Multi-part measurement parameter self-adaptive adjustment method of Bluetooth infrared thermometer

Abstract

The invention relates to the technical field of body temperature measurement, and provides a multi-part measurement parameter adaptive adjustment method of a Bluetooth infrared thermometer. The method comprises the following steps: detecting key points of a human body and generating a measurement path; starting an infrared temperature measurement matrix according to the measurement path to obtain body temperature path data; obtaining measurement environment information and performing self-adaptive compensation to obtain body temperature environment compensation data; constructing a target body temperature fusion model; synchronously compensating the data to the fusion model to obtain a real-time core body temperature; body temperature associated data is obtained and combined with the real-time core body temperature for prediction, and time sequence temperature change information is obtained; storing the core body temperature and the prediction information, and generating a monitoring baseline. The technical problems that the measurement precision of an existing infrared thermometer is greatly influenced by the measurement environment and the individual difference of users, and the authenticity of body temperature measurement is difficult to ensure are solved, and the technical effect that the accuracy of body temperature measurement is improved through cooperative work of multiple Bluetooth infrared thermometers and combination of human body key point detection and self-adaptive compensation is achieved.

Description

Technical Field

[0001] The present application relates to the technical field of body temperature measurement, and in particular to a method for adaptively adjusting multi-site measurement parameters of a Bluetooth infrared thermometer. Background Art

[0002] In recent years, with the growing demand for health monitoring, infrared thermometers have been widely used in medical and family health management due to their non-contact, fast and convenient characteristics. However, most traditional infrared thermometers are based on single-device, single-point measurement, with limited measurement range and lack of intelligent adjustment capabilities, and there are many deficiencies in actual use. In particular, the existing technology cannot achieve collaborative measurement between multiple devices, and it is difficult to fully obtain the temperature data of multiple parts of the user in complex scenarios. With the development of Bluetooth communication technology, new technical support is provided for the collaborative measurement and data interaction of multiple devices of infrared thermometers. Connecting multiple devices through Bluetooth can achieve efficient data transmission and synchronization, overcoming the limitations of single-device measurement. However, relying solely on Bluetooth communication to connect multiple devices is still not enough to solve all problems. Infrared temperature measurement equipment is highly sensitive to environmental factors (such as temperature, humidity, wind speed, etc.), while traditional equipment lacks environmental adaptive compensation capabilities, resulting in poor consistency of measurement results in different scenarios. In addition, the physical characteristics of individual users (such as age, gender, skin characteristics) will significantly affect the distribution of body temperature. Existing thermometers cannot dynamically adjust measurement parameters according to individual characteristics, resulting in low accuracy and personalization of measurement results. Summary of the invention

[0003] The present application provides a method for adaptively adjusting multi-site measurement parameters of a Bluetooth infrared thermometer, aiming to solve the technical problem that the measurement accuracy of existing infrared thermometers is greatly affected by the measurement environment and individual differences of users, making it difficult to ensure the authenticity of temperature measurement.

[0004] In view of the above problems, the present application provides a method for adaptively adjusting multi-site measurement parameters of a Bluetooth infrared thermometer.

[0005] The present application provides a method for adaptively adjusting measurement parameters of multiple parts of a Bluetooth infrared thermometer, the method comprising: performing human body key point detection on a target user, and generating a temperature measurement path based on the detection result; starting an infrared temperature measurement matrix with the temperature measurement path as a constraint to perform temperature measurement of multiple parts of the target user, and obtaining distributed temperature path data, wherein the infrared temperature measurement matrix is ​​composed of M Bluetooth infrared thermometers; interactively obtaining measurement environment information, and adaptively performing environmental compensation for the distributed temperature path data according to the measurement environment information, and obtaining temperature environment compensation data; performing associated data call according to the user number of the target user, and constructing a target temperature fusion model based on the call result; adaptively adjusting the core body temperature by synchronizing the temperature environment compensation data to the target body temperature fusion model, and obtaining real-time core body temperature; interactively obtaining the target user's temperature associated data, and predicting temperature changes according to the temperature associated data and the real-time core body temperature, and obtaining time-series temperature change prediction information; integrating and storing the real-time core body temperature and time-series temperature change prediction information, and generating a long-term and short-term monitoring baseline for the target user.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: The above-mentioned method for adaptively adjusting the measurement parameters of multiple parts of the Bluetooth infrared thermometer is aimed at the target user. First, the key points of the human body are detected to accurately identify the main temperature measurement parts of the user, and the temperature measurement path is generated based on the detection results. This path is used as a measurement constraint to start the infrared temperature measurement matrix composed of M Bluetooth infrared thermometers to measure the body temperature of multiple parts of the user, thereby obtaining distributed body temperature path data covering the whole body. This distributed measurement method effectively improves the comprehensiveness and efficiency of multi-part body temperature collection and avoids the problem of one-sidedness of single-point temperature measurement data. Subsequently, the measurement environment information is obtained through Bluetooth interaction, and the distributed body temperature path data is adaptively compensated for the environment in combination with this information to eliminate the interference of external factors such as ambient temperature and humidity on the temperature measurement results, thereby generating more accurate body temperature environment compensation data. This process ensures the stability and consistency of the measurement data under different environmental conditions. Afterwards, the associated data of the target user is called according to the user number of the target user, and the target body temperature fusion model is constructed based on this. By inputting the compensated body temperature data into the model, the adaptive adjustment of the core body temperature is achieved, thereby obtaining the real-time core body temperature. This dynamic adjustment method can better fit the individual characteristics of users and significantly improve the accuracy of core body temperature measurement. At the same time, the user's body temperature-related data is interactively acquired, and combined with the real-time core body temperature, the temperature change trend is predicted, and the time-series temperature change prediction information is output. This prediction capability can help users understand the temperature change trend in advance. Finally, the real-time core body temperature and time-series temperature change prediction information are integrated and stored to generate the user's long-term and short-term monitoring baseline. This baseline provides a reliable data foundation for the continuous monitoring of the user's health status and provides a scientific basis for the formulation of future health management plans. The entire method has achieved significant improvements in the comprehensiveness of multi-site measurements, the accuracy of data compensation, and the intelligence of core body temperature prediction, providing users with a more scientific and comprehensive health monitoring service.

[0007] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1A schematic diagram of a flow chart of a method for adaptively adjusting measurement parameters of multiple parts of a Bluetooth infrared thermometer in one embodiment; Figure 2 The present invention is a flow chart of generating a temperature measurement path of a method for adaptively adjusting measurement parameters of multiple parts of a Bluetooth infrared thermometer in one embodiment. DETAILED DESCRIPTION

[0010] The embodiment of the present application provides a method for adaptively adjusting multi-site measurement parameters of a Bluetooth infrared thermometer to solve the technical problem that the measurement accuracy of existing infrared thermometers is greatly affected by the measurement environment and individual differences of users, making it difficult to ensure the authenticity of temperature measurement.

[0011] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0012] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0013] Examples, such as Figure 1 As shown, the present application provides a method for adaptively adjusting measurement parameters of multiple parts of a Bluetooth infrared thermometer, the method comprising: The target user's key points on the human body are detected, and a temperature measurement path is generated based on the detection results.

[0014] In an embodiment of the present application, the process of detecting the key points of the human body of the target user is mainly to identify specific key points of the user's body through atlas or sensor data, for example, the commonly used temperature measurement parts such as the forehead, armpit, and neck. The positioning of these key points can accurately reflect the temperature distribution of different parts of the user. Based on these detected key points, a temperature measurement path is analyzed and generated to ensure that each temperature measurement key point can be efficiently covered. The generation of this path is a comprehensive consideration of the human body structure, the measurement range of the device, and the spatial layout of the Bluetooth infrared thermometer. The Bluetooth of the Bluetooth infrared thermometer can synchronize the detection and positioning data between multiple devices, and can also realize the update of the path to ensure the continuity and accuracy of the measurement process. Through this method of human key point detection and path generation, it is possible to provide an efficient and accurate basis for subsequent multi-part temperature measurement, while reducing the possibility of omissions and repeated measurements, and improving the overall quality of temperature data collection.

[0015] Further, if Figure 2 As shown, the present application provides a method for detecting key points of a target user's body and generating a temperature measurement path based on the detection results, including: Access an associated database to collect information according to the user number of the target user, and obtain user vital sign data, wherein the user vital sign data includes static vital sign data and dynamic vital sign data; extract measurement part composition information from a temperature measurement part information library according to the static vital sign data; perform human body key point detection on the target user with the measurement part composition information as a constraint, and determine multiple actual measurement parts; perform measurement trajectory fitting according to the multiple actual measurement parts and the spatial distribution of the infrared temperature measurement matrix, and output the body temperature measurement path.

[0016] Preferably, the user ID of the target user is used to access the associated database and extract the vital sign data related to the user. These vital sign data are divided into static vital sign data and dynamic vital sign data. Static vital sign data include characteristics such as the user's gender, age, height, body shape, etc. that do not change significantly over time. Dynamic vital sign data refers to the user's current activity status, heart rate, etc., which can reflect the dynamic characteristics of the user. Subsequently, according to the static vital sign data of the user, the measurement part composition information is extracted from the temperature measurement part information library. The temperature measurement part information library stores the typical measurement part distribution corresponding to different physical sign types. For example, children are preferentially measured on the forehead and underarms, while adults may prefer the forehead and wrist. By analyzing the static vital sign data of the user, the measurement part combination information suitable for the user can be screened out, providing a basis for subsequent measurement path planning. Afterwards, the target user is subjected to human body key point detection with the measurement part composition information as a constraint. Key point detection collects multi-view images of the user through a camera or other sensor, and uses a measurement identification model to identify and mark the actual measurement part of the target user, such as the specific coordinates of the head, wrist, neck and other areas. Multi-view data can improve the accuracy of key point detection, especially when the user is blocked or has a special posture. Then, according to the actual measurement site detected and the spatial distribution of the infrared temperature measurement matrix, the temperature measurement scheduling simulation of multiple actual measurement sites is performed on multiple candidate call equipment groups, and then the measurement trajectory fitting of multiple actual measurement sites and infrared temperature measurement matrices is completed through multi-level screening to generate an optimal body temperature measurement path, for example, from the left ear canal to the right ear canal to the forehead to the armpit to the mouth. The trajectory fitting takes into account the scanning range of the device, the temperature measurement efficiency, and the key point distribution of the user to ensure that the path coverage is comprehensive and the measurement process is efficient. During the measurement trajectory fitting process, the infrared temperature measurement device shares the measurement range and location information in real time via Bluetooth to ensure that the fitting results cover each measurement site and improve the efficiency of multi-device collaboration. This process realizes the intelligent planning of personalized body temperature measurement paths, improves the comprehensiveness and accuracy of measurements, and improves the measurement efficiency through device collaborative optimization.

[0017] Furthermore, the present application provides a method for extracting measurement site composition information from a temperature measurement site information library according to the static vital sign data, including: Interactively obtain multiple sample measurement distribution coordinates, wherein the multiple sample measurement distribution coordinates have multiple sample vital sign data identifiers; associate and store the multiple sample vital sign data and the multiple sample measurement distribution coordinates to complete the construction of the temperature measurement part information library; traverse and calculate multiple vital sign similarities between the static vital sign data and the multiple sample vital sign data in the temperature measurement part information library; after serializing the multiple vital sign similarities, call the measurement part composition information from the multiple sample measurement distribution coordinates according to the sorting extreme value.

[0018] Optionally, a set of sample measurement distribution coordinates is obtained through a multi-sample data collection method. These coordinates represent the positions of body parts related to temperature measurement in different samples, such as the distribution information of common temperature measurement parts such as ears, foreheads, and armpits. The collection of sample data depends on the collaborative work of Bluetooth infrared thermometers, which realize real-time data transmission through Bluetooth, and synchronize the distribution coordinates of each temperature measurement part and the corresponding physical sign data (such as age, gender, height, and weight) to the historical measurement database. Through Bluetooth communication, an interactive relationship can be quickly established with the historical measurement database to obtain multiple sample measurement distribution coordinates and their corresponding multiple sample physical sign data identifiers. Subsequently, the physical sign data (such as age, gender, height, and weight) of each sample are associated and stored with the corresponding measurement distribution coordinate identifier to form a complete set of sample databases, namely, the temperature measurement part information library. This temperature measurement part information library not only contains the spatial positions of each measurement part of the body, but also contains static information closely related to the individual characteristics of the sample. Afterwards, according to the static physical sign data of the target user (such as age, gender, height, and weight), the sample physical sign data in the temperature measurement part information database are traversed one by one, and the similarity between the user's physical sign and the sample physical sign is calculated, thereby obtaining multiple physical sign similarities. The calculation of similarity can be based on Euclidean distance or other similarity measurement methods. The higher the physical sign similarity, the higher the similarity with the target user in terms of body shape and other aspects. Then, the calculated similarity results are sorted from high to low. By selecting the sample with the highest similarity, the corresponding measurement distribution coordinates are extracted. These coordinates usually represent the temperature measurement part composition information that best matches the characteristics of the target user. For example, if the physical sign data of the target user is close to the characteristics of a group of samples in the sample database (such as height 170cm, age 30 years old, etc.), the temperature measurement part coordinate information of this group of samples will be preferentially selected as the measurement part composition information. The extracted measurement part composition information is used for subsequent measurement path generation and temperature measurement operations, and provides a scientific basis for the fitting of multi-part measurement paths.

[0019] Furthermore, the present application provides a method for detecting key points of the human body of the target user based on the measurement part composition information as a constraint to determine multiple actual measurement parts, including: Perform multi-view image acquisition on the target user to obtain K user images; select the regions of interest of the K user images based on the measurement part composition information as a constraint to obtain K groups of measurement part images; input the K groups of measurement part images into a measurement identification model for automatic identification to obtain K groups of measurement part coordinate intervals; unify the coordinates of the K groups of measurement part coordinate intervals based on the measurement part composition information as a constraint to obtain the multiple actual measurement parts.

[0020] Optionally, multiple image acquisition devices (such as cameras) are used to acquire multi-view images of the target user to generate K user images. These devices communicate via Bluetooth, so that the system terminal can obtain images taken by the user from different angles in real time, ensuring that the key parts of the temperature measurement (such as the forehead, armpits, ears, etc.) are fully captured. The multi-view acquisition method improves the integrity and accuracy of the measurement part identification and provides high-quality input for subsequent processing. Subsequently, with the measurement part composition information as a constraint, the measurement part selection model automatically selects and screens out the region of interest (ROI) related to the temperature measurement in the K user images, and then uses these screened regions of interest as K groups of measurement part images, thereby effectively eliminating the interference of irrelevant areas and ensuring that the measurement target is clear. After that, the K groups of measurement part images are input into the measurement identification model to identify and mark the specific coordinate positions of the measurement parts (such as the center point of the ear or the boundary area of ​​the forehead), and generate K groups of measurement part coordinate intervals. For example, for the ear measurement part, the model can accurately identify the outer boundary of the auricle and the coordinates of its center point, providing high-precision location information. These recognition results are transmitted synchronously to ensure that all identification data can be quickly integrated, laying the foundation for subsequent operations. Then, based on the measurement part composition information, the K groups of measurement part coordinate intervals are unified. In this process, a unified coordinate reference system needs to be selected. The reference system can be defined based on the perspective of the user's front image, or the center point of the space can be selected as the origin. The purpose of this step is to provide a unified spatial reference for all measurement parts and eliminate the deviation caused by different perspectives. For each group of measurement part coordinate intervals, the perspective transformation algorithm is used to convert it into a unified coordinate reference system to obtain multiple actual measurement parts. For example, through the three-dimensional space mapping formula (such as the homogeneous coordinate transformation matrix), the coordinate direction of the side view is rotated and adjusted to align with the coordinate axis of the front view, and then the translation correction is performed according to the distance and position offset between the side view device and the front view device, so that the measurement part coordinates can be accurately represented in the unified coordinate system, ensuring the consistency and accuracy of multi-view data. Through the above process, the actual measurement part of the target user can be accurately identified, and the coordinate deviation caused by different perspectives can be eliminated through multi-view data fusion and coordinate unification. This not only improves the accuracy and robustness of measurement part identification, but also provides reliable data support for subsequent multi-part temperature measurement path planning.

[0021] Furthermore, the present application provides a method of selecting the regions of interest of the K user images based on the measurement part composition information as a constraint to obtain K groups of measurement part images, including: Interactively obtain K sample view image sets, and select the regions of interest of the K sample view image sets according to standard measurement parts to obtain K sample identification image sets; construct K measurement part selection units using the K sample view image sets and the K sample identification image sets; complete the construction of a measurement part selection model by connecting the K measurement part selection units in parallel; automatically select the regions of interest by inputting the K user images into the measurement part selection model to obtain K groups of associated parts; and obtain the K groups of measurement part images by screening the K groups of associated parts using the measurement part composition information as a constraint.

[0022] Optionally, through Bluetooth communication established with multiple image acquisition devices, the system terminal interacts with these image acquisition devices to obtain K sample view image sets. These image sets cover the content of different samples taken from multiple viewpoints, providing a diverse data basis for subsequent analysis of regions of interest. Then, according to the definition of standard measurement parts (such as forehead, ear, armpit, etc.), the K sample view image sets are analyzed, and the regions of interest related to the measurement in each image are framed to generate K sample identification image sets. Subsequently, the K sample view image sets and the corresponding sample identification image sets are divided into K training sets and K validation sets, and the constructed initial measurement part frame selection unit is trained by the training set, and the validation set is used for verification after the training to evaluate the performance of the measurement part frame selection unit. Among them, the measurement part frame selection unit can be constructed based on a convolutional neural network (CNN). Taking CNN as an example, an initial measurement part frame selection unit structure is constructed, including an input layer, a convolution layer, a pooling layer, a fully connected layer, and an output layer. The weight of the box selection unit is assigned using a random initialization method, and a sample view image and an identification image in a training set are input into the box selection unit of the initial measurement part. The training process includes layer-by-layer extraction and analysis of image features. After the input layer receives the sample view image, the edge features, texture information, and geometric shapes of the region of interest are extracted through the convolution layer. The feature dimension is then reduced through the pooling layer to improve the computational efficiency of the box selection unit. In the fully connected layer, the box selection unit conducts a comprehensive analysis of the extracted features and generates the predicted coordinates of the region of interest in the output layer. The difference between the coordinates of the region of interest predicted by the box selection unit and the coordinates of the real region in the sample identification image is calculated by using a loss function (such as L1 or L2 loss), and the gradient of the loss to the weight of each layer is calculated layer by layer through the back propagation algorithm. The weight of the box selection unit is then updated using the Adam optimizer to gradually optimize the parameters of the box selection unit. The above training process is repeated until the maximum number of iterations is reached or the accuracy of the validation set meets expectations. After the training is completed, the box selection unit is tested using the validation set to evaluate the recognition accuracy of the box selection unit for the region of interest. For example, by verifying whether the box selection unit can accurately select the boundaries of target parts such as ears or foreheads. If the model performance meets the requirements, the current box selection unit is output as a measurement part box selection unit. Otherwise, hyperparameters such as learning rate and training batches are adjusted, and retraining is performed to further improve the box selection effect. For other training sets and validation sets, K-1 measurement part box selection units are constructed in the same way. These measurement part box selection units are used to select the regions of interest in images with different perspectives to ensure the recognition efficiency and accuracy of the measurement parts. Afterwards, the K measurement part box selection units are connected in parallel, and the feature extraction capabilities under different perspectives are integrated to complete the construction of the measurement part box selection model. This model has the ability to automatically select regions of interest in multi-perspective images, and can achieve efficient positioning in new input data.Then, K user images are input into the measurement part selection model, and the model automatically selects the region of interest for the measurement part in each image to generate K groups of associated parts. These associated parts include all candidate areas related to the measurement, ensuring comprehensive coverage without missing key parts. Finally, with the measurement part composition information as a constraint, the K groups of associated parts are screened, irrelevant or redundant areas are eliminated, and only the image areas of the parts involved in the measurement part composition information are retained, and finally K groups of measurement part images are generated. For example, if the measurement part composition information only includes the ears and forehead, the image areas corresponding to the ears and forehead will be screened out from the associated parts. Through the above process, the measurement part-related areas can be efficiently extracted from the multi-view sample images, and automatic processing can be achieved by constructing a selection model. This method not only improves the accuracy of measurement part identification, but also significantly reduces the need for manual intervention, providing high-quality data support for subsequent temperature measurement path planning and data fusion.

[0023] For the measurement identification model, the training method is the same as the above-mentioned training measurement part selection unit, the difference is that the training data used are sample measurement part images and sample measurement part coordinates, which will not be repeated here.

[0024] Furthermore, the present application provides a method for fitting a measurement trajectory according to the spatial distribution of the multiple actual measurement locations and the infrared temperature measurement matrix, and outputting the body temperature measurement path, including: Interactively obtain M temperature measurement scanning ranges of the M Bluetooth infrared thermometers in the infrared temperature measurement matrix; arrange and combine the M Bluetooth infrared thermometers to obtain multiple alternative calling device groups, wherein the intersection of the temperature measurement scanning ranges of the multiple alternative calling device groups is an empty set; use the multiple alternative calling device groups to perform temperature measurement scheduling simulation for the multiple actual measurement parts to obtain multiple measurement time consumption information and multiple measurement coverage ratios; extract and obtain multiple alternative calling quantities based on the multiple alternative calling device groups; after performing a primary screening of the multiple alternative calling device groups according to the measurement coverage ratio, perform a secondary screening of the multiple alternative calling device groups according to the weighted calculation results of the multiple alternative calling quantities and the multiple measurement time consumption information to obtain a target calling device group; according to the M temperature measurement scanning ranges, perform measurement trajectory fitting for the multiple actual measurement parts and the target calling device group, and output the temperature measurement path.

[0025] Optionally, the temperature measurement scanning range of each thermometer is obtained by interacting with M Bluetooth infrared thermometers in the infrared temperature measurement matrix through a Bluetooth connection. These ranges are usually described in coordinate form (such as the temperature measurement area in three-dimensional space) for subsequent device combination and measurement path planning. Subsequently, the M Bluetooth infrared thermometers are arranged and combined to generate multiple candidate call device groups. Each device group consists of several thermometers, and the temperature measurement scanning ranges of these devices do not overlap with each other (the intersection is an empty set). This design is to avoid repeated measurement of the same area and improve measurement efficiency. For example, if a device group contains two thermometers, their scanning ranges may cover the user's forehead and ears respectively, and will not act on the same part at the same time. After that, for each generated candidate call device group, a temperature measurement scheduling simulation is performed on multiple actual measurement parts. The simulation process includes verification of the startup sequence, measurement sequence and coverage range of the device group, and outputs the measurement time information and measurement coverage ratio. Among them, the measurement time information records the time required for each device group to complete all measurement tasks. The measurement coverage ratio records the percentage of the device group covering the actual measurement part to evaluate its coverage effect. In addition, multiple candidate call volumes, that is, the number of devices required to be called by each device group, are extracted from multiple candidate call device groups for subsequent screening. After obtaining the measurement time information and measurement coverage ratio, each device group is screened according to its measurement coverage ratio, and the device group with a coverage ratio lower than the preset threshold is eliminated. For example, if a device group only covers 50% of the target part, and the threshold is 75%, the device group will be eliminated. For the device groups that pass the first-level screening, a comprehensive weighted calculation is further performed based on the measurement time information and the candidate call volume, and the weighted results are analyzed. The device group with the highest comprehensive score is selected as the target call device group, thereby completing the second-level screening of the candidate call device group. In this process, weights are assigned according to the proportion of measurement time to standard time and the proportion of call volume to standard call volume. Device groups with shorter measurement time will be assigned higher weights because they have higher overall efficiency. Device groups with fewer call volumes will also be assigned higher weights because they have lower device usage complexity. Then, the target call device group is used to fit the measurement trajectory in combination with the temperature measurement scanning range of each device and the location of the actual measurement part. The fitting process follows the principle of allocating the device closest to a certain part to that part first, so as to make the measurement path as short as possible, and generates an optimal path covering all actual measurement parts to ensure maximum measurement efficiency and coverage. During the allocation process, if the scanning range of a device overlaps with multiple measurement parts, the part closest to its coverage center point is given priority. Finally, according to the actual coverage range of the device, the measurement order is adjusted to reduce the number of switching between devices. After completing the matching of the device and the measurement part, the actual measurement part is used as the key point of the path, and connected point by point according to the optimized measurement order, and finally an optimal body temperature measurement path covering all measurement parts is output.Through the above trajectory fitting process, the generated measurement path can cover the target parts to the maximum extent, while reducing the device movement distance and measurement time, thereby improving the efficiency and accuracy of multi-part body temperature measurement.

[0026] The infrared temperature measurement matrix is ​​started with the temperature measurement path as a constraint to measure the temperature of multiple parts of the target user to obtain distributed temperature path data, wherein the infrared temperature measurement matrix is ​​composed of M Bluetooth infrared thermometers.

[0027] In one embodiment, the generated temperature measurement path is used as a guiding constraint for the measurement process. The path defines multiple measurement sites of the target user and their priority order to ensure the consistency and scientificity of multi-site temperature measurement, for example, from the left ear canal to the right ear canal to the forehead to the armpit to the oral cavity. Subsequently, M Bluetooth infrared thermometers in the infrared temperature measurement matrix are activated through Bluetooth communication. These Bluetooth infrared thermometers will be activated one by one according to the site order of the measurement path to ensure that each Bluetooth infrared thermometer accurately measures the temperature at the measurement site it is responsible for. For example, the Bluetooth infrared thermometer responsible for the left ear canal will be activated first, followed by the right ear canal, forehead, etc., and other Bluetooth infrared thermometers are in standby mode during non-measurement periods to save energy. Afterwards, according to the measurement path, each Bluetooth infrared thermometer is called in turn to measure the temperature of the target site. For example, the first Bluetooth infrared thermometer measures the left ear canal, collects temperature data in real time, and uploads the data to the system terminal via Bluetooth. After completing the left ear canal measurement, the second Bluetooth infrared thermometer is activated to measure the right ear canal. Activate another Bluetooth infrared thermometer to measure the forehead. Activate the relevant Bluetooth infrared thermometers in order according to the path to complete the measurement of the remaining parts. After the temperature measurement is completed at each part, the Bluetooth infrared thermometer will upload the temperature data to the system terminal via Bluetooth in real time. By associating these data with the path information, a distributed temperature path data set is formed. The data set not only records the temperature value of each part, but also includes the timestamp and Bluetooth infrared thermometer identification information to ensure the traceability and integrity of the data, and provide reliable data support for subsequent core temperature analysis.

[0028] The measurement environment information is interactively obtained, and adaptive environment compensation of the distributed body temperature path data is performed according to the measurement environment information to obtain body temperature environment compensation data.

[0029] In one embodiment, the measurement environment information of the target user is obtained in real time through Bluetooth interaction with the environmental sensor device, including wind speed, humidity, etc. These environmental data will be used as the core input of adaptive compensation. Subsequently, these measurement environment information and distributed body temperature path data are input into multiple pre-built body temperature deviation compensation functions, and adaptive environmental compensation is performed on the temperature of each measurement site to solve the problem of body temperature deviation caused by sweat evaporation due to wind speed, humidity, etc. Afterwards, the compensated body temperature data is output in the same format as the distributed body temperature path data to form body temperature environment compensation data. The compensated body temperature data is closer to the user's true core body temperature, providing reliable data support for subsequent health analysis and decision-making.

[0030] Furthermore, the present application provides interactively obtaining measurement environment information, and performing adaptive environment compensation of the distributed body temperature path data according to the measurement environment information to obtain body temperature environment compensation data, including: The measurement site composition information is disassembled to obtain multiple standard measurement sites; network data is called for the first standard measurement site to obtain multiple sample environment information, multiple sample actual measured body temperatures and multiple sample compensated body temperatures; multi-source regression analysis is performed on the multiple sample environment information, multiple sample actual measured body temperatures and multiple sample compensated body temperatures to obtain multiple body temperature deviation compensation functions; the measurement environment information and distributed body temperature path data are input into the multiple body temperature deviation compensation functions to generate the body temperature environment compensation data.

[0031] Optionally, the measurement site composition information is first disassembled into multiple standard measurement sites, such as ears, foreheads, armpits, etc. The definition of each standard measurement site includes its location characteristics and temperature measurement requirements, providing a clear processing target for subsequent data call and analysis. For the first standard measurement site after disassembly, relevant sample data are called from the historical database through networking, including sample environmental information, sample measured body temperature, and sample compensated body temperature. These sample data provide a basis for analyzing body temperature deviation by associating the body temperature values ​​before and after compensation with environmental information. Subsequently, the functional relationship between the body temperature deviation of the site and the environmental variables is constructed using a multi-source regression analysis method in combination with the sample environmental information, the sample measured body temperature, and the sample compensated body temperature as a body temperature deviation compensation function. Specifically, first, the variables in the sample environmental information (wind speed, humidity, etc.) are used as independent variables, and the difference between the sample measured body temperature and the compensated body temperature (body temperature deviation) is used as the dependent variable to construct a multi-source regression model to handle the nonlinear relationship between environmental variables and body temperature deviation. Then input the sample environmental information, sample measured body temperature and sample compensated body temperature into this multi-source regression model, and gradually improve the parameters in the multi-source regression model through the same steps as the aforementioned training measurement part selection unit, through forward propagation, loss calculation, back propagation, parameter optimization and other steps, until the loss value tends to stabilize or reaches the maximum number of iterations. After completing the training of the multi-source regression model, extract the body temperature deviation compensation function corresponding to the part from the multi-source regression model. The compensation function reflects the influence of environmental variables on body temperature deviation. Taking wind speed and humidity as examples, the body temperature deviation compensation function is as follows: ;in, is the body temperature deviation, which indicates the difference between the actual body temperature and the compensated body temperature. x is the wind speed, which indicates the air flow speed in the measurement environment. y is the humidity, which indicates the relative humidity level in the measurement environment. a is the quadratic coefficient of wind speed, which reflects the degree of influence of the square of wind speed on the body temperature deviation. b is the logarithmic relationship coefficient of humidity on the body temperature deviation, which reflects the nonlinear influence of humidity change on the body temperature deviation. c is a constant term, which indicates the fixed part of the basal body temperature deviation, which is usually related to the specific measurement equipment or scene. After that, the same operation is performed on other standard measurement parts to construct multiple body temperature deviation compensation functions. After obtaining multiple body temperature deviation compensation functions, the actual measurement environment information of the target user and the distributed body temperature path data are input into the corresponding body temperature deviation compensation function, and the body temperature compensation value of each measurement part is calculated one by one, and the distributed body temperature path data is automatically corrected according to this body temperature compensation value. The compensated body temperature data is integrated into the body temperature environment compensation data for subsequent core body temperature calculation and health analysis. The compensation data not only solves the interference problem of the measurement environment, but also significantly improves the credibility and accuracy of the body temperature data.

[0032] The associated data is called according to the user number of the target user, and a target body temperature fusion model is constructed based on the call result; the core body temperature is adaptively adjusted by synchronizing the body temperature environment compensation data to the target body temperature fusion model to obtain the real-time core body temperature.

[0033] In one embodiment, historical body temperature data and core body temperature data similar to the target user are called from the associated database according to the user number of the target user. Based on these call results, a personalized body temperature fusion model is constructed, which can provide targeted body temperature analysis capabilities by integrating body temperature data characteristics similar to those of the target user. Subsequently, the body temperature environment compensation data obtained by real-time measurement is synchronized to the constructed body temperature fusion model. The body temperature fusion model analyzes and fuses the measurement results of multiple parts, and finally outputs the real-time core body temperature of the target user. This core body temperature is closer to the real physiological temperature, can accurately reflect the user's current health status, and provide a scientific basis for future health monitoring and diagnosis.

[0034] Furthermore, the present application provides a method for synchronizing the body temperature environment compensation data to the target body temperature fusion model to perform core body temperature adaptive adjustment to obtain real-time core body temperature, including The sample data is called by taking the dynamic vital sign data as a similarity measure to obtain a plurality of sample initial body temperature data and a plurality of sample core body temperatures; the plurality of sample initial body temperature data are preprocessed according to the measurement site composition information to obtain a plurality of sample distributed body temperatures; the target body temperature fusion model is constructed using the plurality of sample distributed body temperatures and the plurality of sample core body temperatures; the body temperature environment compensation data is synchronized to the target body temperature fusion model to perform core body temperature adaptive adjustment to obtain the real-time core body temperature.

[0035] Optionally, based on the dynamic vital signs data of the target user (such as heart rate, blood oxygen level, current activity status, etc.), similar historical sample data is called from the associated database. The calling process is to calculate the similarity between the dynamic vital signs data of the target user and the dynamic vital signs data of the samples in the associated database by means of Euclidean distance, cosine similarity, etc., and screen out samples that meet the similarity threshold as the initial body temperature data of multiple samples and the core body temperature of multiple samples. Subsequently, the initial body temperature data of the samples is preprocessed using the measurement site composition information, that is, the obvious outliers caused by environmental interference in the sample data are first removed to ensure data quality. After the removal, the data is classified and aligned according to the measurement site, for example, the temperature measurement range of the ear is unified. The classified and aligned initial body temperature data is then distributed by site to form sample distributed body temperature features for model construction. After that, the distributed body temperature and core body temperature data of multiple samples are input into the target body temperature fusion model for training. This model can be constructed based on neural network regression, random forest regression, etc. The specific construction process is similar to the above, and is still carried out through steps such as forward propagation, loss calculation, back propagation, and parameter optimization. After the training is completed, the target user's real-time measured temperature environment compensation data is synchronized to the target temperature fusion model. The model fuses the temperature environment compensation data according to the learned mapping relationship to generate real-time core body temperature. This process effectively improves the scientificity and personalization of temperature measurement and provides users with more reliable health data support.

[0036] Interactively obtain the target user's body temperature-related data, and predict the body temperature change based on the body temperature-related data and the real-time core body temperature to obtain time-series temperature change prediction information; integrate and store the real-time core body temperature and time-series temperature change prediction information to generate the target user's long-term and short-term monitoring baselines.

[0037] In one embodiment, the target user's body temperature-related data is obtained by interacting with the user device or health data management platform through a Bluetooth connection, including dynamic physiological indicators (such as heart rate, blood pressure and other physiological parameters related to body temperature changes) and environmental variables (such as room temperature and humidity). These body temperature-related data provide key inputs for the prediction of body temperature changes. Subsequently, these data are input together with the real-time core body temperature into a body temperature change prediction model constructed based on LSTM (long short-term memory network), ARIMA (autoregressive integrated moving average model), etc. The construction method of this body temperature change prediction model is the same as the above. The body temperature change prediction model analyzes the relationship between the user's real-time core body temperature and the body temperature-related data, combines the mapping relationship learned from the sample data to generate a short-term prediction (such as the next 1 hour) and a long-term prediction (such as the next 1 day) of the body temperature change, and outputs the short-term prediction and the long-term prediction together to form a time-series temperature change prediction information. After that, the time-series temperature change prediction information is integrated with the real-time core body temperature and stored to generate a long-term and short-term monitoring baseline for the target user. This long-term and short-term monitoring baseline is divided into a short-term baseline and a long-term baseline. The short-term baseline is a collection of real-time core body temperature and short-term predictions for immediate health status assessment. The long-term baseline is a collection of real-time core body temperature and long-term predictions for long-term health trend analysis. These are stored in a structured form, including key information such as timestamp, core body temperature, predicted body temperature, etc., to facilitate subsequent query and analysis.

[0038] In summary, the embodiments of the present application have at least the following technical effects: The embodiment of the present application proposes a method for adaptively adjusting the measurement parameters of multiple parts of a Bluetooth infrared thermometer, covering the entire process from user temperature measurement path generation to core temperature calculation and time series prediction. The method generates a measurement path through human key point detection, uses a temperature measurement matrix composed of M Bluetooth infrared thermometers to complete multi-part measurement, combines the measurement environment information for adaptive compensation, and generates accurate temperature environment compensation data. At the same time, based on the user number and associated data, a target temperature fusion model is constructed, and the real-time core temperature is accurately calculated through dynamic adjustment, and the temperature change trend is further predicted. The method involves accurate identification of measurement parts, path fitting, environmental compensation, temperature fusion model construction, data storage and prediction, and provides users with an efficient solution for long-term and short-term temperature monitoring. These technical effects jointly solve the technical problem that the measurement accuracy of existing infrared thermometers is greatly affected by the measurement environment and individual differences of users, and it is difficult to ensure the authenticity of temperature measurement, and realizes the technical effect of improving the accuracy of temperature measurement through the collaborative work of multiple Bluetooth infrared thermometers, combined with human key point detection and adaptive compensation.

[0039] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0040] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0041] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for adaptively adjusting measurement parameters of multiple parts of a Bluetooth infrared thermometer, characterized in that: The method comprises: Detect key points of the target user's body and generate a temperature measurement path based on the detection results; Using the temperature measurement path as a constraint, the infrared temperature measurement matrix is ​​started to measure the temperature of multiple parts of the target user to obtain distributed temperature path data, wherein the infrared temperature measurement matrix is ​​composed of M Bluetooth infrared thermometers; interactively obtaining measurement environment information, and performing adaptive environment compensation of the distributed body temperature path data according to the measurement environment information to obtain body temperature environment compensation data; Calling associated data according to the user number of the target user, and constructing a target body temperature fusion model based on the calling result; The real-time core body temperature is obtained by synchronizing the body temperature environment compensation data to the target body temperature fusion model to perform core body temperature adaptive adjustment; interactively obtaining the body temperature associated data of the target user, and performing body temperature change prediction based on the body temperature associated data and the real-time core body temperature to obtain time series temperature change prediction information; The real-time core body temperature and time-series temperature change prediction information are integrated and stored to generate long-term and short-term monitoring baselines for the target user.

2. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 1, characterized in that: Performing human body key point detection on a target user and generating a body temperature measurement path based on the detection result, the method comprising: Accessing a related database to collect information according to the user number of the target user to obtain user vital sign data, wherein the user vital sign data includes static vital sign data and dynamic vital sign data; Extracting measurement part composition information from a temperature measurement part information database according to the static vital sign data; Performing body key point detection on the target user based on the measurement part composition information as a constraint to determine a plurality of actual measurement parts; The measurement trajectory is fitted according to the spatial distribution of the multiple actual measurement locations and the infrared temperature measurement matrix, and the body temperature measurement path is output.

3. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 2, characterized in that: The measurement trajectory is fitted according to the spatial distribution of the multiple actual measurement positions and the infrared temperature measurement matrix, and the body temperature measurement path is outputted. The method includes: Interactively obtain M temperature measurement scanning ranges of the M Bluetooth infrared thermometers in the infrared temperature measurement matrix; Arrange and combine the M Bluetooth infrared thermometers to obtain multiple candidate calling device groups, wherein the intersection of the temperature measurement scanning ranges of the multiple candidate calling device groups is an empty set; Using the multiple candidate calling equipment groups to perform temperature measurement scheduling simulation for the multiple actual measurement locations, and obtaining multiple measurement time-consuming information and multiple measurement coverage ratios; Extracting and obtaining multiple candidate call amounts based on the multiple candidate call device groups; After performing primary screening of the plurality of candidate calling device groups according to the measurement coverage ratio, performing secondary screening of the plurality of candidate calling device groups according to weighted calculation results of the plurality of candidate calling amounts and the plurality of measurement time-consuming information to obtain a target calling device group; According to the M temperature measurement scanning ranges, measurement trajectories are fitted for the multiple actual measurement locations and the target calling device group, and the body temperature measurement path is output.

4. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 2, characterized in that: Interactively obtain measurement environment information, and perform adaptive environment compensation of the distributed body temperature path data according to the measurement environment information to obtain body temperature environment compensation data, the method comprising: Disassembling the measurement part composition information to obtain a plurality of standard measurement parts; Performing online data call on the first standard measurement site to obtain multiple sample environmental information, multiple sample measured body temperatures, and multiple sample compensated body temperatures; Performing multi-source regression analysis on the multiple sample environmental information, the multiple sample measured body temperatures and the multiple sample compensated body temperatures to obtain multiple body temperature deviation compensation functions; The measurement environment information and the distributed body temperature path data are input into the multiple body temperature deviation compensation functions to generate the body temperature environment compensation data.

5. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 2, characterized in that: The real-time core body temperature is obtained by synchronizing the body temperature environment compensation data to the target body temperature fusion model to perform core body temperature adaptive adjustment, and the method includes: Using the dynamic vital sign data as a similarity measure to perform sample data calling, and obtain a plurality of sample initial body temperature data and a plurality of sample core body temperatures; Preprocessing the initial body temperature data of the plurality of samples according to the measurement site composition information to obtain the distributed body temperatures of the plurality of samples; Using the multiple sample distributed body temperatures and the multiple sample core body temperatures to construct the target body temperature fusion model; The body temperature environment compensation data is synchronized to the target body temperature fusion model to perform core body temperature adaptive adjustment to obtain the real-time core body temperature.

6. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 2, characterized in that: Using the measurement part composition information as a constraint, human body key point detection is performed on the target user to determine a plurality of actual measurement parts, the method comprising: Capturing multi-view images of the target user to obtain K user images; Using the measurement part composition information as a constraint, selecting regions of interest of the K user images to obtain K groups of measurement part images; Inputting the K groups of measurement site images into a measurement identification model for automatic identification, and obtaining K groups of measurement site coordinate intervals; The coordinates of the K groups of measurement part coordinate intervals are unified with the measurement part composition information as a constraint to obtain the multiple actual measurement parts.

7. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 6, characterized in that: The region of interest of the K user images is selected based on the measurement part composition information as a constraint to obtain K groups of measurement part images. The method includes: Interactively obtain K sample view image sets, and select the region of interest of the K sample view image sets according to the standard measurement position to obtain K sample identification image sets; The K sample view image sets and the K sample identification image sets are used to construct K measurement part selection units; The construction of the measurement part selection model is completed by connecting the K measurement part selection units in parallel; The K user images are input into the measurement part selection model to automatically select the region of interest, thereby obtaining K groups of associated parts; The K groups of measurement site images are obtained by screening the K groups of associated sites using the measurement site composition information as a constraint.

8. The method for adaptively adjusting the multi-position measurement parameters of a Bluetooth infrared thermometer according to claim 2, characterized in that: Extracting measurement part composition information from a temperature measurement part information library according to the static vital sign data, the method comprising: Interactively obtaining a plurality of sample measurement distribution coordinates, wherein the plurality of sample measurement distribution coordinates have a plurality of sample vital sign data identifiers; The plurality of sample vital sign data and the plurality of sample measurement distribution coordinates are stored in association with each other to complete the construction of the temperature measurement part information library; Traversing and calculating multiple physical sign similarities between the static physical sign data and multiple sample physical sign data in the temperature measurement part information library; After serializing the plurality of vital sign similarities, the measurement site composition information is called from the plurality of sample measurement distribution coordinates according to the sorting extreme values.

Images