic line sample The number of times the mean line coincides with the position of ;

[0092] When testing a certain allergy line sample When the number of times is greater than the number of times it appears in a mixed test, it satisfies ,and , indicating that the location has a stronger preference for this single allergen.

[0093] Further, step S63, introducing the suppression value Used to compensate for position bias; when the allergy line sample The more the position of is biased towards itself, the lower the probability of other lines appearing. Approaching 1, the suppression value Approaching 0; suppression value Calculate according to the following formula:

[0094]

[0095] in, Indicates the maximum number of tests for all allergy line samples at the same position;

[0096] Indicates that at the same position, The number of tests of the closest other allergy line samples, and this number is less than The number of tests;

[0097] Further, step S64, introducing a correction factor , merge the weighted similarity with the suppression value to satisfy:

[0098]

[0099] Correction Factor Controlled by a convex optimization algorithm, the secondary similarity distribution gradually approaches 1 to allow for slight misjudgments within a certain range.

[0100] Furthermore, when any calculated final similarity reaches above 0.95, the suspected allergen is output.

[0101] Furthermore, the output suspected allergens are sampled and used to conduct an increased dose exposure test on the cat or dog to be tested, and their physiological reactions are observed to verify whether they are the ultimate allergens. Specific embodiment 2

[0103] 1. Allergy line sample collection (S1)

[0104] 1.1 Pretreatment of single allergen

[0105] Prepare a variety of different types of cat and dog allergens (which can be obtained from commercially available standards or extracted and purified by yourself) and label them separately. .

[0106] Each single allergen is brought into contact with the test line (T line) and control line (C line) of the immunochromatographic test paper and allowed to react for a certain period of time under a predetermined environment (eg, room temperature or a specified temperature).

[0107] 1.2 Photographing and forming a collection of allergy line samples

[0108] Use an image acquisition device (such as a high-definition camera or a smart phone) to photograph the developed test paper to obtain image data;

[0109] For the The image corresponding to the allergen is extracted, and the color strip generated on the test paper is recorded as the allergy line sample. When multiple test strips or repeated tests are performed, all allergy line samples formed under different conditions for a single allergen can be combined into a set .

[0110] At the image level, allergy line samples It usually contains one or more darkened test line zones, as well as color information of the background or other locations for subsequent comparison.

[0111] 2. Mixed or unknown allergen detection (S2)

[0112] 2.1 Preparation of samples to be tested

[0113] Obtain mixed allergen samples from the cat or dog to be tested, such as saliva, urine, fecal suspension, dander solution, etc., or tissue fluid of uncertain composition;

[0114] If multiple tests are required for a specific individual, the above-mentioned biological samples can be mixed or tested separately to see whether they will produce a significant color reaction with the gold-labeled antibody on the test paper.

[0115] 2.2 Shoot and obtain adsorption line set

[0116] Place the mixed or unknown allergen in contact with a new blank immunochromatographic test strip (or a test strip that has been tested for other items but the corresponding zone is still usable);

[0117] Under the same or similar environmental parameters, after a period of reaction, images are captured using an image acquisition device.

[0118] After image processing and feature extraction, adsorption lines are formed , which means the color zone that appears on the test paper, and summarizes it as an adsorption line set Here The value of can be regarded as a mapping index for different batches or multiple potential allergens.

[0119] 3. Automatic image alignment (S3)

[0120] 3.1 Position detection and reference line calibration

[0121] For each test strip image captured, first identify the approximate positions of the control line (C line) and the test line (T line), and use image processing algorithms (such as edge detection, morphological operations, etc.) to locate the area;

[0122] Transform and align the images of the same test strip or different test strips taken at different times in the same coordinate system (for example, through key point registration or template matching) to ensure that the same position in different photos corresponds to the same pixel coordinates during subsequent analysis.

[0123] 3.2 Obtaining the aligned adsorption line

[0124] After alignment, the adsorption line data may be slightly deviated from the original position. The uniform transformation is ;

[0125] If multiple batches or multiple rounds of testing are performed (see step S4), ,in Indicates the number of times the test or shooting is repeated.

[0126] 4. Multiple rounds of acquisition and fusion (S4)

[0127] 4.1 Cycle N detection

[0128] For the same or multiple unknown allergens, the test strip detection and image acquisition can be repeated N times to obtain several groups of adsorption line results;

[0129] After each test, the alignment algorithm of S3 was applied to calibrate the obtained adsorption lines to a unified reference coordinate system to avoid excessive individual errors.

[0130] 4.2 Obtaining the aligned adsorption line set

[0131] Finally in In the experiment, for a certain allergen number , the alignment result can be obtained ; After all rounds are combined, it forms , for subsequent averaging and similarity calculation.

[0132] 5. Calculation of mean adsorption line (S5)

[0133] 5.1 Color mean processing

[0134] For all the obtained in step S4 , perform arithmetic or weighted average of the color values ​​at each corresponding position (such as the same pixel grid or the same color display area) to form a mean adsorption line .

[0135] This step is equivalent to using the comprehensive results of multiple tests to eliminate random noise or outliers that may exist in a single test environment, making the adsorption line more representative.

[0136] 5.2 Mean Adsorption Line Set

[0137] After summarizing all the test strips and allergen serial numbers, a new set of adsorption line descriptions can be obtained. When making subsequent comparisons, it can be regarded as the standard adsorption line that is "representative after multiple tests".

[0138] 6. Similarity determination and allergen output (S6)

[0139] 6.1 Comparison of allergy line samples and mean adsorption lines

[0140] The single allergen allergy line sample set obtained in step S1 The mean adsorption line calculated in step S5 Perform one-by-one or batch comparison; select the foreground color or pixel area at the same position, calculate the color difference and obtain the preliminary similarity .

[0141] 6.2 Introducing multiple correction mechanisms

[0142] In order to improve the accuracy of interpretation, this scheme uses weights and inhibition value Parameters such as quantification of the number of allergy lines and other factors; correction factors are also added ,The convex optimization algorithm is used to make the final similarity converge toward the threshold 1, allowing a small error within a certain range.

[0143] 6.3 Final allergen determination

[0144] If for a single allergen serial number After multiple weighting and suppression corrections, if the calculated similarity is higher than 0.95 (or other preset critical value), it can be determined that the unknown allergen is The corresponding allergen is highly consistent; this suspected or confirmed allergen is output as the final allergen, and further increased dose exposure test can be carried out for physiological verification.

[0145] 7. Subsequent verification and extended application

[0146] 7.1 Verification test

[0147] In cases where the judgment result is high or has been output as a "suspected allergen", re-exposure testing can be carried out in a laboratory or clinical environment to observe the intensity of the animal's immune or allergic reaction and verify its accuracy.

[0148] 7.2 Cloud Big Data Management

[0149] The obtained interpretation results and test images are uploaded to the cloud platform for subsequent statistical analysis or cross-hospital sharing, gradually building a larger pet allergen visual recognition database and improving the system's ability to recognize more types of allergens.

[0150] 7.3 Adaptation to other scenarios

[0151] If you need to detect allergens in other species (such as small mammals, birds, etc.), or use a simple test kit for testing at home, you can also reuse the ideas of multiple tests-mean fusion-automatic image alignment in this method. The operation process only needs to make corresponding adjustments to the hardware equipment and threshold settings.

[0152] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for interpreting cat and dog allergen test strip results based on artificial intelligence visual recognition, characterized in that: The method comprises the following steps: S1. Take samples from the control line and test line of the cat and dog allergen test strips to obtain an allergy line sample set ,in, The serial number of the allergen tested; S2. Test the cat and dog allergen test strips to obtain the adsorption line set ; S3. Use image recognition method to test the position of the control line and the test line of the cat and dog allergen test paper in the same area, align their positions, and obtain the adsorption line set after position alignment. ; S4, loop step S1 to step S3 Once the position alignment is achieved, the adsorption line set is obtained. ,satisfy: ,in, Indicates the serial number is The first Adsorption line data after secondary image recognition and alignment; S5. Yes Post-adsorption line set Calculate the color mean and obtain the mean adsorption line set ; S6. Collect the mean adsorption lines Allergy line sample set Perform color comparison to obtain the maximum similarity adsorption line, and output the allergen corresponding to the maximum similarity adsorption line; specifically, the following steps are included: S61. Collect allergy line samples of single allergen after gold labeling The average allergy line collection of the test strip after alignment with the position after multiple uses of the gold-labeled antibody Compare and calculate the similarity for the foreground color at the same position; for each allergy line sample, take the difference of the closest foreground color in the mean allergy line set as the similarity; the similarity is expressed as: , calculate and eliminate interference based on the difference between the foreground color of the allergy line of the single allergen and the foreground color of the mean allergy line; S62. Introduce similarity function and add weight , the weight is determined by the number of times the single allergen has appeared on the test strip after position alignment. The more times it appears, the greater the weight; S63, determine the inhibition value based on the ratio of the number of occurrences of this single allergen to other single allergens in the same location ; When the number of occurrences of the single allergen is much smaller than other single allergens at the same position, the inhibition value is calculated and the color shift is suppressed; S64, introduction of correction factor , adjust the similarity function and suppression value after introducing weights; the correction factor is controlled by the convex optimization algorithm, so that the output results gradually approach the threshold 1.

2. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 1, characterized in that: Step S1 is as follows: after a single allergen is in contact with the test line and the control line of the test paper for a period of time, an image acquisition device is used to capture the image of the test paper to obtain an allergy line sample. ; in, Indicates For each single allergen allergy line sample of the serial number, Indicates the serial number of the allergen, and obtains the allergy line sample set of a single allergen through set operation .

3. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 2, characterized in that: The allergens are commercially available standard products or highly purified products obtained through extraction tests. The following conditions are met: T1: The allergen appears on the test line of the test paper and combines with the gold-labeled antibody, and the color deepens; at this time, the allergy line sample Includes foreground colors of the area where the allergen appears and other locations on the test strip; T2: The allergen does not appear on the test line of the test strip, but appears on the control line, and does not bind to the gold-labeled antibody. Its foreground color is lighter than the colors at other positions, so it should be ignored.

4. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 2, characterized in that: Step S2 is as follows: after the test paper is in contact with the mixed allergens from the cat or dog to be tested, or any uncertain allergen for a period of time, an image acquisition device is used to capture the image of the test paper to obtain the adsorption line , and form a collection of adsorption lines ; The above allergens or uncertain allergens come from the genetic information, physiological hormone information, excrement information, saliva information, sweat information, etc. of the cat or dog to be tested, including but not limited to at least one of DNA, RNA, P450 metabolic enzymes, urine, feces, saliva, and sweat.

5. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 1, characterized in that: In step S61, the similarity Calculated as follows: ; in, is the allergy line sample; z is the distribution position mark of the allergen on the test paper; Indicated in the allergy line sample The An allergy line sample area mark, which is used to indicate the color value of the complete area where the allergy is triggered; Indicates that in the mean allergy line set, the allergy line sample For the overlapping parts, the specific color value is determined by arithmetic mean; Indicates the color difference of the same allergen in different samples.

6. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 5, characterized in that: Step S62: Calculate the weight using the following formula: And introduce the initial similarity: ; ; in, is the secondary similarity, which means the similarity after adding weight; represents the weight, which is positively correlated with the secondary similarity, and ; Indicates allergy line sample The actual number of tests; Indicates that in the mean allergic line set, the position judgment and allergic line sample The number of times the mean line coincides with the position of ; When testing a certain allergy line sample When the number of times is greater than the number of times it appears in a mixed test, it satisfies ,and The larger the value, the stronger the preference of the location for this single allergen.

7. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 6, characterized in that: Step S63: Introducing suppression value Used to compensate for position bias; when the allergy line sample The more the position of is biased towards itself, the lower the probability of other lines appearing. Approaching 1, the suppression value Approaching 0; suppression value Calculate according to the following formula: ; in, Indicates the maximum number of tests for all allergy line samples at the same position; Indicates that at the same position, The number of tests of the closest other allergy line samples, and this number is less than The number of tests.

8. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 7, characterized in that: Step S64: Introducing a correction factor , merge the weighted similarity with the suppression value to satisfy: ; Among them, the correction factor Controlled by a convex optimization algorithm, the secondary similarity distribution gradually approaches 1 to allow for slight misjudgments within a certain range.

9. The method for interpreting cat and dog allergen test paper results based on artificial intelligence visual recognition according to claim 8, characterized in that: When any of the final calculated similarities reaches 0.95 or above, the suspected allergen is output; the output suspected allergen is sampled, and the suspected allergen is used to conduct an increased dose exposure test on the cat or dog to be tested, and its physiological reaction is observed to verify whether it is the final allergen.

Images