Non-contact AI vision real-time multi-dimensional physiological and psychological health testing device and method

Abstract

The present invention discloses a non-contact AI vision real-time multi-dimensional physiological and mental health testing device and method, which relates to the technical field of data processing. The device includes: a non-contact AI vision module for collecting psychological test answers, real-time physiological and motion expression information; a physiological health assessment module for assessing the health status based on standard physiological states; a mental health assessment module for generating a mental health score by inputting psychological test questions and answers into a scoring model; a credibility correction module for correcting the score by combining real-time physiological information; a conflict tracking compensation module for optimizing the assessment result based on motion expression information; and a mechanism adjustment module for dynamically optimizing the test mechanism according to the results of psychological and physiological health assessments, thereby achieving the technical effects of improving the objectivity and comprehensiveness of the test and improving the accuracy of the test.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to a non-contact AI vision real-time multi-dimensional physiological and mental health testing device and method. Background Art

[0002] Traditional methods mainly rely on users' subjective answers to psychological test questions and judge their mental health status by analyzing the answers. They rely too much on users' subjective answers. Users may deliberately conceal or adjust answers due to subjectivity or social desirability bias, resulting in the test results not being able to truly reflect their mental health status, and being unable to comprehensively and objectively evaluate the mental and physical health status of users, and the test adaptability is poor. Summary of the Invention

[0003] The present invention provides a non-contact AI vision real-time multi-dimensional physiological and mental health testing device and method to solve the technical problems of single information dimension and insufficient test objectivity in the prior art, and achieve the technical effects of improving the objectivity and comprehensiveness of the test and improving the test accuracy.

[0004] In a first aspect, the present invention provides a non-contact AI vision real-time multi-dimensional physiological and mental health testing device, wherein the non-contact AI vision real-time multi-dimensional physiological and mental health testing device includes:

[0005] A non-contact AI vision module, configured to test a target user according to a registration sequence of psychological test questions, and collect a psychological test answer sequence, a real-time physiological information sequence, and a real-time action and expression information sequence.

[0006] A physiological health assessment module, configured to perform a health assessment on the real-time physiological information sequence according to the standard physiological state vector of the target user to obtain a physiological health assessment result.

[0007] A mental health assessment module, configured to input the registration sequence of psychological test questions and the psychological test answer sequence into a first mental health scoring model to obtain a first mental health scoring sequence.

[0008] A credibility correction module, configured to correct the credibility of the first mental health scoring sequence according to the real-time physiological information sequence to obtain a second mental health scoring sequence.

[0009] A conflict tracking and compensation module, configured to perform conflict tracking and compensation on the second mental health scoring sequence according to the real-time action and expression information sequence to obtain a mental health assessment result.

[0010] A mechanism adjustment module, configured to dynamically adjust the user test mechanism according to the mental health assessment result and the physiological health assessment result.

[0011] In a feasible implementation manner, the credibility correction module includes:

[0012] A credibility detection unit, configured to perform credibility detection on the psychological test answer sequence according to the real-time physiological information sequence based on an answer credibility threshold, and determine a credible abnormal answer sequence corresponding to an answer credibility abnormal coefficient sequence.

[0013] A mental health score extraction unit, configured to perform data extraction on the first mental health score sequence based on the credible abnormal answer sequence to obtain a credible abnormal mental health score sequence.

[0014] A correction coefficient calculation unit, configured to perform a ratio calculation on the answer credibility abnormal coefficient sequence and the answer credibility threshold to obtain an answer credibility correction coefficient sequence.

[0015] A corrected score generation unit, configured to correct the credible abnormal mental health score sequence based on the answer credibility correction coefficient sequence to obtain a mental health corrected score sequence.

[0016] A score update unit, configured to update the first mental health score sequence according to the mental health corrected score sequence to generate the second mental health score sequence.

[0017] In a feasible implementation manner, the credibility detection unit includes:

[0018] An answer credibility evaluation subunit, configured to perform credibility evaluation on the psychological test answer sequence according to the real-time physiological information sequence to obtain M answer credibility coefficients.

[0019] An abnormal coefficient sequence establishment subunit, configured to determine whether the M answer credibility coefficients are less than the answer credibility threshold, and establish an answer credibility abnormal coefficient sequence less than the answer credibility threshold.

[0020] An abnormal identification subunit, configured to perform credible abnormal answer identification on the psychological test answer sequence according to the answer credibility abnormal coefficient sequence to obtain the credible abnormal answer sequence.

[0021] In a feasible implementation manner, the conflict tracking and compensation module includes:

[0022] A conflict evaluation unit, configured to perform conflict degree evaluation on the psychological test answer sequence according to the real-time action and expression information sequence to obtain an answer conflict evaluation sequence.

[0023] A score compensation and optimization unit, configured to perform compensation and optimization on the second mental health score sequence according to the answer conflict evaluation sequence to obtain a third mental health score sequence.

[0024] The tension evaluation unit is used to evaluate the answering tension of the target user according to the real-time physiological information sequence and the real-time action expression information sequence, and obtain M answering tension coefficients.

[0025] The weight coefficient generation unit is used to perform weight allocation on the third sequence of mental health scores according to the M answering tension coefficients, and obtain M scoring weight coefficients.

[0026] The evaluation result generation unit is used to perform weighted calculation on the third sequence of mental health scores according to the M scoring weight coefficients, and generate the mental health evaluation result.

[0027] In a feasible implementation manner, the scoring compensation and optimization unit includes:

[0028] The coefficient extraction subunit is used to traverse the answering conflict evaluation sequence to extract the first answering conflict evaluation coefficient.

[0029] The threshold judgment subunit is used to judge whether the first answering conflict evaluation coefficient is greater than or equal to the answering conflict evaluation threshold.

[0030] The second activation subunit is used to activate the second mental health scoring model if the first answering conflict evaluation coefficient is greater than or equal to the answering conflict evaluation threshold.

[0031] The first parameter recording subunit is used to record the mental health score corresponding to the first answering conflict evaluation coefficient in the second sequence of mental health scores as the first parameter of the mental health score.

[0032] The second parameter obtaining subunit is used to input the psychological test answer and real-time action expression information corresponding to the first answering conflict evaluation coefficient into the second mental health scoring model, and obtain the second parameter of the mental health score.

[0033] The third parameter calculation subunit is used to perform mean calculation according to the first parameter of the mental health score and the second parameter of the mental health score, and obtain the third parameter of the mental health score.

[0034] The third sequence generation subunit is used to update and optimize the second sequence of mental health scores according to the third parameter of the mental health score, and generate the third sequence of mental health scores.

[0035] In a feasible implementation manner, the mechanism adjustment module includes:

[0036] The user test mechanism includes the physiological health test cycle and the mental health test cycle corresponding to the target user.

[0037] A mental health test cycle optimization unit, which is used to optimize the mental health test cycle in association with the mental health assessment result to obtain an optimized mental health test cycle.

[0038] A physical health test cycle adjustment unit, which is used to optimize and adjust the physical health test cycle according to the physical health assessment result, combine it with the optimized mental health test cycle to obtain a user test optimization mechanism, and perform physical and mental health tests on the target user according to the user test optimization mechanism.

[0039] In a feasible implementation manner, the physical health assessment module includes:

[0040] A real-time physical information vector construction unit, which is used to organize the real-time physical information sequence and construct a plurality of real-time physical information vectors.

[0041] A physical state deviation vector acquisition unit, which is used to identify the deviation of the plurality of real-time physical information vectors according to the standard physical state vector to obtain a plurality of physical state deviation vectors.

[0042] A physical health assessment coefficient acquisition unit, which is used to input the plurality of physical state deviation vectors into a physical health assessment model to obtain a plurality of physical health assessment coefficients.

[0043] A physical health assessment result output unit, which is used to construct a physical health assessment graph according to the plurality of physical health assessment coefficients and output the physical health assessment graph as the physical health assessment result.

[0044] In a feasible implementation manner, the non-contact AI vision real-time multi-dimensional physical and mental health test device further includes:

[0045] A user expansion unit, which is used to interconnect users with the same characteristics according to the target user to obtain a global user.

[0046] An extended retrieval unit, which is used to retrieve historical normal physical information according to the global user to obtain a normal physical sample set.

[0047] A centralized analysis unit, which is used to perform a centralized tendency analysis according to the normal physical sample set to obtain the standard physical state vector.

[0048] In a feasible implementation manner, the non-contact AI vision real-time multi-dimensional physical and mental health test device further includes:

[0049] A user information collection unit, which is used to obtain the basic user information and historical mental health test results of the target user.

[0050] A problem matching evaluation unit, configured to evaluate the matching degree of each mental health test question in the mental health test question bank according to the user's basic information and the historical mental health test results, and obtain a plurality of question matching coefficients.

[0051] A serialization unit, configured to sort the mental health test question bank in descending order according to the plurality of question matching coefficients, and construct a registered mental health test question bank.

[0052] A priority extraction unit, configured to screen the first M mental health test questions in the registered mental health test question bank to construct the mental health test question registration sequence, where M is a positive integer greater than 1.

[0053] In a second aspect, the present invention also provides a non-contact AI vision real-time multi-dimensional physiological and mental health test method, where the non-contact AI vision real-time multi-dimensional physiological and mental health test method includes:

[0054] Testing a target user according to the mental health test question registration sequence, and collecting a mental health test answer sequence, a real-time physiological information sequence, and a real-time action and expression information sequence.

[0055] Performing a health assessment on the real-time physiological information sequence according to the standard physiological state vector of the target user to obtain a physiological health assessment result.

[0056] Inputting the mental health test question registration sequence and the mental health test answer sequence into a first mental health scoring model to obtain a first mental health scoring sequence.

[0057] Performing credibility correction on the first mental health scoring sequence according to the real-time physiological information sequence to obtain a second mental health scoring sequence.

[0058] Performing conflict tracking compensation on the second mental health scoring sequence according to the real-time action and expression information sequence to obtain a mental health assessment result.

[0059] Dynamically adjusting the user test mechanism according to the mental health assessment result and the physiological health assessment result.

[0060] The present invention discloses a non-contact AI vision real-time multi-dimensional physiological and mental health testing device and method, including: a non-contact AI vision module for testing a target user according to a psychological test question registration sequence and collecting its psychological test answer sequence, real-time physiological information sequence and real-time action expression information sequence; a physiological health assessment module for health-assessing the real-time physiological information sequence based on the standard physiological state vector of the target user to generate a physiological health assessment result; a mental health assessment module for inputting the psychological test question registration sequence and the psychological test answer sequence into a first mental health scoring model to generate a first mental health scoring sequence; a credibility correction module for correcting the credibility of the first mental health scoring sequence according to the real-time physiological information sequence to generate a second mental health scoring sequence; a conflict tracking compensation module for performing conflict tracking compensation on the second mental health scoring sequence based on the real-time action expression information sequence to generate a mental health assessment result; and a mechanism adjustment module for dynamically adjusting the user test mechanism according to the mental health assessment result and the physiological health assessment result. The non-contact AI vision real-time multi-dimensional physiological and mental health testing device and method disclosed by the present invention solve the technical problems of single information dimension and insufficient test objectivity, and achieve the technical effects of improving test objectivity and comprehensiveness and improving test accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 It is a schematic structural diagram of the non-contact AI vision real-time multi-dimensional physiological and mental health testing device of the present invention;

[0062] Figure 2 It is a schematic flow diagram of the non-contact AI vision real-time multi-dimensional physiological and mental health testing method of the present invention.

[0063] Description of reference numerals: non-contact AI vision module 11, physiological health assessment module 12, mental health assessment module 13, credibility correction module 14, conflict tracking compensation module 15, mechanism adjustment module 16. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] The above technical solutions will be described in detail below in combination with the drawings in the specification and specific embodiments to better understand the above technical solutions. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used to explain the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention. In addition, it should be noted that only parts related to the present invention are shown in the drawings for the convenience of description, rather than all.

[0065] Example 1, as Figure 1This is a schematic structural diagram of the non-contact AI vision real-time multi-dimensional physiological and mental health testing device of the present invention. Among them, the non-contact AI vision real-time multi-dimensional physiological and mental health testing device includes:

[0066] A non-contact AI vision module 11, configured to test a target user according to a psychological test question registration sequence, and collect a psychological test answer sequence, a real-time physiological information sequence, and a real-time action and expression information sequence.

[0067] Specifically, first, based on a preset psychological test question registration sequence, the target user is questioned, and the user's answer content is collected and output as a psychological test answer sequence. Since the start of the test, a real-time physiological information sequence and a real-time action and expression information sequence are synchronously collected through computer vision technology, physiological signal analysis technology, and action and expression recognition technology; among them, the psychological test question registration sequence refers to an ordered arrangement and combination of a series of psychological test questions preset according to user characteristics and test purposes.

[0068] Exemplarily, the real-time physiological information sequence includes real-time change data of physiological indicators such as the user's heart rate, blood pressure, and respiratory rate; the real-time action and expression information sequence includes real-time change situations of the user's facial expressions, body movements, etc.

[0069] In the above test process, the non-contact AI vision module, as the core part of data collection, provides rich, comprehensive, and real-time multi-dimensional data support for subsequent physiological health assessment, mental health assessment, etc. It is the basis for the entire device to achieve multi-dimensional physiological and mental health testing; among them, the non-contact data collection method avoids the discomfort and interference that traditional contact sensors may bring to users, improves the user's test experience and cooperation degree, helps to reduce the deviation caused by simply relying on the user's subjective answers in traditional psychological tests, provides a reliable data basis for accurately evaluating the user's physiological and mental health status subsequently, and thus improves the accuracy of the overall test.

[0070] A physiological health assessment module 12, configured to perform a health assessment on the real-time physiological information sequence according to the standard physiological state vector of the target user, and obtain a physiological health assessment result.

[0071] Specifically, the real-time physiological information sequence refers to a series of time-series data collected during the monitoring of the target user's physiological state, such as heart rate, blood oxygen, blood pressure, skin temperature, electrocardiogram signal, etc.; the standard physiological state vector defines the physiological state of the target user under the standard state and is obtained through data analysis and model training.

[0072] The process of quantitatively evaluating the physiological health status of a user by comparing and analyzing real-time physiological data with a standard physiological state vector, and the finally obtained physiological health assessment result can reflect whether the user's current physiological state is normal and in which aspects there are abnormalities.

[0073] In some embodiments, the physiological health assessment module 12 includes:

[0074] A real-time physiological information vector construction unit, configured to sort the real-time physiological information sequence and construct a plurality of real-time physiological information vectors. A physiological state deviation vector acquisition unit, configured to perform deviation identification on the plurality of real-time physiological information vectors according to the standard physiological state vector to obtain a plurality of physiological state deviation vectors. A physiological health assessment coefficient acquisition unit, configured to input the plurality of physiological state deviation vectors into a physiological health assessment model to obtain a plurality of physiological health assessment coefficients. A physiological health assessment result output unit, configured to construct a physiological health assessment graph according to the plurality of physiological health assessment coefficients and output the physiological health assessment graph as the physiological health assessment result.

[0075] Specifically, first, the real-time physiological information vector construction unit sorts the collected real-time physiological information sequence, constructs a plurality of real-time physiological information vectors from different types of physiological data according to the time series and data characteristics, each vector corresponding to one or more specific physiological indicators, generating standardized physiological information data; then, the physiological state deviation vector acquisition unit compares these real-time physiological information vectors with the standard physiological state vector of the target user, identifies the deviation between each real-time physiological information vector and the standard vector, forming a plurality of physiological state deviation vectors, which intuitively reflect the degree and direction of the difference between the user's current physiological state and the normal state, quantifying the deviation degree of the health state; next, the physiological health assessment coefficient acquisition unit inputs these physiological state deviation vectors into a pre-trained physiological health assessment model, and the pre-defined evaluation mapping relationship (such as a mathematical model) of the model comprehensively analyzes and calculates the deviation vectors to obtain a plurality of physiological health assessment coefficients, each coefficient corresponding to one or more specific physiological health dimensions, such as a cardiovascular health coefficient, a nervous system health coefficient, etc., through which the health degree of the user in different physiological aspects is quantified; finally, the physiological health assessment result output unit constructs a physiological health assessment graph according to these physiological health assessment coefficients, and this assessment graph shows the assessment results of the user in each physiological health dimension in an intuitive visual way (such as a radar chart, a bar chart, etc.) and outputs this assessment graph as the final physiological health assessment result, providing an intuitive health status feedback.

[0076] The mental health assessment module 13 is configured to input the mental test question registration sequence and the mental test answer sequence into the first mental health scoring model to obtain the first mental health scoring sequence.

[0077] Specifically, the mental health assessment module 13 is an analysis component in the entire test device, used to quantitatively evaluate the mental state of the user. Among them, the mental test question registration sequence refers to a set of questions screened from the mental test question bank and arranged in a certain logical order based on the user's background information, historical test results, etc. These questions are carefully designed to detect the user's inner state and mental characteristics from different angles; the mental test answer sequence is a set of a series of answers given by the user in real time for the above questions.

[0078] Specifically, the first mental health scoring model is a machine learning or deep learning model pre-trained with a large amount of sample data. It is used to receive questions and answers as input features, and use the mapping rules obtained through pre-training to conduct a preliminary quantitative scoring of the user's mental health status, and output the first mental health scoring sequence, which contains the preliminary scoring values of the user in different mental dimensions.

[0079] Preferably, the first mental health scoring model is an existing mental health evaluation model, including a model based on machine learning or a mathematical model based on regression analysis, so as to reduce repetitive work and improve the overall efficiency of the device.

[0080] The credibility correction module 14 is configured to perform credibility correction on the first mental health scoring sequence according to the real-time physiological information sequence to obtain the second mental health scoring sequence.

[0081] Specifically, the credibility correction module 14 is used to adjust the credibility of the first mental health scoring sequence based on the real-time physiological information, so as to obtain the second mental health scoring sequence, thereby improving the accuracy of the mental health scoring and reducing the scoring deviation caused by the influence of the user's physiological state and emotional fluctuation factors when answering questions.

[0082] In some embodiments, the credibility correction module 14 includes:

[0083] The credibility detection unit is configured to perform credibility detection on the mental test answer sequence according to the real-time physiological information sequence based on the answer credibility threshold, and determine the credible abnormal answer sequence corresponding to the answer credible abnormal coefficient sequence.

[0084] The mental health scoring extraction unit is configured to perform data extraction on the first mental health scoring sequence based on the credible abnormal answer sequence to obtain the credible abnormal mental health scoring sequence.

[0085] A correction coefficient calculation unit for calculating a ratio based on the answer credibility anomaly coefficient sequence and the answer credibility threshold to obtain an answer credibility correction coefficient sequence.

[0086] A correction score generation unit for correcting the credible anomaly mental health score sequence based on the answer credibility correction coefficient sequence to obtain a mental health correction score sequence.

[0087] A score update unit for updating the first mental health score sequence according to the mental health correction score sequence to generate the second mental health score sequence.

[0088] Specifically, the answer credibility threshold is a standard value preset based on a large amount of experimental data and statistical analysis, used to measure the credibility of answers. Physiological information in different dimensions has different thresholds; the credibility detection unit compares each physiological index in the real-time physiological information sequence with the answer credibility threshold, calculates the answer credibility anomaly coefficient sequence, and thus determines the credible anomaly answer sequence, that is, the sequence of questions that may have untrue or unreliable answers.

[0089] Furthermore, the mental health score extraction unit extracts the mental health scores corresponding to these abnormal answers from the first mental health score sequence according to the credible anomaly answer sequence to form a credible anomaly mental health score sequence; then, the correction coefficient calculation unit calculates the ratio of each coefficient in the answer credibility anomaly coefficient sequence to the answer credibility threshold to obtain the answer credibility correction coefficient sequence, which quantitatively reflects the credibility deviation degree of each abnormal answer.

[0090] Furthermore, the correction score generation unit corrects the credible anomaly mental health score sequence using the answer credibility correction coefficient sequence, such as by multiplying the original score by the corresponding correction coefficient or performing other mathematical operations to adjust the deviation caused by the user's dishonest or unreliable answers, and generates a mental health correction score sequence; finally, the score update unit replaces or updates the values in the mental health correction score sequence to the corresponding positions in the first mental health score sequence, thereby generating a more accurate and credible second mental health score sequence.

[0091] Through the objective correction method based on physiological data, it effectively makes up for the defect of simply relying on the user's subjective answers in traditional psychological tests, improves the credibility and accuracy of mental health scores, and provides a more reliable data basis for subsequent conflict tracking compensation and test mechanism adjustment.

[0092] In some implementation manners, the credibility detection unit includes:

[0093] The answer credibility evaluation subunit is used to evaluate the credibility of the psychological test answer sequence according to the real-time physiological information sequence, and obtain M answer credibility coefficients.

[0094] The abnormal coefficient sequence establishment subunit is used to determine whether the M answer credibility coefficients are less than the answer credibility threshold, and establish an answer credibility abnormal coefficient sequence of those less than the answer credibility threshold.

[0095] The abnormal identification subunit is used to identify credible abnormal answers in the psychological test answer sequence according to the answer credibility abnormal coefficient sequence, and obtain the credible abnormal answer sequence.

[0096] Specifically, first, perform a difference analysis on the real-time physiological information sequence and the standard physiological information sequence of the target user in a calm state, and calculate the relative difference as the M answer credibility coefficients (such as using methods like Pearson correlation coefficient, Euclidean distance, etc.) to measure the credibility of the answers. For example, when a user answers a certain question, if there are physiological reactions such as a sudden increase in heart rate (such as heart rate greater than 100 beats per minute), a significant increase in skin conductance, etc., it may indicate that the user has tension, anxiety or dishonesty in this question, thus affecting the credibility of the answer. Then, determine whether the M answer credibility coefficients are less than the answer credibility threshold, and record the answer credibility coefficients less than the answer credibility threshold as answer credibility abnormal coefficients, store them in the answer credibility abnormal coefficient sequence, and then screen the corresponding psychological test answers as the credible abnormal answer sequence.

[0097] Exemplarily, the answer credibility coefficient is a value between 0 and 1, used to represent the credibility of each answer, 1 means completely credible, and 0 means completely incredible; the answer credibility threshold is a standard value preset according to a large amount of experimental data and statistical analysis, used to judge whether the answer credibility coefficient is abnormal, usually set around 0.5, and the specific value can be adjusted according to the characteristics of the test scenario and the user group. Through the comprehensive analysis of the real-time physiological information sequence and the psychological test answer sequence, the credibility of each answer can be accurately evaluated, and abnormal answers can be effectively identified.

[0098] The conflict tracking and compensation module 15 is used to perform conflict tracking and compensation on the second mental health score sequence according to the real-time action and expression information sequence, and obtain the mental health evaluation result.

[0099] Specifically, the real-time action expression information sequence includes information such as the facial expressions, body movements, and eye movement trajectories of the target user, indirectly reflecting the user's emotional state, psychological activities, and possible internal conflicts when answering questions; the second sequence of mental health scores has been corrected for credibility, but the possible contradictions or conflicts between the user's behavioral performance and the content of the answers have not been considered. Therefore, it is necessary to analyze the consistency or contradiction between the user's behavioral responses and the content of the answers through conflict tracking compensation, and further adjust and optimize the mental health scores to obtain a more real and comprehensive mental health assessment result.

[0100] In some embodiments, the conflict tracking compensation module 15 includes:

[0101] A conflict evaluation unit for evaluating the degree of conflict of the psychological test answer sequence according to the real-time action expression information sequence to obtain an answer conflict evaluation sequence.

[0102] A scoring compensation and optimization unit for compensating and optimizing the second sequence of mental health scores according to the answer conflict evaluation sequence to obtain a third sequence of mental health scores.

[0103] A tension degree evaluation unit for evaluating the answering tension degree of the target user according to the real-time physiological information sequence and the real-time action expression information sequence to obtain M answering tension coefficients.

[0104] A weight coefficient generation unit for assigning weights to the third sequence of mental health scores according to the M answering tension coefficients to obtain M scoring weight coefficients.

[0105] An evaluation result generation unit for performing weighted calculation on the third sequence of mental health scores according to the M scoring weight coefficients to generate the mental health assessment result.

[0106] Specifically, first, the conflict evaluation unit analyzes the real-time action expression information sequence, extracts feature information such as the user's facial expression changes, action frequencies, and micro-expressions when answering each question, matches and compares these behavioral features with the content of the corresponding questions, calculates the degree of conflict between the mental health scores and the expressions and actions, and forms an answer conflict evaluation sequence, which contains the quantitative values of the conflict degree between the behavior and content of each answer; for example, when the user shows a smiling or relaxed body movement when answering a question related to anxiety, it may indicate that the user is covering up or has an internal conflict.

[0107] Specifically, then, in combination with the answer conflict assessment sequence, through the conflict tracking compensation mechanism, the second mental health score sequence is corrected to generate the third mental health score sequence. For example, the overly high scores that may be caused by the user's concealment are reduced or the overly low scores caused by the user's suppression of true emotions are increased to generate a third mental health score sequence that is closer to the user's true mental state. Among them, exemplarily, different scoring items correspond to different compensation factors λ, and the compensation factor is an empirical value or empirical formula determined based on statistical analysis.

[0108] Furthermore, by combining the real-time physiological information sequence and the real-time action expression information sequence, the tension level of the user is evaluated to obtain M answer tension coefficients, which reflect the user's psychological pressure and tension state on different questions. Exemplarily, first, feature engineering processing is performed on the real-time physiological information sequence and the real-time action expression information sequence to obtain feature indicators. Then, based on the preset feature indicator fusion and mapping rules, the obtained feature indicators are converted into corresponding tension level measurement values.

[0109] Furthermore, the weight coefficient generation unit assigns weights to the third mental health score sequence according to the M answer tension coefficients, that is, a weight coefficient related to the tension level is assigned to each score. The scores corresponding to the answers with a higher tension level will be given a greater weight because this may indicate that the user has stronger emotional reactions or psychological problems on these questions. Finally, the evaluation result generation unit performs a weighted calculation on the third mental health score sequence according to the M score weight coefficients to generate the mental health evaluation result.

[0110] In the entire test scheme, the conflict tracking compensation module 15 plays a key role in the refined adjustment and optimization of the mental health score. By comprehensively analyzing multi-dimensional information such as the user's behavior, physiology, and answer content, the comprehensiveness and accuracy of the evaluation result are ensured.

[0111] In some implementation manners, the score compensation and optimization unit includes:

[0112] A coefficient extraction subunit for traversing the answer conflict evaluation sequence to extract the first answer conflict evaluation coefficient. A threshold judgment subunit for judging whether the first answer conflict evaluation coefficient is greater than or equal to the answer conflict evaluation threshold. A second activation subunit for activating the second mental health scoring model if the first answer conflict evaluation coefficient is greater than or equal to the answer conflict evaluation threshold. A first parameter recording subunit for recording the mental health score corresponding to the first answer conflict evaluation coefficient in the second mental health scoring sequence as the first mental health scoring parameter. A second parameter obtaining subunit for inputting the psychological test answer and real-time action expression information corresponding to the first answer conflict evaluation coefficient into the second mental health scoring model to obtain the second mental health scoring parameter. A third parameter calculation subunit for performing a mean calculation based on the first mental health scoring parameter and the second mental health scoring parameter to obtain the third mental health scoring parameter. A third sequence generation subunit for updating and optimizing the second mental health scoring sequence according to the third mental health scoring parameter to generate the third mental health scoring sequence.

[0113] Specifically, first, the coefficient extraction subunit traverses the answer conflict evaluation sequence and extracts each first answer conflict evaluation coefficient one by one. These coefficients represent the quantified values of the conflict degree between each answer's behavior and content. Then, the extracted first answer conflict evaluation coefficient is compared with the preset answer conflict evaluation threshold to judge whether it is greater than or equal to the threshold. Among them, the answer conflict evaluation threshold is a standard value preset according to a large amount of experimental data and statistical analysis, used to determine whether the answer conflict degree needs to be compensated and optimized. It is usually set near the medium conflict degree, and the specific value can be adjusted according to the characteristics of the test scenario and user group.

[0114] Specifically, if the first answer conflict evaluation coefficient is greater than or equal to the answer conflict evaluation threshold, it can be considered that there is an obvious conflict in this answer and further compensation and optimization are needed. At this time, the second activation subunit activates the second mental health scoring model to re-evaluate and score the answer with conflict. Among them, the second mental health scoring model is a mental health evaluation model that considers the action expression dimension information and is used to map the psychological test answer and the corresponding action expression information into the second mental health scoring parameter through the embedded mapping rule.

[0115] Specifically, the psychological test answer and real-time action expression information corresponding to the first answer conflict evaluation coefficient are input into the second mental health scoring model, and the second mental health scoring parameter corresponding to this answer is output. This parameter is the re-evaluation result of the user's mental state after considering the answer conflict situation.

[0116] Synchronously, record the mental health score corresponding to the conflict evaluation coefficient in the second sequence of mental health scores, and use it as the first parameter of the mental health score. Furthermore, calculate the mean value based on the first parameter of the mental health score and the second parameter of the mental health score to obtain the third parameter of the mental health score. By calculating the mean value, the original score and the re-evaluation score are comprehensively considered, avoiding the extremity of a single score and making the final score more balanced and reasonable.

[0117] Optionally, in practical applications, other more complex fusion algorithms can be adopted according to specific situations, such as weighted average, regression analysis, etc.

[0118] Furthermore, update the third parameter of the mental health score to the corresponding position in the second sequence of mental health scores to generate a new third sequence of mental health scores, and make targeted score adjustments for conflicting answers, so that the final mental health score can more truly reflect the actual mental state of the user, improving the accuracy and credibility of the evaluation results.

[0119] The above-mentioned targeted compensation process based on the degree of conflict effectively solves the problem of inconsistent answers and behaviors caused by users' concealment and suppression in traditional psychological tests, improves the objectivity and accuracy of mental health evaluation results, and helps to accurately evaluate and effectively intervene in users' mental health.

[0120] The mechanism adjustment module 16 is used to dynamically adjust the user test mechanism according to the mental health evaluation result and the physical health evaluation result.

[0121] Specifically, through the mental health evaluation result and the physical health evaluation result, the real-time state of the target user can be comprehensively reflected, and the strategy for testing and analyzing the target user can be correspondingly adjusted, such as adjusting the test cycle, optimizing test questions (quantity, type, questioning depth, etc.), improving the test environment, etc., to ensure that the test process can better adapt to the current state and needs of the user.

[0122] In some embodiments, the mechanism adjustment module 16 includes:

[0123] The user test mechanism includes the physical health test cycle and the mental health test cycle corresponding to the target user. The mental health test cycle optimization unit is used to perform associated optimization on the mental health test cycle according to the mental health evaluation result to obtain an optimized mental health test cycle. The physical health test cycle adjustment unit is used to optimize and adjust the physical health test cycle according to the physical health evaluation result, combine it with the optimized mental health test cycle to obtain a user test optimization mechanism, and perform physical and mental health tests on the target user according to the user test optimization mechanism.

[0124] Optionally, the user testing mechanism includes a physiological health testing cycle and a psychological health testing cycle for the target user; specifically, the user testing mechanism is adjusted as follows: First, the assessment results of psychological health and physiological health are received; then, the psychological health testing cycle optimization unit adjusts the psychological health testing cycle according to the psychological health assessment result. For example, if the user scores low in certain psychological dimensions, indicating potential problems, the optimization unit can appropriately shorten the testing intervals for these dimensions and increase the testing frequency according to the preset cycle mapping relation table to more closely monitor the user's psychological changes; at the same time, through the physiological health testing cycle adjustment unit, the physiological health testing cycle is optimized according to the physiological health assessment result. Similarly, if there are abnormal fluctuations in certain physiological indicators of the user, the testing cycles of these indicators are adjusted accordingly to ensure that changes in the physiological state can be captured in a timely manner.

[0125] Furthermore, on the basis of adjusting the testing cycle, the mechanism adjustment module 16 also optimizes the selection and presentation method of the test questions. For example, according to the user's health status and test history, those test questions with higher relevance to the user's current problems are preferentially selected, or the presentation order of the questions is adjusted, and the more diagnostic questions are placed in the front to improve the efficiency and accuracy of the test.

[0126] Optionally, the depth and breadth of the test can also be dynamically adjusted according to the user's health change trend. For example, for users with relatively stable health conditions, the testing frequency and the number of questions can be appropriately reduced to reduce the user's testing burden; while for users with large fluctuations in health conditions or potential risks, the depth and coverage of the test are increased to ensure a comprehensive and in-depth understanding of the user's health status.

[0127] By comprehensively analyzing the psychological health assessment result and the physiological health assessment result, the dynamic and intelligent adjustment of the user testing mechanism is realized. This adjustment can ensure that the testing process always matches the user's health status, avoiding problems of insufficient testing or over-testing that may be brought about by a fixed testing mechanism.

[0128] In some embodiments, the non-contact AI vision real-time multi-dimensional physiological and psychological health testing device further includes:

[0129] A user expansion unit for interconnecting users with the same characteristics based on the target user to obtain global users. An extended retrieval unit for retrieving historical normal physiological information based on the global users to obtain a normal physiological sample set. A centralized analysis unit for performing centralized tendency analysis based on the normal physiological sample set to obtain the standard physiological state vector.

[0130] Specifically, first, the user interconnection module identifies the characteristic information of the target user (such as age, gender, living habits, physiological characteristics, etc.), and based on methods such as cluster analysis, screens for similar users, establishes user associations, forms a user group with the same characteristics, and outputs it as the global user set; then, historical normal physiological information is screened from the global user set to form a normal physiological sample set; next, a central tendency analysis is performed on the normal physiological sample set to obtain a standard physiological state vector. Exemplarily, the mean, median or standard deviation is calculated as the result of the central tendency analysis to ensure the robustness of the physiological state benchmark.

[0131] In some embodiments, the non-contact AI vision real-time multi-dimensional physiological and mental health testing device further includes:

[0132] A user information collection unit for obtaining the basic user information of the target user and the historical mental health test results. A problem matching evaluation unit for evaluating the matching degree of each mental health test problem in the mental health test question bank according to the basic user information and the historical mental health test results, and obtaining a plurality of problem matching coefficients. A serialization unit for sorting the mental health test question bank in descending order according to the plurality of problem matching coefficients to construct a registered mental health test question bank. A priority extraction unit for screening the first M mental health test problems in the registered mental health test question bank to construct the mental test question registration sequence, where M is a positive integer greater than 1.

[0133] Specifically, the user information collection unit is used to collect the basic user information of the target user and the historical mental health test results. Among them, the basic user information includes age, gender, occupation, education level, etc., which can help to understand the background characteristics of the user, equivalent to a user portrait; the historical mental health test results are the records of the mental health tests participated by the user in the past, including test time, test items, test scores, etc., reflecting the change trend of the user's mental health status.

[0134] Specifically, the problem matching evaluation unit evaluates the matching degree of each problem in the mental health test question bank based on the cosine similarity according to the basic user information and the historical mental health test results, and obtains a plurality of problem matching coefficients; among them, each problem in the mental health test question bank is marked with corresponding customer group information (corresponding to the basic user information and the historical mental health test results); the problem matching coefficient is a quantitative index used to measure the degree of fit between each test problem and the user characteristics and needs.

[0135] Further, the serialization unit sorts the mental health test question bank in descending order according to the question matching coefficient, constructs a registered mental health test question bank, and screens the first M mental health test questions in the registered mental health test question bank through the priority extraction unit to construct a mental health test question registration sequence, where M is a positive integer greater than 1, which determines the depth and breadth of the test.

[0136] Through the above process, the most relevant questions are selected based on the user characteristics, ensuring the pertinence and personalization of the test question selection, which can better meet the user's needs and improve the efficiency and accuracy of the test. For example, for users with depressive tendencies, questions related to depression are preferentially selected to more deeply evaluate their mental state; while for users with good mental health, some preventive and developmental test questions can be selected to help them further improve their mental quality. In addition, by dynamically adjusting the value of M, a balance can be achieved between the test time and the test depth, improving the flexibility and practicality of the test.

[0137] In summary, the non-contact AI vision real-time multi-dimensional physiological and mental health test device provided by the present invention has the following technical effects:

[0138] Through the non-contact AI vision module, it is used to test the target user according to the mental health test question registration sequence and collect the mental health test answer sequence, real-time physiological information sequence and real-time action expression information sequence of the target user; the physiological health assessment module is used to perform a health assessment on the real-time physiological information sequence based on the standard physiological state vector of the target user to generate a physiological health assessment result; the mental health assessment module is used to input the mental health test question registration sequence and the mental health test answer sequence into the first mental health scoring model to generate the first mental health scoring sequence; the credibility correction module is used to correct the credibility of the first mental health scoring sequence according to the real-time physiological information sequence to generate the second mental health scoring sequence; the conflict tracking compensation module is used to perform conflict tracking compensation on the second mental health scoring sequence based on the real-time action expression information sequence to generate a mental health assessment result; the mechanism adjustment module is used to dynamically adjust the user test mechanism according to the mental health assessment result and the physiological health assessment result, so as to achieve the technical effects of improving the objectivity and comprehensiveness of the test and improving the test accuracy.

[0139] Embodiment 2, as Figure 2 is a schematic flowchart of the non-contact AI vision real-time multi-dimensional physiological and mental health test method of the present invention. For example, Figure 1 in the schematic structural diagram of the non-contact AI vision real-time multi-dimensional physiological and mental health test device of the present invention can be used to implement the process as Figure 2 shown.

[0140] Based on the same concept as the non-contact AI vision real-time multi-dimensional physiological and mental health testing device in the above embodiments, the non-contact AI vision real-time multi-dimensional physiological and mental health testing method provided by the present invention includes:

[0141] S100: Test the target user according to the psychological test question registration sequence, and collect the psychological test answer sequence, real-time physiological information sequence, and real-time action and expression information sequence.

[0142] S200: Perform a health assessment on the real-time physiological information sequence according to the standard physiological state vector of the target user to obtain a physiological health assessment result.

[0143] S300: Input the psychological test question registration sequence and the psychological test answer sequence into the first mental health scoring model to obtain the first mental health scoring sequence.

[0144] S400: Perform credibility correction on the first mental health scoring sequence according to the real-time physiological information sequence to obtain the second mental health scoring sequence.

[0145] S500: Perform conflict tracking compensation on the second mental health scoring sequence according to the real-time action and expression information sequence to obtain a physiological and mental health assessment result.

[0146] S600: Dynamically adjust the user test mechanism according to the physiological and mental health assessment results and the physiological health assessment result.

[0147] In some embodiments, performing credibility correction on the first mental health scoring sequence according to the real-time physiological information sequence to obtain the second mental health scoring sequence includes:

[0148] Based on the answer credibility threshold, perform credibility detection on the psychological test answer sequence according to the real-time physiological information sequence to determine the credible abnormal answer sequence corresponding to the answer credibility abnormal coefficient sequence. Extract data from the first mental health scoring sequence based on the credible abnormal answer sequence to obtain a credible abnormal mental health scoring sequence. Calculate the ratio of the answer credibility abnormal coefficient sequence to the answer credibility threshold to obtain an answer credibility correction coefficient sequence. Correct the credible abnormal mental health scoring sequence based on the answer credibility correction coefficient sequence to obtain a mental health corrected scoring sequence. Update the first mental health scoring sequence according to the mental health corrected scoring sequence to generate the second mental health scoring sequence.

[0149] In some implementation manners, based on the answer credibility threshold, performing credibility detection on the psychological test answer sequence according to the real-time physiological information sequence to determine the credible abnormal answer sequence corresponding to the answer credibility abnormal coefficient sequence includes:

[0150] Perform credibility evaluation on the psychological test answer sequence according to the real-time physiological information sequence to obtain M answer credibility coefficients. Determine whether the M answer credibility coefficients are less than the answer credibility threshold, and establish an answer credibility anomaly coefficient sequence less than the answer credibility threshold. Perform credible anomaly answer recognition on the psychological test answer sequence according to the answer credibility anomaly coefficient sequence to obtain the credible anomaly answer sequence.

[0151] In some embodiments, conflict tracking compensation is performed on the second mental health score sequence according to the real-time action expression information sequence to obtain a mental health assessment result, including:

[0152] Perform conflict degree evaluation on the psychological test answer sequence according to the real-time action expression information sequence to obtain an answer conflict evaluation sequence. Perform compensation optimization on the second mental health score sequence according to the answer conflict evaluation sequence to obtain a third mental health score sequence. Evaluate the answer tension degree of the target user according to the real-time physiological information sequence and the real-time action expression information sequence to obtain M answer tension coefficients. Perform weight assignment on the third mental health score sequence according to the M answer tension coefficients to obtain M scoring weight coefficients. Perform weighted calculation on the third mental health score sequence according to the M scoring weight coefficients to generate the mental health assessment result.

[0153] In some embodiments, performing compensation optimization on the second mental health score sequence according to the answer conflict evaluation sequence to obtain a third mental health score sequence includes:

[0154] Traverse the answer conflict evaluation sequence to extract the first answer conflict evaluation coefficient. Determine whether the first answer conflict evaluation coefficient is greater than or equal to the answer conflict evaluation threshold. If the first answer conflict evaluation coefficient is greater than or equal to the answer conflict evaluation threshold, activate the second mental health score model. Denote the mental health score corresponding to the first answer conflict evaluation coefficient in the second mental health score sequence as the first mental health score parameter. Input the psychological test answer and real-time action expression information corresponding to the first answer conflict evaluation coefficient into the second mental health score model to obtain the second mental health score parameter. Perform mean calculation according to the first mental health score parameter and the second mental health score parameter to obtain the third mental health score parameter. Update and optimize the second mental health score sequence according to the third mental health score parameter to generate the third mental health score sequence.

[0155] In some embodiments, dynamically adjusting the user test mechanism according to the mental health assessment result and the physical health assessment result includes:

[0156] The user test mechanism includes the physiological health test cycle and the mental health test cycle corresponding to the target user. The mental health test cycle is associated and optimized according to the mental health assessment result to obtain an optimized mental health test cycle. According to the physiological health assessment result, the physiological health test cycle is optimized and adjusted. Combining the optimized mental health test cycle, a user test optimization mechanism is obtained, and the target user is subjected to physiological and mental health tests according to the user test optimization mechanism.

[0157] In some embodiments, the real-time physiological information sequence is health-assessed according to the standard physiological state vector of the target user to obtain a physiological health assessment result, including:

[0158] The real-time physiological information sequence is sorted out to construct a plurality of real-time physiological information vectors. Deviation identification is performed on the plurality of real-time physiological information vectors according to the standard physiological state vector to obtain a plurality of physiological state deviation vectors. The plurality of physiological state deviation vectors are input into a physiological health assessment model to obtain a plurality of physiological health assessment coefficients. According to the plurality of physiological health assessment coefficients, a physiological health assessment graph is constructed, and the physiological health assessment graph is output as the physiological health assessment result.

[0159] In some embodiments, the non-contact AI vision real-time multi-dimensional physiological and mental health test method further includes:

[0160] The target user is interconnected with users with the same characteristics to obtain a global user. Historical normal physiological information is retrieved according to the global user to obtain a normal physiological sample set. Central tendency analysis is performed on the normal physiological sample set to obtain the standard physiological state vector.

[0161] In some embodiments, the non-contact AI vision real-time multi-dimensional physiological and mental health test method further includes:

[0162] The user basic information and historical mental health test results of the target user are obtained. The matching degrees of each mental health test question in the mental health test question bank are evaluated according to the user basic information and the historical mental health test results to obtain a plurality of question matching coefficients. The mental health test question bank is sorted in descending order according to the plurality of question matching coefficients to construct a registered mental health test question bank. The first M mental health test questions in the registered mental health test question bank are selected to construct the mental test question registration sequence, where M is a positive integer greater than 1.

[0163] It should be understood that the embodiments mentioned in this specification focus on their differences from other embodiments. The specific embodiments in the foregoing Embodiment 1 are equally applicable to the non-contact AI vision real-time multi-dimensional physiological and mental health testing method described in Embodiment 2. For the sake of brevity of the specification, no further elaboration will be made here.

[0164] It should be understood that the disclosed embodiments of the present invention and the above descriptions enable those skilled in the art to implement the present invention using the present invention. At the same time, the present invention is not limited to the part of the embodiments mentioned above. It should be understood that those of ordinary skill in the art can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements on some of the technical features; and 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 embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. Non-contact AI visual real-time multi-dimensional physiological and psychological health testing device, characterized by: include: The non-contact AI vision module is used to test the target user according to the psychological test question registration sequence, and collect the psychological test answer sequence, real-time physiological information sequence and real-time action expression information sequence; A physiological health assessment module, used to perform a health assessment on the real-time physiological information sequence according to the standard physiological state vector of the target user to obtain a physiological health assessment result; A mental health assessment module, used for inputting the psychological test question registration sequence and the psychological test answer sequence into a first mental health score model to obtain a first mental health score sequence; A credibility correction module, used for performing credibility correction on the first sequence of mental health scores according to the real-time physiological information sequence to obtain a second sequence of mental health scores; A conflict tracking compensation module, used for performing conflict tracking compensation on the second sequence of mental health scores according to the real-time action expression information sequence to obtain a mental health assessment result; A mechanism adjustment module, used for dynamically adjusting the user testing mechanism according to the mental health assessment result and the physiological health assessment result; The credibility correction module comprises: A credibility detection unit, configured to perform credibility detection on the psychological test answer sequence according to the real-time physiological information sequence based on the answer credibility threshold, and determine a credible abnormal answer sequence corresponding to the answer credibility abnormal coefficient sequence; A mental health score extraction unit, configured to extract data from the first sequence of mental health scores based on the credible abnormal answer sequence to obtain a credible abnormal mental health score sequence; A correction coefficient calculation unit, used for performing ratio calculation based on the answer credibility abnormal coefficient sequence and the answer credibility threshold value to obtain an answer credibility correction coefficient sequence; A correction score generating unit, used for correcting the credible abnormal mental health score sequence based on the answer credible correction coefficient sequence to obtain a mental health correction score sequence; A score updating unit, configured to update the first sequence of mental health scores according to the mental health correction score sequence to generate the second sequence of mental health scores; The conflict tracking and compensation module includes: A conflict assessment unit, configured to perform conflict degree assessment on the psychological test answer sequence according to the real-time action expression information sequence, and obtain an answer conflict assessment sequence; A scoring compensation optimization unit, used for compensating and optimizing the second sequence of mental health scores according to the answer conflict evaluation sequence to obtain a third sequence of mental health scores; A tension evaluation unit, used to evaluate the target user's answer tension according to the real-time physiological information sequence and the real-time action expression information sequence, and obtain M answer tension coefficients; A weight coefficient generating unit, used for weighting the third sequence of mental health scores according to the M answer stress coefficients to obtain M score weight coefficients; An evaluation result generating unit is used to perform weighted calculation on the third sequence of mental health scores according to the M scoring weight coefficients to generate the mental health evaluation result.

2. The non-contact AI visual real-time multi-dimensional physiological and psychological health testing device according to claim 1, characterized in that: The credibility detection unit comprises: An answer credibility evaluation subunit is used to perform credibility evaluation on the psychological test answer sequence according to the real-time physiological information sequence to obtain M answer credibility coefficients; An abnormal coefficient sequence establishing subunit, used for judging whether the M answer credibility coefficients are less than the answer credibility threshold, and establishing an answer credibility abnormal coefficient sequence less than the answer credibility threshold; The abnormality identification subunit is used to identify credible abnormal answers to the psychological test answer sequence according to the credible abnormality coefficient sequence of the answers to obtain the credible abnormal answer sequence.

3. The non-contact AI visual real-time multi-dimensional physiological and psychological health testing device according to claim 1, characterized in that: The scoring compensation optimization unit includes: A coefficient extraction subunit, used for traversing the answer conflict evaluation sequence to extract a first answer conflict evaluation coefficient; a threshold judgment subunit, configured to judge whether the first answer conflict assessment coefficient is greater than or equal to an answer conflict assessment threshold; A second activation subunit is used to activate a second mental health score model if the first answer conflict assessment coefficient is greater than or equal to the answer conflict assessment threshold; A first parameter recording subunit is used to record the mental health score corresponding to the first answer conflict assessment coefficient in the second sequence of mental health scores as a first parameter of the mental health score; A second parameter acquisition subunit is used to input the psychological test answer and real-time action expression information corresponding to the first answer conflict assessment coefficient into the second mental health score model to obtain a second mental health score parameter; A third parameter calculation subunit is used to perform mean calculation according to the first mental health score parameter and the second mental health score parameter to obtain a third mental health score parameter; The third sequence generating subunit is used to update and optimize the second sequence of mental health scores according to the third parameter of the mental health scores to generate the third sequence of mental health scores.

4. The non-contact AI visual real-time multi-dimensional physiological and psychological health testing device according to claim 1, characterized in that: The mechanism adjustment module includes: The user testing mechanism includes a physiological health testing cycle and a mental health testing cycle corresponding to the target user; A mental health test cycle optimization unit, used to optimize the mental health test cycle according to the mental health assessment result to obtain a mental health test optimization cycle; The physiological health test cycle adjustment unit is used to optimize and adjust the physiological health test cycle according to the physiological health assessment results, combine the psychological health test optimization cycle to obtain a user test optimization mechanism, and perform physiological and psychological health tests on the target user according to the user test optimization mechanism.

5. The non-contact AI visual real-time multi-dimensional physiological and psychological health testing device according to claim 1, characterized in that: The physiological health assessment module comprises: A real-time physiological information vector construction unit, used to organize the real-time physiological information sequence and construct a plurality of real-time physiological information vectors; A physiological state deviation vector acquisition unit, configured to perform deviation identification on the plurality of real-time physiological information vectors according to the standard physiological state vector to obtain a plurality of physiological state deviation vectors; A physiological health assessment coefficient acquisition unit, used for inputting the plurality of physiological state deviation vectors into a physiological health assessment model to obtain a plurality of physiological health assessment coefficients; The physiological health assessment result output unit is used to construct a physiological health assessment graph according to the multiple physiological health assessment coefficients, and output the physiological health assessment graph as the physiological health assessment result.

6. The non-contact AI visual real-time multi-dimensional physiological and psychological health testing device according to claim 1, characterized in that: Also includes: A user expansion unit is used to interconnect users with the same characteristics according to the target user to obtain global users; An expanded retrieval unit, configured to retrieve historical normal physiological information according to the global user to obtain a normal physiological sample set; The centralized analysis unit is used to perform centralized trend analysis based on the normal physiological sample set to obtain the standard physiological state vector.

7. The non-contact AI visual real-time multi-dimensional physiological and psychological health testing device according to claim 1, characterized in that: Also includes: A user information collection unit, used to obtain basic user information and historical mental health test results of the target user; A question matching evaluation unit, used to evaluate the matching degree of each mental health test question in the mental health test question library according to the user basic information and the historical mental health test results, and obtain multiple question matching coefficients; A serialization unit, for arranging the mental health test question library in descending order according to the plurality of question matching coefficients, and constructing a registration psychological test question library; The priority extraction unit is used to screen the first M psychological health test questions in the registered psychological test question library and construct the psychological test question registration sequence, wherein M is a positive integer greater than 1.

8. Non-contact AI visual real-time multi-dimensional physiological and psychological health testing method, characterized in that: A non-contact AI visual real-time multi-dimensional physiological and psychological health testing device for executing any one of claims 1 to 7, comprising: Test the target user according to the psychological test question registration sequence, and collect the psychological test answer sequence, real-time physiological information sequence and real-time action expression information sequence; Performing a health assessment on the real-time physiological information sequence according to the standard physiological state vector of the target user to obtain a physiological health assessment result; Inputting the psychological test question registration sequence and the psychological test answer sequence into a first mental health score model to obtain a first mental health score sequence; Performing credibility correction on the first sequence of mental health scores according to the real-time physiological information sequence to obtain a second sequence of mental health scores; Performing conflict tracking compensation on the second sequence of mental health scores according to the real-time action expression information sequence to obtain a mental health assessment result; The user testing mechanism is dynamically adjusted according to the mental health assessment results and the physiological health assessment results.

Images