Facial recognition-based wounded consciousness state evaluation system

Abstract

The invention provides a wounded consciousness state evaluation system based on face recognition, and relates to the technical field of health evaluation. Comprising a facial information extraction module, a facial feature analysis module, a consciousness state calculation module, an anomaly detection module and a consciousness state evaluation module. According to the invention, the infrared camera carefully captures the fine movement of the muscles around the eyes, the mouth corner and the frontal part, the accuracy and comprehensiveness of data acquisition are improved, the accuracy and dimension of evaluation are improved by evaluating the blinking frequency, the mouth corner displacement and the frontal muscle stretching amplitude, and by utilizing the extraction of the deformation characteristics and the calculation of the movement amplitude gradient, the accuracy and the comprehensiveness of the evaluation are improved. The time sequence analysis is combined, the evaluation of the consciousness state becomes more sensitive and accurate, potential changes of consciousness can be found in time, and therefore, rich data interpretation is provided for diagnosis and monitoring through quick response, and the density and distribution characteristic analysis of sudden change points, so that the consciousness state can be quickly and accurately evaluated under the condition that a wounded person is not contacted.

Description

Technical Field

[0001] The present invention relates to the technical field of health assessment, and particularly to a system for assessing the consciousness state of wounded personnel based on facial recognition. Background Art

[0002] The technical field of health assessment includes a variety of methods and devices for monitoring and evaluating an individual's health status. Especially in clinical medicine and remote medical monitoring, this field covers from basic vital sign monitoring to complex pathological state analysis, including the use of biosensors, medical imaging techniques, and various physiological parameter analysis methods. In addition, health assessment technology increasingly adopts digital tools and algorithms to provide real-time and accurate health status information, which is crucial for disease prevention, early diagnosis, and treatment monitoring.

[0003] Among them, a system for assessing the consciousness state of wounded personnel based on facial recognition refers to a system that uses facial image processing technology to analyze the consciousness level of wounded personnel. It evaluates the consciousness state through facial expression recognition and eye movement analysis. Specifically, it includes using an image capture device to obtain the facial image of the injured person, analyzing the expression changes through a facial expression recognition algorithm, and evaluating the response ability and consciousness level of the injured person through eye movement tracking. By integrating facial recognition technology with medical assessment requirements, a non-invasive and rapidly responsive consciousness state monitoring tool is constructed.

[0004] The prior art relies on traditional biosensors and basic vital sign monitoring, and cannot quickly provide consciousness state information in dynamic and complex environments. Relying on basic physiological parameter analysis and medical imaging, it is difficult to analyze facial expression changes and their impact on the consciousness state. In emergency medical rescue, it is impossible to quickly evaluate the accurate consciousness level of wounded personnel, delaying treatment and increasing the risk of wounded personnel. Especially when dealing with large-scale disaster responses, it cannot meet the requirements of rapid deployment and efficient operation. Summary of the Invention

[0005] In order to solve the technical problems existing in the prior art, which rely on traditional biosensors and basic vital sign monitoring, cannot quickly provide consciousness state information in dynamic and complex environments, rely on basic physiological parameter analysis and medical imaging, are difficult to analyze facial expression changes and their impact on the consciousness state, in emergency medical rescue, cannot quickly evaluate the accurate consciousness level of wounded personnel, delaying treatment and increasing the risk of wounded personnel, especially when dealing with large-scale disaster responses, cannot meet the requirements of rapid deployment and efficient operation, the embodiments of the present invention provide a system for assessing the consciousness state of wounded personnel based on facial recognition. The technical solution is as follows:

[0006] On the one hand, a system for assessing the consciousness state of wounded personnel based on facial recognition is provided, including:

[0007] The facial information extraction module uses an infrared camera to collect the motion data of the eye area, the corners of the mouth, and the frontal muscles, calculates the blink frequency, the displacement of the corners of the mouth, and the stretching amplitude of the frontal muscles, and obtains the key sequence of facial movements;

[0008] The facial feature analysis module extracts deformation features based on the key sequence of facial movements, calculates the motion amplitude gradient of the facial area, analyzes the gradient difference, identifies feature points, and obtains the facial dynamic features;

[0009] The consciousness state calculation module analyzes the facial movement trend based on the facial dynamic features, calculates the movement change rate, identifies and counts the distribution of change rate mutation points, determines the influence of the injured area on the mutation points, and obtains the analysis result of the movement trend deviation;

[0010] The anomaly detection module measures the mutation point density based on the analysis result of the movement trend deviation, compares the distribution characteristics of the mutation points, determines whether the number of mutation points exceeds the average value of the mutation points, marks the abnormal mutation points and classifies them, and generates the mutation point marking data;

[0011] The consciousness state evaluation module calculates the duration, distribution density, and time stability of the mutation points based on the mutation point marking data, determines the type and pattern of abnormal mutations, classifies the consciousness level, and obtains the consciousness state evaluation result.

[0012] Optionally, the key sequence of facial movements includes the eyelid movement sequence, the corner-of-mouth movement sequence, and the frontal expression sequence, the facial dynamic features include the deformation rate, the deformation period, and the deformation amplitude, the analysis result of the movement trend deviation includes the mutation frequency, the mutation persistence, and the mutation intensity, the mutation point marking data includes the abnormal timestamp, the abnormal duration, and the abnormal intensity level, and the consciousness state evaluation result includes the consciousness state stability, the consciousness state change frequency, and the consciousness state classification level..

[0013] Optionally, the facial information extraction module includes;

[0014] The facial movement data acquisition sub-module uses an infrared camera to collect the motion change data of the eye area, the corners of the mouth, and the frontal muscles, extracts the position information of the facial monitoring points in adjacent time frames, calculates the contraction amplitude of the eye area, the movement trajectory of the corners of the mouth, and the stretching length of the frontal muscles, calibrates the movement interval range, and generates the facial movement trajectory;

[0015] The facial deformation analysis sub-module calculates the deformation gradient of the facial monitoring points within adjacent time frames based on the facial movement trajectory, screens the movement points that meet the deformation criteria, and analyzes the blink closing cycle, the offset of the corners of the mouth, and the stretching degree of the frontal muscles within consecutive time frames to obtain the facial deformation features;

[0016] The key sequence screening sub-module calls the facial deformation features, screens the facial motion sequences, judges the continuity of the motion sequences, screens the stable motion points according to the time axis, calculates the matching degree of the facial motion patterns, determines the time period with stable motion patterns, and obtains the key facial motion sequences.

[0017] Optionally, the facial feature analysis module includes;

[0018] Based on the key facial motion sequences, the deformation feature extraction sub-module analyzes the motion trajectories of the areas around the eyes, the corners of the mouth, and the frontal muscles, identifies and marks the key feature points, obtains the position changes of each feature point in consecutive time frames, and generates the key feature motion data;

[0019] The gradient calculation sub-module calls the key feature motion data, calculates the motion amplitude gradient values of the areas around the eyes, the corners of the mouth, and the frontal muscles, compares the positions of the motion points in consecutive time frames, analyzes the gradient change rate between adjacent time frames, screens the gradient data exceeding the motion amplitude standard, and obtains the motion gradient feature values;

[0020] Based on the motion gradient feature values, the trend analysis sub-module analyzes the change trend of the facial motion, compares the facial dynamics in different time periods, tracks the motion states of the feature points in local areas, judges the stability and change patterns of the facial dynamics, and obtains the facial dynamic feature trend.

[0021] Optionally, when calculating the motion amplitude gradients of the areas around the eyes, the corners of the mouth, and the frontal muscles, the formula is used:

[0022]

[0023] Compare the positions of the motion points in consecutive time frames, analyze the gradient change rate between adjacent time frames, screen the gradient data exceeding the motion amplitude standard, and obtain the motion gradient feature values;

[0024] where G ij represents the motion amplitude gradient value of the j-th feature point in the i-th area between adjacent time frames, represents the displacement of the feature point in the k-th frame of the i-th area, represents the displacement of the j-th feature point in the adjacent frame, w ij represents the weight coefficient of the j-th feature point in the i-th area, Δt k represents the time interval between the k-th frame and the k - 1-th frame, α ij represents the deformation correction parameter of the j-th feature point in the i-th area.

[0025] Optionally, the consciousness state calculation module includes;

[0026] The rate-of-change calculation sub-module analyzes the rate of change of the facial movement amplitude based on the facial dynamic feature trend, calculates the difference value of the facial movement speed within adjacent time windows, filters out the time intervals where the rate of change exceeds the set standard, and generates a movement amplitude deviation feature.

[0027] The mutation point extraction sub-module calls the movement amplitude deviation feature, compares the rate-of-change gradients of adjacent time windows, identifies the rate-of-change mutation points, filters out the time points where the movement speed deviates from the standard, extracts the distribution of the mutation points on the time axis, analyzes the continuity and change trend of the mutation points, and obtains a mutation point characteristic map.

[0028] The error weight allocation sub-module analyzes the corresponding relationship between the time positions of the mutation points and the injured areas based on the mutation point characteristic map, determines the degree of influence of the mutation points on the facial movement pattern, adjusts the error allocation ratio of the injured areas, determines the influence range of the mutation points, and obtains the analysis result of the movement trend deviation.

[0029] Optionally, the anomaly detection module includes;

[0030] The mutation point density calculation sub-module extracts the mutation point data and the number of mutation points within a time window based on the movement trend deviation analysis result, calculates the average density of the mutation points within a time period, and obtains the mutation point density distribution.

[0031] The anomaly screening sub-module calls the mutation point density distribution, compares the mutation point distribution characteristics within adjacent time periods, calculates the change amplitude of the number of mutation points, determines whether the number of mutation points within adjacent time periods exceeds the average density, screens out the abnormal mutation points and marks the time nodes, and obtains the abnormal distribution nodes.

[0032] The anomaly classification sub-module analyzes the time distribution law of the abnormal mutation points based on the abnormal distribution nodes, calculates the weight frequency ratio of the abnormal mutation points, classifies the abnormal types of the mutation points, determines the change situation of the abnormal types within each time period, and generates mutation point marking data.

[0033] Optionally, the change density ratio of the abnormal mutation points is calculated using the formula:

[0034]

[0035] Classify the abnormal types of the mutation points, determine the change situation of the abnormal types within each time period, and generate mutation point marking data;

[0036] Where D r represents the change density ratio of the abnormal mutation points, Q e represents the number of abnormal mutation points within the e-th time interval, Q e-1 represents the number of abnormal mutation points within the previous time interval, f represents the number of time intervals, Ze represents the end time of the e-th time interval, Z e-1 represents the end time of the previous time interval.

[0037] Optionally, the consciousness state evaluation module includes;

[0038] Based on the mutation point marking data, the mutation point feature calculation sub-module extracts the start time of the mutation point, calculates the mutation point interval time and duration, analyzes the time stability of the mutation point within the differential time window, and obtains the mutation time stability index;

[0039] The abnormal pattern analysis sub-module calls the mutation time stability index, analyzes the change law of the mutation point on the time axis, compares the distribution of mutation points within the differential time period, identifies the fluctuation trend of the number of mutation points, and judges the type and law of abnormal mutations to obtain abnormal change identification data;

[0040] Based on the abnormal change identification data, the state level division sub-module calculates the proportion of mutation point types, determines the corresponding relationship between the mutation mode and the consciousness state according to the consciousness state evaluation standard, divides the consciousness state level, calculates the time proportion corresponding to each level, and obtains the consciousness state evaluation result.

[0041] Optionally, when calculating the proportion of mutation point types, the formula is used:

[0042]

[0043] According to the consciousness state evaluation standard, determine the corresponding relationship between the mutation mode and the consciousness state, divide the consciousness state level, calculate the time proportion corresponding to each level, and obtain the consciousness state evaluation result;

[0044] where, P p represents the proportion of the p-th type of mutation point, C p represents the number of the p-th type of mutation point, U q,p represents the cumulative duration of the p-th type of mutation point within the q-th time segment, N q represents the total duration of the q-th time segment, represents the average duration of the p-th type of mutation point within all time segments, N q,p represents the average duration of the p-th type of mutation point within the q-th time segment, κ represents the adjustment term weight coefficient, p represents the mutation point type index, q represents the time segment index, r represents the number of time segments, and s represents the number of mutation point types.

[0045] The beneficial effects brought by the technical solution provided by the embodiments of the present invention at least include:

[0046] The innovative solution carefully captures the subtle movements of the muscles around the eyes, corners of the mouth, and forehead through an infrared camera, improving the accuracy and comprehensiveness of data collection. By evaluating the blink frequency, displacement of the corners of the mouth, and stretching amplitude of the frontal muscles, the precision and dimension of the assessment are enhanced. Using the extraction of deformation features and the calculation of the motion amplitude gradient, combined with time series analysis, the assessment of the state of consciousness becomes more sensitive and accurate, capable of promptly detecting potential changes in consciousness, thus enabling a rapid response. The analysis of the density and distribution characteristics of mutation points provides rich data interpretation for diagnosis and monitoring, enabling the rapid and accurate assessment of the state of consciousness of the injured person without physical contact, which is particularly suitable for first aid scenes and battlefield environments, providing immediate data to support medical decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0048] Figure 1 System schematic diagram of the present invention;

[0049] Figure 2 System framework schematic diagram of the present invention;

[0050] Figure 3 Flowchart of the facial information extraction module of the present invention;

[0051] Figure 4 Flowchart of the facial feature analysis module of the present invention;

[0052] Figure 5 Flowchart of the state of consciousness calculation module of the present invention;

[0053] Figure 6 Flowchart of the anomaly detection module of the present invention;

[0054] Figure 7 Flowchart of the state of consciousness assessment module of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] The following will describe the technical solutions in the present invention in conjunction with the drawings.

[0056] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either of the two can be selected.

[0057] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meaning they express is the same. "(of)", "corresponding" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meaning they express is the same.

[0058] In the embodiments of the present invention, sometimes subscripts such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0059] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0060] The embodiments of the present invention provide a wounded awareness state assessment system based on face recognition, as Figure 1 shown, the system includes:

[0061] The facial information extraction module uses an infrared camera to collect the motion change data of the eye area, the corners of the mouth and the frontalis muscle, analyzes the deformation degree of the facial motion in adjacent time frames, calculates the blink frequency, the displacement amount of the corners of the mouth and the stretching amplitude of the frontal muscle, and obtains the key sequence of facial motion;

[0062] The facial feature analysis module extracts the deformation features of the key facial areas based on the key sequence of facial motion, calculates the motion amplitude gradients of the eye area, the corners of the mouth and the frontalis muscle areas, analyzes the gradient differences in continuous time frames, and tracks the facial feature points in the local area to obtain the facial dynamic features;

[0063] The awareness state calculation module analyzes the motion trends of the eye area, the corners of the mouth and the frontalis muscle based on the facial dynamic features, calculates the change rate of the facial motion within the differential time window, extracts the mutation points of the change rate, counts the distribution of the mutation points on the time axis, determines the influence degree of the injured area on the facial mutation points, and assigns the error weights of the injured area to obtain the analysis result of the motion trend deviation;

[0064] Based on the analysis result of the motion trend deviation, the anomaly detection module calculates the density of mutation points on the time axis, compares the distribution characteristics of mutation points in adjacent time periods, determines whether the number of mutation points in adjacent time periods exceeds the average value of mutation points, screens out abnormal mutation points, marks the time nodes, classifies the anomalies, and generates mutation point marking data;

[0065] Based on the mutation point marking data, the consciousness state evaluation module calculates the duration, distribution density, and time stability of the mutation points, analyzes the variation law of the mutation points on the time axis, determines the type and pattern of abnormal mutations, divides the consciousness state levels, and obtains the consciousness state evaluation result.

[0066] The key sequences of facial movements include eyelid movement sequences, mouth corner movement sequences, and forehead expression sequences. The dynamic facial features include deformation rate, deformation period, and deformation amplitude. The analysis results of motion trend deviation include mutation frequency, mutation persistence, and mutation intensity. The mutation point marking data includes abnormal timestamps, abnormal durations, and abnormal intensity levels. The consciousness state evaluation result includes consciousness state stability, consciousness state change frequency, and consciousness state classification level.

[0067] As Figure 2 and Figure 3 shown, the facial information extraction module includes;

[0068] The facial motion data acquisition sub-module uses an infrared camera to collect the motion change data of the eye perimeter, mouth corners, and frontal muscles, extracts the position information of the facial monitoring points in adjacent time frames, calculates the contraction amplitude of the eye perimeter, the movement trajectory of the mouth corners, and the stretching length of the frontal muscles, calibrates the range of the motion interval, and generates the facial motion trajectory;

[0069] Set the acquisition frame rate and acquisition time period, collect the facial continuous time frame image sequence through the device, extract the position information of the facial monitoring points in adjacent time frames. The facial monitoring points are the center points of the upper and lower eyelids of the eye perimeter, the left and right edge points of the mouth corners, and the reference points in the middle area of the forehead. Convert the pixel coordinates to the actual measurement coordinates, and use the Euclidean distance formula to calculate the displacement change amount between the motion points. The formula is where (x1,y1) and (x2,y2) respectively represent the position coordinates of the monitoring points in adjacent time frames. Assume that the position of the left upper eyelid in time frame 1 is (120,200), and the position in time frame 2 is (118,198), then the calculated displacement is The pixels, after the displacement is converted into the actual length, are used for subsequent analysis of the movement amplitude. The calculation of the contraction amplitude around the eyes uses the maximum and minimum displacement differences of consecutive time frames. Suppose the maximum displacement within 10 consecutive frames is 4.2 mm and the minimum displacement is 1.1 mm, then the contraction amplitude is 3.1 mm. The movement trajectory of the corners of the mouth is obtained by tracking the displacement curve of the corner points of the mouth to calculate the moving path length, which is calculated by cumulative displacement. For example, if the displacement of the corner point of the mouth within consecutive time frames is 1.0 mm, 1.2 mm, and 1.5 mm respectively, then the trajectory length is 1.0 + 1.2 + 1.5 = 3.7 mm. The stretching length of the frontalis muscle is calculated by monitoring the displacement change of the forehead reference point in the vertical direction. Suppose the position of the forehead reference point in the initial frame is (150, 100), and the position in the fifth frame is (150, 105), then the stretching length is 5 mm. The range of the calibrated movement interval is set by displacement threshold. The threshold is determined with reference to the mean and standard deviation of the statistical sample. Suppose the sample mean is 2.5 mm and the standard deviation is 0.8 mm. The interval where the displacement is greater than the mean plus 1.5 times the standard deviation is defined as the effective movement interval, that is, the threshold is 2.5 + 1.5×0.8 = 3.7 mm. According to the data of all monitoring points, a set of displacement curves of facial monitoring points is generated to obtain the facial movement trajectory.

[0070] Based on the facial movement trajectory, the facial deformation analysis sub-module calculates the deformation gradient of the facial monitoring points within adjacent time frames, screens the movement points that meet the deformation criteria, analyzes the blinking closure cycle, the offset of the corners of the mouth, and the stretching degree of the frontalis muscle within consecutive time frames, and obtains the facial deformation characteristics;

[0071] Call the data of the displacement curves of the monitoring points, calculate the deformation gradient of the facial monitoring points within adjacent time frames, and use the gradient formula where Δd represents the displacement difference and Δt is the time difference. Suppose the time interval is 0.04 seconds and the displacement difference between adjacent frames is 1.6 mm, then the gradient Screen the movement points that meet the deformation criteria. The deformation criteria are set based on the mean and standard deviation of the sample. Suppose the mean is 25 mm / s and the standard deviation is 10 mm / s. The threshold is set as the mean plus 2 times the standard deviation, that is, 25 + 2×10 = 45 mm / s. The points with a deformation gradient greater than 45 mm / s are regarded as effective movement points. When analyzing the blinking closure cycle within consecutive time frames, by tracking the change in the distance between the upper and lower eyelids. Suppose the distance sequence within consecutive frames is (5.2, 4.1, 2.0, 1.0, 0.8, 1.2, 3.0, 5.0) mm. The closure cycle is determined as the time required from the minimum distance to the maximum distance. If it takes 0.16 seconds from closure to opening, the closure cycle is 0.16 seconds. The offset of the corners of the mouth is calculated by calculating the displacement changes in the horizontal and vertical directions of the corner points of the mouth. Suppose the cumulative displacement in the horizontal direction is 3.5 mm and in the vertical direction is 2.8 mm, the total offset is The degree of frontalis muscle stretching is determined by the difference between the maximum and minimum values in the displacement curve of consecutive time frames. Assuming that the stretching displacement difference within consecutive time frames is 6.2 mm, the stretching degree is 6.2 mm. All valid motion points and related deformation data are sorted out to generate facial deformation features.

[0072] The key sequence screening sub-module calls the facial deformation features, screens the facial motion sequences, judges the continuity of the motion sequences, screens stable motion points according to the time axis, calculates the facial motion pattern matching degree, determines the time period with stable motion patterns, and obtains the key facial motion sequences.

[0073] Set the sequence length threshold and select the data segment that meets the requirement of the number of consecutive frames. Assuming that the length of the valid consecutive frame segment in the sequence needs to exceed 8 frames, the continuity of the motion sequence is calculated through the inter-frame interval. If the inter-frame time difference is 0.04 seconds and there is no frame skipping, it is regarded as a continuous sequence. When screening stable motion points according to the time axis, the motion stability judgment index is used where σ d is the standard deviation of displacement, μ d is the mean value of displacement. Assuming that the mean value of sequence displacement is 3.0 mm and the standard deviation is 0.5 mm, then the stability index Set the threshold to 0.2. S<0.2 is judged as a stable motion point. The dynamic time warping algorithm (DTW) is used to calculate the facial motion pattern matching degree. The reference motion template and test sequence data are input. Assuming that the result after DTW distance calculation is 12.5 and the matching threshold is set to 15, when the DTW distance is less than 15, it is judged as a highly matching sequence. When determining the stable time period of the motion pattern, it is judged by the proportion of stable motion points within the time period. Assuming that 9 out of 10 frames within the time period are stable points, accounting for 90%, exceeding the set threshold of 80%, select the time period that meets the requirements of stability and matching degree to obtain the key facial motion sequences.

[0074] As Figure 2 and Figure 4 shown, the facial feature analysis module includes;

[0075] The deformation feature extraction sub-module analyzes the motion trajectories of the areas around the eyes, the corners of the mouth, and the frontalis muscle based on the key facial motion sequences, identifies and marks the key feature points, obtains the position changes of each feature point in consecutive time frames, and generates the key feature motion data;

[0076] The deformation feature extraction sub-module analyzes the motion trajectories of the areas around the eyes, the corners of the mouth, and the frontalis muscle based on the key facial motion sequences, extracts the images of consecutive time frames in the target areas, locates the coordinates of the feature points through the facial feature point detection method, and sets the feature points around the eyes as The feature points at the corners of the mouth are The feature points of the frontalis muscle are Where x is the feature point index, u is the time frame number, and the feature point displacement is calculated using the following formula: Where, and are the horizontal and vertical coordinates of the x-th feature point in the u-th frame respectively. During the monitoring process, assuming that the coordinates of the corner of the mouth feature point in the 8th frame are (112.4 mm, 198.6 mm) and those in the 7th frame are (111.1 mm, 197.9 mm), substitute them into the calculation:

[0077] Among all the feature points, set the displacement validity threshold as the weighted standard deviation of the sample displacement mean, that is: Where the sample displacement mean The standard deviation σ D = 0.4 mm, the weight coefficient β = 1.5, calculate to get T d = 1.2 + 1.5×0.4 = 1.8 mm. If is less than the threshold, it is not marked as significant movement. The displacements of the feature points in the eye area, corner of the mouth and frontalis muscle area are calculated in sequence to generate a feature matrix: Generate key feature motion data.

[0078] The gradient calculation sub-module calls the key feature motion data, calculates the motion amplitude gradient values of the eye area, corner of the mouth and frontalis muscle area, compares the positions of the motion points in consecutive time frames, analyzes the gradient change rate between adjacent time frames, filters the gradient data exceeding the motion amplitude standard, and obtains the motion gradient feature values;

[0079] Calculate the motion amplitude gradients of the eye area, corner of the mouth and frontalis muscle area using the formula:

[0080]

[0081] Compare the positions of the motion points in consecutive time frames, analyze the gradient change rate between adjacent time frames, filter the gradient data exceeding the motion amplitude standard, and obtain the motion gradient feature values;

[0082] Where, G ij represents the motion amplitude gradient value of the j-th feature point in the i-th area between adjacent time frames, represents the displacement of the feature point in the i-th area in the k-th frame, represents the displacement of the j-th feature point in the adjacent frame, w ij represents the weight coefficient of the j-th feature point in the i-th area, Δt k represents the time interval between the k-th frame and the k - 1-th frame, α ij represents the deformation correction parameter of the j-th feature point in the i-th area;

[0083] Assume that in the k-th frame, the feature points in the i-th region are located at the coordinates while in the (k - 1)-th frame, the corresponding j-th feature point is located at By calculating the Euclidean distance between these two feature points, the displacement difference is obtained:

[0084]

[0085] Assume that the following data is measured by a high-precision imaging device:

[0086] In the k-th frame, the coordinates of the feature points in the i-th region:

[0087] In the (k - 1)-th frame, the coordinates of the j-th feature point:

[0088] Substitute into the calculation:

[0089]

[0090] Determine the weight coefficient w of the feature point ij , for example, the weight of the eye area is set to 1.2, the mouth corner area is 1.0, and the frontalis area is 0.8. Assume that the weight coefficient w of the i-th region ij = 1.0.

[0091] The time interval Δt k is the time difference between two frames. For example, if the frame rate of the imaging device is 30 frames per second, then:

[0092]

[0093] The deformation correction parameter α ij is used to adjust the deformation influence of different feature points in motion. For example, by measuring the displacement changes of feature points under different expressions and calculating their standard deviations as α ij value. Assume that for the feature points in the i-th region, α ij = 0.5mm.

[0094] Substitute the above values into the original formula:

[0095]

[0096] Calculate the denominator:

[0097]

[0098] Therefore, the motion amplitude gradient G ij is:

[0099]

[0100] The results show that in adjacent time frames, the feature points in the i-th region move at a speed of approximately 5.59 millimeters per second. By calculating the G ij value of all feature points, the motion gradient eigenvalue of the entire face can be obtained.

[0101] Based on the motion gradient eigenvalue, the trend analysis sub-module analyzes the change trend of facial motion, compares the facial dynamics in different time periods, tracks the motion state of feature points in local regions, judges the stability and change pattern of facial dynamics, and obtains the facial dynamic feature trend.

[0102] Based on the motion gradient eigenvalue, the trend analysis sub-module analyzes the change trend of facial motion and extracts an effective gradient sequence from the input data where p is the feature point number and q is the time window number. The average motion gradient is calculated within each time window: If the gradient values of M = 4 feature points in the 3rd time window are {5.2, 6.1, 5.8, 5.6} mm / s, calculate the average value:

[0103] The dynamic stability is calculated by the fluctuation coefficient: where is the standard deviation of the gradient within the window,

[0104] Calculate:

[0105] The fluctuation coefficient is: The difference rate is used to judge the trend change: If the average value in the previous time window Substitute into the calculation: If the preset threshold T Δ = 0.07, then It is determined that the trend is stable. Organize the dynamic parameters of all windows to generate a trend matrix: Obtain the facial dynamic feature trend.

[0106] As Figure 2 and Figure 5 shown, the consciousness state calculation module includes;

[0107] Based on the facial dynamic feature trend, the change rate calculation sub-module analyzes the change rate of the facial motion amplitude, calculates the difference value of the facial motion speed in adjacent time windows, screens the time intervals where the change rate exceeds the set standard, and generates the motion amplitude offset feature;

[0108] Extract the motion rate data within consecutive time windows from the trends of facial dynamic features. Assume the total monitoring time is 10 seconds, the time window size is 2 seconds, and there are a total of 5 window segments. Use a high - frame - rate camera device for data acquisition, set the frame rate to 60 frames per second, and the interval time for each frame is 0.0167 seconds. During the detection period, collect the motion data of the areas around the eyes, the corners of the mouth, and the frontalis muscle, calculate the average rate of the facial monitoring points within each time window, and set the rate means of the 1st - 5th time windows as To calculate the rate of change of the facial motion rate within adjacent time windows, use the following formula: Where, represents the rate of change of the motion rate within the v - th time window, and are the average motion rates of adjacent time windows, is the weight coefficient of the v - th time window, T δ is the time window length, is the time correction term. Assume the weight coefficient of the 4th time window Time correction term Time window length T δ = 2.0 seconds, substitute the rate means of the 4th and 3rd time windows into the calculation:

[0109]

[0110] After completing the calculation of the rate of change, it is necessary to set a rate - of - change threshold for anomaly screening. Based on historical monitoring data and experimental samples, set the rate - of - change threshold to 0.4 mm / s2. When , mark this time window as an abnormal time period. In this calculation exceeds the threshold, determine that the 4th time window is an abnormal window, conduct abnormal time - window screening, count all the window sequences that exceed the threshold, and form a set of abnormal time intervals; compare the selected abnormal time intervals with the rate trend within the continuous time period to confirm the distribution characteristics and change patterns of the continuous abnormal windows, and generate the motion amplitude offset characteristics.

[0111] The mutation - point extraction sub - module calls the motion amplitude offset characteristics, compares the rate - of - change gradients of adjacent time windows, identifies the rate - of - change mutation points, screens the time points where the motion rate deviates from the standard, extracts the distribution of the mutation points on the time axis, analyzes the continuity and change trend of the mutation points, and obtains the mutation - point characteristic map;

[0112] The mutation - point extraction sub - module calls the motion amplitude offset characteristics, compares the rate - of - change gradients within adjacent time windows, identifies the mutation points, and uses the gradient - difference rate formula: Where, is the rate - of - change gradient value of the m - th time window, λ m is the weight - correction coefficient, τm is the window time span, θ m is the mutation correction term, and the change rates of the 6th and 7th time windows are extracted The time span τ7 = 0.06 s, the correction term θ7 = 0.03 s, and the weight coefficient λ7 = 1.05 are substituted into the calculation:

[0113]

[0114] Set the mutation point threshold T G = 25.0 mm / s2, because exceeds the threshold and is marked as a mutation point. The time index set {t3, t7, t 10} of the mutation points is screened. The continuity calculation uses the time interval difference rate: If t7 = 3.4 s, t 10 = 4.2 s, and the reference time ΔT = 1.0 s is set, then: The mutation point characteristic map is obtained.

[0115] Based on the mutation point characteristic map, the error weight allocation sub-module analyzes the correspondence between the time position of the mutation point and the injured area, judges the influence degree of the mutation point on the facial movement mode, adjusts the error allocation ratio of the injured area, determines the influence range of the mutation point, and obtains the analysis result of the motion trend deviation.

[0116] Based on the mutation point characteristic map, the error weight allocation sub-module analyzes the correspondence between the mutation point time position and the injured area, and sets the mutation point position and the injured area number Calculate the influence coefficient of the mutation point on the injured area: Among them, is the influence value of the kth mutation point on the nth area, is the mutation intensity, ω n is the area weight, β k is the time offset correction term, ψ n is the area correction coefficient, which is obtained through data measurement ω2 = 1.2, β3 = 0.02 ms, ψ2 = 0.05 ms are substituted into the calculation:

[0117]

[0118] If the influence threshold T I = 50.0, it is marked as a significantly affected area, and the adjustment ratio of the error allocation of the injured area is calculated: If the cumulative influence value of the 2nd area The number of mutation points K = 3, then: Obtain the analysis result of the motion trend deviation.

[0119] As Figure 2 and Figure 6 shown, the anomaly detection module includes:

[0120] The mutation point density calculation sub-module extracts the mutation point data and the number of mutation points within the time window based on the analysis result of the motion trend deviation, calculates the average density of the mutation points within the time period, and obtains the mutation point density distribution.

[0121] Set the time window size W α and the total time T α , calculate the number of time windows Statistically count the number of mutation points in each time window The formula for calculating the mutation point density is: Where represents the mutation point density in the a-th time window, is the window correction coefficient, is the adjustment term, set the time window size W α = 2.0 seconds, the total time T α = 10.0 seconds, calculate the number of windows N α = 5, the number of mutation points in the 3rd time window in actual monitoring Window correction coefficient Adjustment term Substitute the values into the formula: By calculating the mutation point density of all time windows, the distribution of the mutation point density in each time period can be obtained, and the mutation point density distribution can be obtained.

[0122] The anomaly screening sub-module calls the mutation point density distribution, compares the mutation point distribution characteristics in adjacent time periods, calculates the change amplitude of the number of mutation points, determines whether the number of mutation points in adjacent time periods exceeds the average density, screens out the abnormal mutation points and marks the time nodes to obtain the abnormal distribution nodes.

[0123] Extract the data of the number of mutation points within the time period, set the time window size and the total time period length, compare and analyze according to the number of mutation points and the average density of the mutation points in each time period in the input data. Set the total monitoring time to 12 seconds and the time window size to 2 seconds, divide the monitoring data into 6 time periods, and measure the number of mutation points in each time period to be respectively. The formula for calculating the change amplitude of the number of mutation points is as follows: Where is the change amplitude of the number of mutation points in the q-th time period, and are the number of mutation points in the current and previous time periods respectively, is the time period weight coefficient, is the average value of the mutation point density, is the density correction term. The number of mutation points in the 4th time period and the 3rd time period are respectively Let the weight coefficient average value of the mutation point density density correction term Substitute into the formula for calculation:

[0124]

[0125] According to the calculation result of the change amplitude of the number of mutation points, set the threshold T of the change amplitude of the number of mutation points τ = 1.2 mpoints. When it is determined that the fluctuation of the number of mutation points is significant. Here Therefore, the change in the number of mutation points in the 4th time period is determined to be an abnormal mutation interval;

[0126] Judge whether the number of mutation points in adjacent time periods exceeds the average density. Let the reference value of the mutation point density be the average value of the number of mutation points in all time periods:

[0127]

[0128] If the number of mutation points in a certain time period exceeds then it is marked as a high-density section. The number of mutation points in the 4th time period is marked as the time period of abnormal mutation points. Combining all the detection data, the abnormal mutation points are screened out and the corresponding time nodes t4 = 6.0 s and t5 = 8.0 s are marked to obtain the abnormal distribution nodes.

[0129] Based on the abnormal distribution nodes, the abnormal classification sub-module analyzes the time distribution law of the abnormal mutation points, calculates the weight frequency ratio of the abnormal mutation points, classifies the abnormal types of the mutation points, judges the change of the abnormal types in each time period, and generates the mutation point marking data.

[0130] Calculate the change density ratio of the abnormal mutation points, using the formula:

[0131]

[0132] Classify the abnormal types of the mutation points, judge the change of the abnormal types in each time period, and generate the mutation point marking data;

[0133] Among them, D r represents the change density ratio of the abnormal mutation points, Q e represents the number of abnormal mutation points in the e-th time interval, Q e-1represents the number of abnormal mutation points in the previous time interval, f represents the number of time intervals, Z e represents the end time of the e-th time interval, Z e-1 represents the end time of the previous time interval;

[0134] The gait data of the user is monitored by a laser sensor and an inertial measurement unit. The total number of mutation points is counted with every 30 seconds as a time interval. An abnormal mutation point is defined as a time node where the step length change exceeds 15%. The step length change data is recorded and analyzed in real time by the inertial measurement unit. The sampling frequency of the sensor is 100Hz, and the number of abnormal mutation points is Q e In actual detection, they are respectively recorded as:

[0135] Q0 = 4 points, Q1 = 6 points, Q2 = 7 points, Q3 = 5 points, Q4 = 8 points, Q5 = 6 points;

[0136] The parameter f represents the number of time intervals. The data acquisition source is the preset detection period division. The total monitoring time this time is 180 seconds, 30 seconds for each interval, and there are a total of 6 time intervals. Therefore, f = 6;

[0137] The parameter Z e represents the end time of the e-th time interval. Based on the data acquisition time setting, the end time points of the time intervals are:

[0138] Z0 = 30 seconds, Z1 = 60 seconds, Z2 = 90 seconds, Z3 = 120 seconds, Z4 = 150 seconds, Z5 = 180 seconds;

[0139] The calculation steps are as follows:

[0140] Calculation of the sum of step length differences:

[0141]

[0142] Calculation of the square root of the sum of squares of time intervals:

[0143]

[0144] Formula calculation:

[0145]

[0146] The threshold for abnormal mutation points is that the step length change exceeds 15%. According to the literature on gait monitoring, the normal walking step length fluctuation range is between 5% and 10%. Exceeding 15% can be judged as abnormal. The sampling frequency of the sensor is selected as 100Hz to meet the real-time monitoring accuracy requirements, and the time interval length is 30 seconds to cover the natural gait cycle without overly extending the detection time;

[0147] The result shows that the change density ratio of abnormal mutation points within the detection period is 0.0249. This value is used to quantify the fluctuation degree of the number of mutation points. The higher the value, the more drastic the change in the number of mutation points. This index value is the input data for subsequent abnormal classification and the analysis of the change of abnormal types over time periods, which helps to distinguish the abnormal types of mutation points and generate mutation point marking data.

[0148] As Figure 2 and Figure 7 shown, the consciousness state evaluation module includes;

[0149] Based on the mutation point marking data, the mutation point feature calculation sub-module extracts the start time of the mutation point, calculates the interval time and duration of the mutation point, analyzes the time stability of the mutation point within different time windows, and obtains the mutation time stability index;

[0150] Based on the mutation point marking data, the mutation point feature calculation sub-module extracts the start time of the mutation point Set the total number of mutation points as N δ , and calculate the interval time and duration of the mutation point. The formula for the interval time of the mutation point is: Where, is the interval time between the d-th mutation point and the previous mutation point, is the time correction coefficient, is the data sampling interval, is the time offset correction term. Extracting the data, the time of the 3rd mutation point is the time of the 2nd mutation point time correction coefficient sampling interval offset correction term Substitute into the formula:

[0151]

[0152] The duration of the mutation point The calculation formula is: Where, correction term Substitute into the calculation: The time stability of the mutation point Adopt the coefficient of variation method: Let seconds, substitute into the calculation: Obtain the mutation time stability index.

[0153] The abnormal mode analysis sub-module calls the mutation time stability index, analyzes the variation law of mutation points on the time axis, compares the distribution of mutation points in different time periods, identifies the fluctuation trend of the number of mutation points, determines the type and law of abnormal mutations, and obtains the abnormal change identification data;

[0154] The formula for the fluctuation trend is: Where, is the mutation point fluctuation trend value in the e-th time period, is the trend correction coefficient, is the time period span, is the correction parameter. The stability indexes of the 4th and 3rd time periods are Let Time period span Correction parameter Substitute into the calculation:

[0155]

[0156] Judgment criteria for abnormal mutation types: is low fluctuation, is medium fluctuation, is high fluctuation. The current calculation belongs to the medium fluctuation type. Obtain the abnormal change identification data.

[0157] The state level division sub-module calculates the proportion of mutation point types based on the abnormal change identification data, determines the corresponding relationship between the mutation mode and the consciousness state according to the consciousness state evaluation standard, divides the consciousness state level, calculates the time proportion corresponding to each level, and obtains the consciousness state evaluation result.

[0158] Calculate the proportion of mutation point types using the formula:

[0159]

[0160] According to the consciousness state evaluation standard, determine the corresponding relationship between the mutation mode and the consciousness state, divide the consciousness state level, calculate the time proportion corresponding to each level, and obtain the consciousness state evaluation result;

[0161] Where, P p represents the proportion of the p-th type of mutation point, C p represents the number of the p-th type of mutation point, U q,p represents the cumulative duration of the p-th type of mutation point in the q-th time segment, N q represents the total duration of the q-th time segment, represents the average duration of the p-th type of mutation point in all time segments, N q,prepresents the average duration of the p-th type of mutation point within the q-th time segment, κ represents the weight coefficient of the adjustment term, p represents the mutation point type index, q represents the time segment index, r represents the number of time segments, and s represents the number of mutation point types;

[0162] C p represents the number of the p-th type of mutation point, which is obtained by classifying and counting the mutation points detected within the time segment. For example, within a 10-minute monitoring period, by counting the facial expression data, the number of the 1st type of mutation point is identified as 5;

[0163] U q,p represents the cumulative duration of the p-th type of mutation point within the q-th time segment. This value is obtained by collecting data through video or sensors and accumulating the recorded durations of the mutation points. For example, within the first time segment, the duration of the 1st type of mutation point is 2.5 seconds, and within the second time segment, it is 1.8 seconds;

[0164] N q represents the total duration of the q-th time segment. This value is determined by the time segment division criterion. For example, if each time segment is set to 10 seconds, then N q = 10 seconds;

[0165] represents the average duration of the p-th type of mutation point within all time segments, and the calculation method is:

[0166]

[0167] where N q,p represents the average duration of the p-th type of mutation point within the q-th time segment, and r is the number of time segments. For example, within 3 time segments, the average durations of the 1st type of mutation point are 1.5 seconds, 2.0 seconds, and 1.8 seconds respectively, then the calculation result is:

[0168]

[0169] κ represents the weight coefficient of the adjustment term, which is used to balance the influence of different types of mutation points during calculation, determined through experiments, and varies with data fluctuations. For example, κ = 0.8 is set;

[0170] Calculation process:

[0171] In the first step, calculate the square of the proportion of the duration of the p-th type of mutation point within each time segment and sum them up:

[0172]

[0173] Suppose that within 3 time segments, the cumulative durations of the first type of mutation points are 2.5 seconds, 1.8 seconds, and 3.2 seconds respectively, and the length of the time segment is 10 seconds, then the calculation is as follows:

[0174]

[0175] Take the square root:

[0176]

[0177] Second step, calculate the numerator part:

[0178] |C p |×0.444;

[0179] Let C p =5:

[0180] 5×0.444 = 2.22;

[0181] Third step, calculate the first part of the denominator:

[0182]

[0183] Suppose there are 3 types of mutation points in total, and the quantities of each type are 5, 3, and 4 respectively, then:

[0184] 5 + 3 + 4 = 12;

[0185] Fourth step, calculate the second part of the denominator:

[0186] Calculate the square root of the sum of the absolute values of the differences between the duration of the p-th type of mutation point within the time segment and the overall average value:

[0187]

[0188] Calculate with the above data:

[0189] |1.5 - 1.77| = 0.27, |2.0 - 1.77| = 0.23, |1.8 - 1.77| = 0.03;

[0190] ∑ = 0.27 + 0.23 + 0.03 = 0.53;

[0191]

[0192] Multiply by the adjustment term weight coefficient κ = 0.8:

[0193] 0.8×0.728 = 0.582;

[0194] Fifth step, calculate the denominator:

[0195] 12 + 0.582 = 12.582;

[0196] Finally, calculate the proportion of mutation point types:

[0197]

[0198] This result indicates that the proportion of the p - type mutation points among all mutation point types is 17.65%. This value can be used to analyze the distribution of mutation points and further conduct a corresponding analysis with the state of consciousness to identify the influence degree of the mutation pattern.

[0199] It should be understood that the term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. Additionally, the character " / " in this article generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context before and after.

[0200] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one (item)" or its similar expression refers to any combination of these items, including any combination of single - item (item) or plural - item (items). For example, at least one (item) of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0201] It should be understood that in various embodiments of the present invention, the magnitudes of the serial numbers of the above - mentioned processes do not imply the sequence of execution. The execution sequence of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0202] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present invention.

[0203] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the devices, systems, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0204] In several embodiments provided by the present invention, it should be understood that the disclosed devices, systems, and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the couplings, direct couplings, or communication connections shown or discussed with each other can be through some interfaces. The indirect couplings or communication connections of systems or units can be in electrical, mechanical, or other forms.

[0205] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0206] In addition, in each embodiment of the present invention, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0207] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, etc., which can store program codes.

[0208] As described above, the above are only specific implementation manners of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A system for evaluating the consciousness state of wounded personnel based on facial recognition, characterized in that, The system includes: The facial information extraction module uses an infrared camera to collect the motion data of the eye area, the corners of the mouth, and the frontalis muscle, calculates the blink frequency, the displacement of the corners of the mouth, and the stretching amplitude of the frontal muscles, and obtains the key sequence of facial motion; The facial feature analysis module extracts deformation features based on the key sequence of facial motion, calculates the motion amplitude gradient of the facial area, analyzes the gradient difference, identifies facial feature points, and obtains the dynamic facial features; The consciousness state calculation module analyzes the facial motion trend based on the dynamic facial features, calculates the motion change rate, identifies and counts the distribution of change rate mutation points, determines the influence of the injured area on the mutation points, and obtains the analysis result of the motion trend deviation; The anomaly detection module measures the mutation point density based on the analysis result of the motion trend deviation, compares the distribution characteristics of the mutation points, determines whether the number of mutation points exceeds the average value of the mutation points, marks the abnormal mutation points and classifies them, and generates the mutation point marking data; The consciousness state evaluation module calculates the duration, distribution density, and time stability of the mutation points based on the mutation point marking data, determines the type and pattern of the abnormal mutation, divides the consciousness level, and obtains the consciousness state evaluation result.

2. The system for evaluating the consciousness state of the wounded based on face recognition according to claim 1, characterized in that The key sequence of facial motion includes the eyelid movement sequence, the corner-of-mouth movement sequence, and the frontal expression sequence. The dynamic facial features include the deformation rate, the deformation period, and the deformation amplitude. The analysis result of the motion trend deviation includes the mutation frequency, the mutation persistence, and the mutation intensity. The mutation point marking data includes the abnormal timestamp, the abnormal duration, and the abnormal intensity level. The consciousness state evaluation result includes the consciousness state stability, the consciousness state change frequency, and the consciousness state classification level.

3. The face recognition-based casualty awareness status assessment system according to claim 1, characterized in that The facial information extraction module includes; The facial motion data acquisition sub-module uses an infrared camera to collect the motion change data of the eye area, the corners of the mouth, and the frontalis muscle, extracts the position information of the facial monitoring points in adjacent time frames, calculates the contraction amplitude of the eye area, the movement trajectory of the corners of the mouth, and the stretching length of the frontalis muscle, calibrates the range of the motion interval, and generates the facial motion trajectory; The facial deformation analysis sub-module calculates the deformation gradient of the facial monitoring points within adjacent time frames based on the facial motion trajectory, screens the motion points that meet the deformation criteria, analyzes the blink closing cycle, the offset of the corners of the mouth, and the stretching degree of the frontalis muscle within consecutive time frames, and obtains the facial deformation features; The key sequence screening sub-module calls the facial deformation features, screens the facial motion sequences, judges the continuity of the motion sequences, screens the stable motion points according to the time axis, calculates the facial motion pattern matching degree, determines the time period with stable motion patterns, and obtains the key sequence of facial motion.

4. The system for evaluating the consciousness state of the wounded based on face recognition according to claim 1, characterized in that, The facial feature analysis module includes; The deformation feature extraction sub-module analyzes the motion trajectories of the eye area, the corners of the mouth, and the frontalis muscle regions based on the key sequence of facial motion, identifies and marks the key feature points, obtains the position changes of each feature point in consecutive time frames, and generates the key feature motion data; The gradient calculation sub-module calls the key feature motion data, calculates the motion amplitude gradient values of the eye perimeter, mouth corners, and frontalis regions, compares the motion point positions of consecutive time frames, analyzes the gradient change rate between adjacent time frames, filters the gradient data exceeding the motion amplitude standard, and obtains the motion gradient feature values; The trend analysis sub-module analyzes the change trend of facial motion based on the motion gradient feature values, compares the facial dynamics of different time periods, tracks the motion states of local region feature points, judges the stability and change pattern of facial dynamics, and obtains the facial dynamic feature trend.

5. The system for evaluating the conscious state of the wounded based on facial recognition according to claim 4, wherein The calculation of the motion amplitude gradient of the eye perimeter, mouth corners, and frontalis regions uses the formula: Compare the motion point positions of consecutive time frames, analyze the gradient change rate between adjacent time frames, filter the gradient data exceeding the motion amplitude standard, and obtain the motion gradient feature values; Among them, G ij represents the motion amplitude gradient value of the j-th feature point in the i-th region between adjacent time frames, represents the displacement of the feature point in the k-th frame of the i-th region, represents the displacement of the j-th feature point in adjacent frames, w ij represents the weight coefficient of the j-th feature point in the i-th region, Δt k represents the time interval between the k-th frame and the k-1-th frame, α ij represents the deformation correction parameter of the j-th feature point in the i-th region.

6. The face recognition-based casualty awareness status assessment system according to claim 1, wherein The consciousness state calculation module includes; The change rate calculation sub-module analyzes the change rate of facial motion amplitude based on the facial dynamic feature trend, calculates the difference value of facial motion speed within adjacent time windows, filters the time intervals with a change rate exceeding the set standard, and generates the motion amplitude offset feature; The mutation point extraction sub-module calls the motion amplitude offset feature, compares the change rate gradients of adjacent time windows, identifies the change rate mutation points, filters the time points with a motion speed deviating from the standard, extracts the distribution of mutation points on the time axis, analyzes the continuity and change trend of the mutation points, and obtains the mutation point characteristic map; The error weight allocation sub-module analyzes the correspondence between the time positions of the mutation points and the injured regions based on the mutation point characteristic map, judges the influence degree of the mutation points on the facial motion pattern, adjusts the error allocation ratio of the injured regions, determines the influence range of the mutation points, and obtains the motion trend deviation analysis result.

7. The system for evaluating the conscious state of the wounded based on face recognition according to claim 1, characterized in that, The anomaly detection module includes; The mutation point density calculation sub-module extracts the mutation point data and the number of mutation points within a time window based on the motion trend deviation analysis result, calculates the average density of mutation points within a time period, and obtains the mutation point density distribution; The anomaly screening sub-module calls the mutation point density distribution, compares the mutation point distribution characteristics within adjacent time periods, calculates the change amplitude of the number of mutation points, judges whether the number of mutation points within adjacent time periods exceeds the average density, screens the abnormal mutation points and marks the time nodes, and obtains the abnormal distribution nodes; The anomaly classification sub-module analyzes the time distribution law of the abnormal mutation points based on the abnormal distribution nodes, calculates the weight frequency ratio of the abnormal mutation points, classifies the types of mutation point anomalies, judges the change situation of the anomaly types in each time period, and generates the mutation point marking data; 8. The face recognition-based casualty awareness status assessment system according to claim 7, characterized in that, The calculation of the change density ratio of the abnormal mutation points uses the formula: Classify the types of mutation point anomalies, judge the change situation of the anomaly types in each time period, and generate the mutation point marking data; Among them, D r represents the change density ratio of abnormal mutation points, Q e represents the number of abnormal mutation points in the e-th time interval, Q e-1 represents the number of abnormal mutation points in the previous time interval, f represents the number of time intervals, Z e represents the end time of the e-th time interval, Z e-1 represents the end time of the previous time interval.

9. The system for evaluating the consciousness state of the wounded based on facial recognition according to claim 1, characterized in that, The consciousness state evaluation module includes; The mutation point feature calculation sub-module extracts the start time of the mutation points based on the mutation point marking data, calculates the mutation point interval time and duration, analyzes the time stability of the mutation points within different time windows, and obtains the mutation time stability index; The abnormal mode analysis sub-module calls the mutation time stability index, analyzes the variation law of mutation points on the time axis, compares the distribution of mutation points in different time periods, identifies the fluctuation trend of the number of mutation points, judges the type and law of abnormal mutations, and obtains abnormal change identification data; The state level division sub-module calculates the proportion of mutation point types based on the abnormal change identification data, judges the corresponding relationship between the mutation mode and the consciousness state according to the consciousness state evaluation standard, divides the consciousness state level, calculates the time proportion corresponding to each level, and obtains the consciousness state evaluation result.

10. The system for evaluating the conscious state of the wounded based on facial recognition according to claim 9, characterized in that, The calculation of the proportion of mutation point types adopts the formula: According to the consciousness state evaluation standard, judge the corresponding relationship between the mutation mode and the consciousness state, divide the consciousness state level, calculate the time proportion corresponding to each level, and obtain the consciousness state evaluation result; Among them, P p represents the proportion of the p-th type of mutation point, C p represents the number of the p-th type of mutation point, U q,p represents the cumulative duration of the p-th type of mutation point within the q-th time segment, N q represents the total duration of the q-th time segment, represents the average duration of the p-th type of mutation point over all time segments, N q,p represents the average duration of the p-th type of mutation point within the q-th time segment, κ represents the adjustment term weight coefficient, p represents the mutation point type index, q represents the time segment index, r represents the number of time segments, and s represents the number of mutation point types.

Images