Peritoneal dialysis fluid detection data intelligent analysis method based on visual recognition

Abstract

The application provides a peritoneal dialysis liquid detection data intelligent analysis method based on visual recognition, which comprises the following steps: acquiring a test paper image to be recognized, and performing test paper region segmentation, color correction and test paper detection region brightness equalization processing on the test paper image to obtain a processed test paper detection region; acquiring a standardized color feature vector of a test paper color developing region from the test paper detection region; mapping and matching the standardized color feature vector with a standard colorimetric database to obtain a colorimetric matching value, performing color abnormality degree recognition according to the colorimetric matching value, obtaining a test paper colorimetric analysis result according to the color abnormality degree and a preset abnormal value, and performing self-adaptive correction on the test paper colorimetric analysis result with an abnormality; and constructing a personalized colorimetric model according to the colorimetric matching value combined with a historical colorimetric matching value to perform health trend analysis and early warning.

Description

Technical Field

[0001] The present invention belongs to the field of visual recognition, and in particular relates to an intelligent analysis method for peritoneal dialysis fluid detection data based on visual recognition. Background Art

[0002] Peritoneal dialysis is a long-term treatment for patients with end-stage renal disease. It relies on changes in the composition of peritoneal dialysis fluid to assess the dialysis effect and patient health status. Peritoneal dialysis-related peritonitis is the most common serious complication of end-stage renal disease. Early diagnosis is crucial to reducing hospitalization rates and reducing the risk of complications. Existing peritoneal dialysis fluid testing methods mainly include laboratory testing and test paper testing. Laboratory testing methods, such as white blood cell counts, Gram staining smears, culture analysis, etc., rely on professional equipment and medical personnel, have a long testing cycle, and are not suitable for home self-testing. In order to improve convenience, test paper-based testing methods have appeared on the market. Patients can use test paper colorimetry to evaluate key indicators in peritoneal dialysis fluid (such as white blood cells, pH, protein, glucose, etc.). However, this method still has significant shortcomings, affecting the accuracy and practicality of the test.

[0003] First, test strip testing relies on colorimetry with the naked eye or comparison with mobile phone photos, which are seriously affected by environmental factors. Colorimetry with the naked eye is easily affected by subjective judgment and has large errors. Although colorimetry with mobile phone photos can reduce subjective errors, the color analysis methods used in existing technologies (such as colorimetric analysis based on RGB color space) are more sensitive to changes in lighting, angle, and equipment hardware, resulting in inconsistent color recognition of the same test strip under different conditions. In addition, the colorimetric standards of test strips of different brands may vary, and existing detection methods usually use fixed colorimetric cards for comparison, which makes it difficult to ensure consistency across brands, thus affecting the comparability of long-term test data.

[0004] Secondly, existing methods have difficulty in accurately identifying slight changes in the color of the test paper. Many early lesions of peritonitis only manifest as mild leukocytosis or abnormal pH values, and existing colorimetric methods are not sensitive enough to these slight changes in color, which can easily lead to misdetection or missed detection of early lesions. In addition, peritoneal dialysis fluid turbidity is an important indicator for judging peritonitis, but its causes are more complex, including infectious and non-infectious factors (such as chyle, bleeding, protein precipitation, etc.). Current color-based detection methods cannot effectively distinguish different types of turbidity, resulting in a high risk of misdiagnosis.

[0005] Furthermore, existing technologies lack the ability to monitor and analyze long-term health trends. Current testing methods rely primarily on single measurements and cannot integrate historical data for dynamic assessment, making it difficult to predict the risk of peritonitis. For long-term dialysis patients, the lack of continuous health data analysis reduces the guiding value of home testing, leading to many potential health risks being overlooked.

[0006] In summary, the existing peritoneal dialysis fluid test strip detection method has obvious shortcomings in environmental adaptability, colorimetric accuracy, early disease identification, cross-brand consistency, and long-term health monitoring. There is an urgent need for a more accurate, stable, and intelligent detection method to improve the reliability of home testing and provide health assessment capabilities with greater clinical value. Summary of the Invention

[0007] The purpose of this invention is to propose an intelligent analysis method for peritoneal dialysis fluid detection data based on visual recognition, which provides a new detection method by improving colorimetric accuracy, reducing environmental interference, enhancing the ability to detect diseases early and supporting long-term health monitoring.

[0008] In order to achieve the above object, the present invention provides a method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition, the method comprising the following steps:

[0009] S1. Obtain a test paper image to be identified, and perform test paper region segmentation and color correction on the test paper image, as well as brightness equalization on the test paper detection region, to obtain a processed test paper detection region;

[0010] S2. Obtaining a standardized color feature vector of the test paper color development area from the test paper detection area;

[0011] S3. Mapping and matching the standardized color feature vector with a standard colorimetric database to obtain a colorimetric matching value, identifying a color abnormality degree based on the colorimetric matching value, obtaining a test paper colorimetric analysis result based on the color abnormality degree and a preset abnormality value, and adaptively correcting the test paper colorimetric analysis result with abnormalities;

[0012] S4. Build a personalized colorimetric model based on the colorimetric matching value and the historical colorimetric matching value to conduct health trend analysis and early warning.

[0013] Furthermore, the test paper area segmentation adaptively determines the threshold value according to the gray value of the image to generate a binary image, and at the same time removes the noise of the binary image through an opening operation, retaining the connected area with the largest area as the test paper detection area;

[0014] The color correction uses the color mean of the test paper background area as a benchmark to calculate the color correction coefficient, performs color correction on the test paper detection area based on the color correction coefficient, and introduces a regularization term to compensate for illumination changes and test paper brand differences; the regularization term is based on the ratio of the color variance to the mean of the test paper detection area;

[0015] The brightness equalization processing of the test paper detection area is achieved by guided filtering, wherein the guided image is the test paper background area, and the adjustment factor is adaptively determined according to the background brightness.

[0016] Furthermore, the S2 specifically includes:

[0017] Convert the color data of the test paper's color display area from RGB color space to HSV color space;

[0018] Normalize the color data of the test strip detection area to compensate for color shift;

[0019] Perform weighted adjustment on the color features to generate a standardized color feature vector.

[0020] Furthermore, mapping and matching the standardized color feature vector with a standard colorimetric database is specifically performed as follows:

[0021] Constructing a standard colorimetric database; the standard colorimetric database includes a plurality of standard colorimetric vectors and their corresponding health status labels;

[0022] The mapping matching is performed by weighted minimization of color difference calculation to establish a mapping relationship between the color feature vector and the standard colorimetric database.

[0023] Furthermore, the abnormality degree is calculated as follows:

[0024]

[0025] Among them, δ c is the color abnormality; Var(C std ) Calculate the color variance of the test paper color area to measure color stability; C b,avg is the mean color vector in the standard colorimetric database B; Std(C b ) is the color standard deviation of the standard colorimetric database B, which represents the color distribution range of the standard colorimetric; λ is the anomaly detection adjustment coefficient, which ensures adaptive adjustment of the color anomaly calculation of different test paper brands; ∈ is a decimal to prevent the zero division problem;

[0026] The final detection result is determined by comparing the color abnormality with the preset abnormal value, and the abnormal situation is adaptively corrected.

[0027] Furthermore, the S4 specifically includes:

[0028] Using historical test data, a personalized colorimetric model is established to calibrate the health significance of individual color changes;

[0029] Use time series modeling methods to predict changing trends in health status and provide risk warnings in the early stages of disease;

[0030] Provides a personalized feedback mechanism to enable patients to record relevant information when abnormal conditions occur, and optimize the colorimetric calibration model in combination with the doctor's diagnosis results.

[0031] Furthermore, the personalized colorimetric model is updated by minimizing the individual colorimetric deviation:

[0032]

[0033] Among them, M p is the output of the personalized colorimetric calibration model, C b is the best matching color vector in the standard colorimetric database B, which is used to provide a colorimetric benchmark under an ideal health state; ω t is the time decay factor, ensuring that the most recent test results contribute more to the model update; N is the total time; the colorimetric results of the past N tests are The corresponding color feature vector is in, Limit the short-term and drastic changes in colorimetric results, ensure the smoothness of colorimetric model updates, and improve the adaptability of the colorimetric model to changes in individual health status.

[0034] Furthermore, the predicted changing trend of the health status is combined with historical health scores, colorimetric results, and personalized colorimetric models to perform trend prediction; if the predicted value is lower than the set threshold for multiple consecutive times, a health warning is triggered, prompting the user that peritonitis or other health abnormalities may exist.

[0035] The beneficial technical effects of the present invention are at least as follows:

[0036] It can adapt to different lighting conditions and shooting environments more stably, reduce the colorimetric error caused by ambient light interference, and improve the stability of color recognition.

[0037] Secondly, the present invention optimizes the test paper colorimetric analysis method, which can improve the system's ability to recognize slight color changes, reduce false detection or missed detection due to colorimetric errors, and enhance the system's adaptability to test papers of different brands, ensuring the comparability and consistency of test results.

[0038] In addition, the present invention enhances the intelligent analysis capability of test data, enabling the system to combine the user's historical test data, build personalized colorimetric calibration standards, and conduct long-term health trend analysis to provide users with disease risk warnings and improve the early detection capability of peritonitis.

[0039] The invention can significantly improve the accuracy and stability of peritoneal dialysis fluid test strip detection, while enhancing the clinical guidance value of home testing, and providing peritoneal dialysis patients with a more convenient and reliable intelligent health management solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative effort.

[0041] Figure 1 This is a flow chart of the intelligent analysis method for peritoneal dialysis fluid detection data based on visual recognition disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following describes embodiments of the present invention in detail. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only to explain the present invention and are not to be construed as limiting the present invention.

[0043] like Figure 1 As shown, an embodiment of the present invention provides a method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition, the method comprising the following steps:

[0044] S1. Obtain a test paper image to be identified, and perform test paper region segmentation and color correction on the test paper image, as well as brightness equalization processing on the test paper detection region, to obtain a processed test paper detection region.

[0045] Specifically, obtain the standardized test paper image I std , perform the following steps:

[0046] Test paper area segmentation: Use edge detection algorithm to obtain the test paper area contour, and remove background interference through morphological operations to extract the test paper detection area R t .

[0047] Assume the gray value of the original image is I g , use the threshold segmentation method to obtain the binary image B t :

[0048]

[0049] Among them, T b The threshold is dynamically calculated and adaptively adjusted according to the background brightness of the test paper to ensure stable segmentation under different lighting conditions.

[0050] The improved morphological opening operation is used to remove small noise points, and combined with the connected region analysis, the test paper area with the largest area is retained as R t , ensuring the integrity of the test strip area.

[0051] Reference area extraction and color benchmark correction: Use the uncolored area on the test paper as a reference to calculate the color offset and ensure color consistency.

[0052] Let R r is the background area of ​​the test paper, and its color mean is (R r ,G r ,B r ) as a color reference.

[0053] Calculate the color correction coefficient α c , and add an innovative regularization term λ to enhance color stability:

[0054]

[0055] Among them, Var(R t ) is the color variance of the test paper’s color display area, which measures the stability of the color distribution. t ) is the color mean of the test paper’s color-rendering area. λ is the adaptive regularization coefficient, which enables color compensation to adapt to different lighting environments while avoiding color distortion caused by overcorrection.

[0056] Based on α c R t Normalize:

[0057]

[0058] Compared with traditional color normalization methods, this normalization method can further eliminate color deviations caused by test paper brand differences and lighting changes, and improve the stability and cross-brand consistency of colorimetric detection.

[0059] Brightness equalization of the test paper detection area: In order to reduce the color anomaly caused by uneven local illumination, a method based on guided filtering is used to equalize the brightness of the test paper detection area R t ' Perform brightness smoothing.

[0060] Assume that the input of the guided filter is R t ′, the guide image is the test paper background area R r , the filter output is R t ",but:

[0061] R t ″=R t ′-γ·(R r -Mean(R r )) (4)

[0062] Among them, γ is the brightness adjustment factor, which can be adaptively adjusted according to the background brightness of the test paper to ensure that the brightness of the color display area of ​​the test paper remains stable under different environments.

[0063] Compared with the traditional histogram equalization method, this method can enhance the contrast while keeping the color information of the test paper unchanged, thereby improving the accuracy of colorimetric analysis.

[0064] Output the test paper detection area R after color correction and brightness equalization t ″, serves as the input for subsequent color feature extraction, ensuring the comparability of color information under different lighting environments and reducing colorimetric deviations caused by differences in test paper brands.

[0065] S2. Obtain a standardized color feature vector of the test paper color development area from the test paper detection area.

[0066] Among them, the key to the colorimetric detection of test paper is to accurately identify the color information of different detection areas on the test paper, so it is necessary to t Extract pixel-level color values ​​and convert them into more stable color feature vectors.

[0067] Specifically, the color data C of the color-developing area of ​​the test paper raw It is composed of (R, G, B), but because the RGB color space is easily affected by lighting, we first convert it to the HSV color space:

[0068] C h =(H,S,V)=f c (R,G,B) (5)

[0069] Among them, f c It is a color conversion function that converts RGB color to HSV because the HSV hue component H is more robust to lighting changes.

[0070] Adaptive color normalization across the test strip detection area: Test strip color is affected by the chemical reaction of the reagents, and its color characteristics vary depending on the test item. Therefore, we need to improve color stability across different test strip brands and detection areas through adaptive adjustment of color distribution.

[0071] Assume that the color distribution mean of the reference area (background area) of the test paper is C r =(H r ,S r ,V r ), used to compare the color of the test paper detection area C h , calculate the normalization coefficient α c :

[0072]

[0073] Among them, λ is a dynamic adjustment factor that enables color normalization to adapt to the color rendering characteristics of different test paper brands. Mean(C h ) Calculate the mean of the overall color of the detection area to prevent color normalization from excessively affecting the colorimetric results.

[0074] Normalized color vector:

[0075] C n =C h α c (7)

[0076] This normalization method can compensate for the color shift caused by differences in test strip brands and improve the cross-brand consistency of colorimetric analysis.

[0077] Color weight adaptive enhancement: The color of the test strip detection area may be affected by factors such as local uneven lighting and uneven chemical reaction of the test strip. Therefore, we weight the color features to highlight key color information.

[0078] Calculate color enhancement weight ω c :

[0079]

[0080] Among them, β is the color enhancement coefficient, which is used to control the adjustment range of color weight. Var(C n ) calculates the variance of the normalized color distribution and identifies inhomogeneities in the color region. This weighting term ensures that highly variable portions of the color region are enhanced, improving the sensitivity of the colorimetric analysis.

[0081] Construct the final normalized color feature vector: Combine the color enhancement weight ω c , get the final test paper color characteristics:

[0082] C std =ω c ·C n (9)

[0083] The color feature vector C std It has a more stable color representation, ensuring the comparability of color information under different test strip brands and environments, and providing input for subsequent intelligent colorimetric analysis.

[0084] Output normalized color feature vector C std , used for subsequent colorimetric analysis, ensuring that the color information of the test paper detection area remains consistent across environments and devices, and improving the accuracy and robustness of colorimetric analysis.

[0085] S3. Map and match the standardized color feature vector with a standard colorimetric database to obtain a colorimetric matching value, identify the color abnormality according to the colorimetric matching value, obtain a test paper colorimetric analysis result according to the color abnormality and a preset abnormality value, and adaptively correct the test paper colorimetric analysis result with abnormalities.

[0086] Among them, the input standardized color feature vector C std .

[0087] Specifically, we construct a color matching mapping for the test strips: the color change of the test strips corresponds to the change of the chemical composition in the peritoneal dialysis fluid, so we need to establish a color feature vector C std The mapping relationship between the colorimetric database B and the standard colorimetric database B is used for accurate colorimetry.

[0088] Assume that the standard colorimetric database B contains multiple standard colorimetric vectors C b and its corresponding health status label L b , then the color matching function M b The weighted minimization of color difference is calculated as follows:

[0089]

[0090] Among them, C b,i Represents the i-th standard color vector in the colorimetric database. i It is a dynamic weighting factor that is adaptively adjusted by factors such as test paper brand, chemical reaction time, and lighting environment to optimize colorimetric matching under different test paper conditions.

[0091] This optimization method ensures that the colorimetric calculation results are stable and are not affected by ambient light and test paper brand.

[0092] Calculate the color abnormality: Peritoneal dialysis test strips may produce abnormal color due to factors such as test strip quality, user operation, and ambient lighting. Therefore, we need to calculate the abnormality of the test strip color to reduce misjudgment and missed detection. Color abnormality δ c The calculation is as follows:

[0093]

[0094] Among them, Var(C std ) Calculate the color variance of the test paper's color-developing area to measure color stability. b,avg is the mean color vector in the standard colorimetric database B. Std(C b ) is the color standard deviation of the standard colorimetric database B, representing the color distribution range of the standard colorimetric database. λ is the anomaly detection adjustment coefficient, ensuring adaptive adjustment of the color anomaly calculation for different test paper brands. ∈ is a decimal to prevent division by zero.

[0095] This color anomaly detection method can eliminate colorimetric errors caused by factors such as uneven chemical reactions of the test paper and abnormal lighting, thereby improving the robustness of the test paper colorimetry.

[0096] Preferably, after colorimetric calculation and color anomaly detection, we need to make a final detection result determination and perform adaptive correction for abnormal situations.

[0097] Assume the final test category of the test paper is L c Determined by the following decision function:

[0098]

[0099] Among them, τ c is the anomaly determination threshold, which is adaptively adjusted by historical data to ensure the sensitivity and stability of anomaly detection. c >τ c , the system enters the adaptive anomaly correction mode and recalculates the color matching function M b , and combines the user's historical colorimetric data to self-update the colorimetric model to optimize future detection accuracy.

[0100] Output test paper colorimetric analysis result L c , including test paper detection category and color abnormality δ c The results will be used for subsequent health trend analysis and personalized colorimetric optimization to ensure the long-term stability and accuracy of the colorimetric analysis results.

[0101] S4. Build a personalized colorimetric model based on the colorimetric matching value and the historical colorimetric matching value to conduct health trend analysis and early warning.

[0102] Specifically, build a personalized colorimetric calibration model:

[0103] Due to individual differences in the composition of peritoneal dialysis fluid among different patients, the standard value of the test strip color needs to be fine-tuned for each individual. We use historical test data to establish a personalized colorimetric model M p , calibrating the health significance of individual color changes, so that colorimetric analysis is not only suitable for health assessment in a statistical sense of the group, but also adapts to individual health characteristics.

[0104] Assume that the colorimetric results of the past N tests are The corresponding color feature vector is The personalized colorimetric calibration model M p Update by minimizing individual colorimetric deviations:

[0105]

[0106] Among them, C b is the best matching color vector in the standard colorimetric database B, which is used to provide a colorimetric benchmark under an ideal health state. t is the time decay factor, ensuring that the most recent detection results contribute more to the model update:

[0107] ω t =e -λ(N-t ) (14)

[0108] λ is the time decay parameter, which controls the influence of historical data on model update. Limit the short-term and drastic changes in colorimetric results, ensure the smoothness of colorimetric model updates, and improve the adaptability of the colorimetric model to changes in individual health status.

[0109] Preferably, since the early symptoms of peritonitis may manifest as slow changes in the colorimetric results of the test strip rather than a single mutation, we use time series modeling methods to predict the changing trend of health status and provide risk warnings in the early stages of the disease.

[0110] Let H(t) represent the health score of the detection data at time t, which is defined as follows:

[0111]

[0112] Where f(·) is the health trend modeling function, which combines historical health scores, the latest colorimetric results, and the personalized calibration model to predict trends. To enhance the stability of trend modeling, a dynamic change constraint is introduced:

[0113]

[0114] Among them, α, β, and γ are model weights that control the dynamic changes of health scores. Limit large fluctuations in health scores within a short period of time and improve the reliability of abnormal change detection.

[0115] If H(t) is lower than the set threshold τ for multiple consecutive times h , then a health warning W is triggered, prompting the user that peritonitis or other health abnormalities may occur:

[0116]

[0117] Among them, τ h Dynamic adjustments are made through individual historical data to ensure the accuracy of early warnings.

[0118] Preferably, since the patient's living habits, diet, medication use and other factors may affect the colorimetric results of peritoneal dialysis fluid, we provide a personalized feedback mechanism so that patients can record relevant information when abnormal conditions occur and optimize the colorimetric calibration model M in combination with the doctor's diagnosis results. p .

[0119] During the next colorimetric analysis, the system will adjust the colorimetric calibration parameters based on the personalized feedback data and optimize the calculation of the health score H(t), so that the model is more in line with the actual changes in the individual's health status.

[0120] Output personalized health trend H(t) and intelligent warning signal W for long-term health monitoring and feedback to personalized colorimetric calibration model M p , forming a closed-loop optimization to improve the accuracy and individual adaptability of colorimetric analysis.

[0121] The foregoing description of specific embodiments of the present disclosure is intended to illustrate a method for performing a multi-tasking process. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0122] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0123] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0124] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the embodiments of this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0125] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0128] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0129] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0130] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0131] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0132] This specification may be described in the general context of computer-executable instructions, such as program modules, executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media, including storage devices.

[0133] The various embodiments in this specification are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the system embodiments are generally similar to the method embodiments, so the description is relatively simple. For relevant parts, refer to the description of the method embodiments.

[0134] Finally, it should be noted that the lithium battery pack chip balancing control platform disclosed in the embodiment of the present invention is only a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, ordinary technicians in this field should understand that it is still possible to modify the technical solutions recorded in the aforementioned embodiments, or to replace some of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An intelligent analysis method for peritoneal dialysis fluid detection data based on visual recognition, characterized in that: The method comprises the following steps: S1. Obtain a test paper image to be identified, and perform test paper region segmentation and color correction on the test paper image, as well as brightness equalization on the test paper detection region, to obtain a processed test paper detection region; S2. Obtaining a standardized color feature vector of the test paper color development area from the test paper detection area; S3. Mapping and matching the standardized color feature vector with a standard colorimetric database to obtain a colorimetric matching value, identifying a color abnormality degree based on the colorimetric matching value, obtaining a test paper colorimetric analysis result based on the color abnormality degree and a preset abnormality value, and adaptively correcting the test paper colorimetric analysis result with abnormalities; S4. Build a personalized colorimetric model based on the colorimetric matching values ​​combined with historical colorimetric matching values ​​to perform health trend analysis and early warning; The color abnormality is calculated as follows: ； in, is the degree of color abnormality; Calculate the color variance of the test paper's color-developing area to measure color stability; Standard colorimetric database The mean color vector in ; Standard colorimetric database The color standard deviation indicates the color distribution range of the standard colorimetric; Adjust the coefficient for anomaly detection to ensure adaptive adjustment of color anomaly calculation for different test strip brands; It is a decimal to prevent division by zero; The final detection result is determined by comparing the color abnormality with the preset abnormal value, and the abnormal situation is adaptively corrected.

2. The method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition according to claim 1, characterized in that: The test paper area segmentation method adaptively determines the threshold value according to the gray value of the image to generate a binary image, and removes the noise of the binary image through an opening operation, retaining the connected area with the largest area as the test paper detection area; The color correction uses the color mean of the test paper background area as a benchmark to calculate the color correction coefficient, performs color correction on the test paper detection area based on the color correction coefficient, and introduces a regularization term to compensate for illumination changes and test paper brand differences; the regularization term is based on the ratio of the color variance to the mean of the test paper detection area; The brightness equalization processing of the test paper detection area is achieved by guided filtering, wherein the guided image is the test paper background area, and the adjustment factor is adaptively determined according to the background brightness.

3. The method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition according to claim 1, characterized in that: Said S2 specifically includes: Convert the color data of the test paper's color display area from RGB color space to HSV color space; Normalize the color data of the test strip detection area to compensate for color shift; Perform weighted adjustment on the color features to generate a standardized color feature vector.

4. The method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition according to claim 1, characterized in that: The mapping and matching of the standardized color feature vector with the standard colorimetric database is specifically as follows: Constructing a standard colorimetric database; the standard colorimetric database includes a plurality of standard colorimetric vectors and their corresponding health status labels; The mapping matching is performed by weighted minimization of color difference calculation to establish a mapping relationship between the color feature vector and the standard colorimetric database.

5. The method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition according to claim 1, characterized in that: Said S4 specifically includes: Using historical test data, a personalized colorimetric model is established to calibrate the health significance of individual color changes; Use time series modeling methods to predict changing trends in health status and provide risk warnings in the early stages of disease; Provides a personalized feedback mechanism to enable patients to record relevant information when abnormal conditions occur, and optimize the colorimetric calibration model in combination with the doctor's diagnosis results.

6. The method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition according to claim 5, characterized in that: The personalized colorimetric model is updated by minimizing the individual colorimetric deviation: ； in, is the output of the personalized colorimetric calibration model, Standard colorimetric database The best matching color vector in is used to provide a colorimetric benchmark under an ideal health state; is the time decay factor, ensuring that the most recent detection results contribute more to the model update; Total duration of time; past The colorimetric results of the second test were , and its corresponding color feature vector is ;in, Limit the short-term and drastic changes in colorimetric results, ensure the smoothness of colorimetric model updates, and improve the adaptability of the colorimetric model to changes in individual health status.

7. The method for intelligent analysis of peritoneal dialysis fluid detection data based on visual recognition according to claim 5, characterized in that: The predicted health status change trend is combined with historical health scores, colorimetric results, and personalized colorimetric models to perform trend prediction; If the predicted value is lower than the set threshold for multiple consecutive times, a health warning will be triggered, prompting the user that peritonitis or other health abnormalities may exist.

Images