Headgear tightness test method and apparatus, computer readable storage medium

Abstract

The application provides a helmet wearing tightness test method and device, computer readable storage medium, comprising: obtaining effective subject data when the subject wears the helmet in different simulation scenarios; monitoring the pressure distribution of the contact surface between the helmet and the head of the subject, and generating a pressure distribution map; inputting the pressure distribution map and the effective subject data into a pattern recognition model to obtain a tightness pattern; in the case that the tightness pattern is not tight, automatically adjusting the helmet wearing parameters by using an adaptive adjustment mechanism; reacquiring the subject feeling data according to the automatically adjusted helmet wearing parameters and verifying the adjustment effect; in the case that the adjustment effect is invalid, reacquiring the data and adjusting the helmet until the helmet is tightly worn. The application solves the problem that the existing helmet test results deviate from the actual use, which leads to the difficulty in accurately evaluating the fitting degree of the helmet and the head.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of helmet tightness testing, in particular to a helmet wearing tightness testing method, a helmet wearing tightness testing device, a computer readable storage medium and a computer program product. BACKGROUND

[0002] With the development of technology and the improvement of safety production consciousness, helmets as personal protective equipment have been widely used in many fields. Traditional helmet design mainly relies on static size measurement and simple pressure test to evaluate the wearing tightness, however, this test method cannot fully reflect the influence of individual differences, motion state and environmental changes and other factors on the wearing tightness of the helmet in actual use. In recent years, the progress of sensor technology, electroencephalogram signal acquisition technology and deep learning algorithm provides new possibilities for the accurate test and optimization of helmet wearing tightness.

[0003] Although the existing technology has made certain progress in helmet design and testing, there are still many deficiencies. First, the existing helmet tightness testing method often ignores the diversity of actual use scenarios, such as climate conditions, motion states, etc., resulting in deviation between the test results and the actual use. Second, the traditional pressure test cannot provide comprehensive three-dimensional pressure distribution information, making it difficult to accurately evaluate the fit of the helmet and the head. In addition, the existing technology lacks real-time monitoring of the physiological and psychological state of the subjects, and cannot carry out individualized adaptation of the helmet from the perspective of the human body. SUMMARY

[0004] The main purpose of the present application is to provide a helmet wearing tightness testing method, a helmet wearing tightness testing device, a computer readable storage medium and a computer program product, to at least solve the problem that the existing technology cannot accurately evaluate the fit of the helmet and the head due to the deviation between the helmet test results and the actual use.

[0005] In order to achieve the above object, according to one aspect of the present application, a helmet wearing tightness test method is provided, comprising: an obtaining step of obtaining effective subject data of a subject wearing a helmet in different simulated scenarios, the effective subject data comprising effective eye movement tracking data, head micro-motion data and electroencephalogram after data processing, the eye movement tracking data comprising at least the line-of-sight movement trajectory, gaze point and blink frequency of the subject, and the head micro-motion data comprising at least the linear acceleration, speed and angular velocity data of the head of the subject; a generating step of generating a pressure distribution map by monitoring the pressure distribution of the contact surface between the helmet and the head of the subject through 3D pressure mapping technology; a pattern recognition step of inputting the pressure distribution map and the effective subject data into a pattern recognition model to determine the tightness of the helmet wearing, obtaining the tightness pattern of the helmet wearing of the subject, the tightness pattern being divided into helmet wearing tightness or helmet wearing not tightness; an adjusting step of, in the case of the tightness pattern being the helmet wearing not tightness, automatically adjusting the helmet wearing parameters according to the pressure distribution map and the effective subject data by using an adaptive adjustment mechanism, so as to adapt the helmet to the head of the subject; a verifying step of re-obtaining subject feeling data according to the automatically adjusted helmet wearing parameters, and verifying the adjustment effect according to the subject feeling data, the subject feeling data comprising the feeling data, physiological data, electroencephalogram data and pressure distribution data of the subject, the adjustment effect being adjustment invalid or adjustment valid, the adjustment invalid indicating that the helmet wearing is still not tight, and the adjustment valid indicating that the helmet wearing is tight; in the case of the adjustment effect being the adjustment invalid, sequentially repeating the obtaining step, the generating step, the pattern recognition step, the adjusting step and the verifying step at least once until the helmet wearing is tight.

[0006] Optionally, the effective subject data of the subject wearing the helmet under different simulation scenarios is acquired, comprising: starting the 4D environment simulator and randomly generating different simulation scenarios under the condition that the preparation condition is met, the preparation condition being that the subject cleans the head, wears the helmet according to the instruction and sits in the simulation environment to prepare for different scene tests; superimposing a virtual instruction of the helmet wearing state in the field of view of the subject in real time according to the simulation scenario through AR technology, so that the subject completes the corresponding action according to the virtual instruction and acquires the original eye movement tracking data and the original head micro-motion data in real time; performing a data cleaning operation on the original eye movement tracking data and the original head micro-motion data to obtain effective eye movement tracking data and effective head micro-motion data; monitoring the original electroencephalogram signal of the subject in real time by using a high-resolution electroencephalogram signal acquisition system combined with signal processing technology; performing a preprocessing operation on the original electroencephalogram signal to obtain an intermediate electroencephalogram signal, and extracting characteristic signals in the intermediate electroencephalogram signal by using Fourier transformation to obtain effective electroencephalogram signals, the preprocessing operation at least including a de-artifact operation.

[0007] Optionally, the preprocessing operation on the original electroencephalogram signal to obtain an intermediate electroencephalogram signal, and the extraction of characteristic signals in the intermediate electroencephalogram signal by using Fourier transformation to obtain effective electroencephalogram signals, comprises: performing the de-artifact operation on the original electroencephalogram signal according to a de-artifact formula to obtain the intermediate electroencephalogram signal, the de-artifact formula being wherein Artifact = |e in (t)-μ e |>ασ e , e in (t) is the original electroencephalogram signal at t time, μ e is the average value of the original electroencephalogram signal, σ e is the standard deviation of the original electroencephalogram signal, and a is a threshold coefficient, Artifact is a Boolean value, e clean (t) is the intermediate electroencephalogram signal after de-artifact, e in (t-1) is the original electroencephalogram signal value without artifact at t-1 time, e in (t+1) is the original electroencephalogram signal value without artifact at t+1 time; and the characteristic extraction on the intermediate electroencephalogram signal according to the Fourier transform formula to obtain the electroencephalogram signal, the Fourier transform formula being wherein F(ω) is the electroencephalogram signal obtained by Fourier transform, f(t) is a time domain signal, representing a function of the intermediate electroencephalogram signal, t is time, ω is angular frequency, and j is an imaginary unit.

[0008] Optionally, the pressure distribution of the contact surface between the inside of the helmet and the head of the subject is monitored by a 3D pressure mapping technology to generate a pressure distribution map, including: collecting pressure data on the contact surface between the helmet and the head by a 3D pressure mapping sensor fixed inside the helmet in real time; converting the pressure data into the pressure distribution map according to a pressure mapping formula, wherein the pressure mapping formula is wherein P(x, y) is a pressure value at coordinates (x, y), (x, y) is a coordinate position on the contact surface inside the helmet, w i is a weight factor of the i-th 3D pressure mapping sensor, p i is a pressure value measured by the i-th 3D pressure mapping sensor, i is a variable index, N is the total number of the 3D pressure mapping sensors, and the expression of the pressure value of each coordinate point in the pressure distribution map is: wherein p'(x, y) is an interpolated pressure value, (x1, y1) and (x2, y2) are coordinates of adjacent points with known pressure values, P(x1, y1), P(x2, y1), P(x1, y2) and P(x2, y2) are the known pressure values at the adjacent points.

[0009] Optionally, the pressure distribution map and the effective subject data are input into a pattern recognition model to obtain a tightness pattern of the helmet worn by the subject, including: inputting the pressure distribution map and the effective subject data into the pattern recognition model to extract spatial features of the pressure distribution map and the effective subject data by a convolution layer in the pattern recognition model to generate a feature map, performing dimension reduction pooling operation by a pooling layer in the pattern recognition model, converting the feature map into a one-dimensional feature vector by a full connection layer in the pattern recognition model, performing feature processing on the one-dimensional feature vector by an output layer in the pattern recognition model, and outputting the tightness pattern, wherein the formula for generating the feature map is y(n, i, j) is a value of the n-th output feature map at position (i, j), n, i, j are variable indexes, W(n, k, l) is a weight of the corresponding n-th convolution kernel at position (k, l), x(n, i+k, j+l) is a value of input data x at position (i+k, j+l), b(n) is a bias coefficient of the corresponding n-th convolution kernel, σ is an activation function, and x is input data representing the pressure distribution map and the effective subject data.

[0010] Optionally, automatically adjusting the helmet wearing parameters according to the pressure distribution map and the effective subject data by using an adaptive adjustment mechanism comprises: performing feature extraction on the pressure distribution map and the effective subject data to obtain corresponding feature data, wherein the feature data corresponding to the pressure distribution map at least includes a pressure peak value, a pressure average value and a pressure distribution uniformity, and the feature data corresponding to the effective subject data at least includes a gaze point position, a gaze duration, a blink frequency, a speed, an acceleration and a motion trajectory of head movement; performing feature fusion on all the feature data to obtain a feature vector; and adjusting the helmet wearing parameters according to an adaptive adjustment mechanism formula based on the feature vector, wherein the adaptive adjustment mechanism formula includes a pressure adjustment formula and a helmet position adjustment formula, the pressure adjustment formula is ΔP = K p ·(P desired -P current ), wherein ΔP is a pressure adjustment amount, K p is a proportional adjustment coefficient, P desired is an expected pressure distribution, P current is a currently monitored pressure distribution, and the helmet position adjustment formula is θ = θ0 + Δθ, wherein θ is a new position of the helmet, θ0 is an initial position of the helmet, and Δθ is an adjustment amount of the helmet.

[0011] Optionally, verifying the adjustment effect according to the subject feeling data comprises: bringing the feeling data in the subject feeling data into a first analysis formula to obtain a feeling data score average, wherein the first analysis formula is wherein is the feeling data score average, S i is the feeling data score of the i-th subject, i is a variable index, and Z is the total number of subjects; bringing the physiological data in the subject feeling data into a second analysis formula to obtain heart rate variability, wherein the second analysis formula is: wherein HRV is the heart rate variability, σ RR is a heart rate standard deviation, is a heart rate average value; and substituting the electroencephalogram data in the subject feeling data into a third analysis formula to obtain the power of α waves in the electroencephalogram, wherein the third analysis formula is: wherein P α is the power of the α waves, P(f) is a power spectral density of the electroencephalogram, f α,high and f α,low are respectively an upper limit and a lower limit of a frequency range of the α waves; and substituting the electroencephalogram data in the subject feeling data into a fourth analysis formula to obtain a pressure distribution uniformity index, wherein the fourth analysis formula is: wherein UI is the pressure distribution uniformity index, Pj a pressure value detected by a jth 3D pressure mapping sensor, j being a variable index, an average of the pressure values detected by all the 3D pressure mapping sensors, V being a total number of the 3D pressure mapping sensors; in a case where a closeness condition is satisfied, determining that the adjustment effect is valid, the closeness condition being that the average of the perception data scores is greater than a first standard value, the heart rate variability is greater than or equal to a second set threshold value and less than or equal to a third standard value, the power of the alpha wave is less than or equal to a fourth standard value, and the pressure distribution uniformity index is less than a fifth standard value; in a case where the closeness condition is not satisfied, determining that the adjustment effect is invalid.

[0012] According to another aspect of the present application, a helmet wearing closeness test device is provided, the device comprising: an acquisition unit configured to perform an acquisition step of acquiring valid subject data of a subject wearing a helmet in different simulated scenarios, the valid subject data comprising eye movement tracking data, head micro-motion data and electroencephalogram data after data processing, the eye movement tracking data comprising at least a visual line movement trajectory, a gaze point and a blink frequency of the subject, the head micro-motion data comprising at least head linear acceleration, speed and angular velocity data of the subject; a generation unit configured to perform a generation step of generating a pressure distribution map by monitoring a pressure distribution of a contact surface between the helmet and the head of the subject through a 3D pressure mapping technology; a pattern recognition unit configured to perform a pattern recognition step of inputting the pressure distribution map and the valid subject data into a pattern recognition model to determine a closeness of helmet wearing, obtaining a closeness pattern of helmet wearing of the subject, the closeness pattern being divided into helmet wearing being close or helmet wearing being not close; an adjustment unit configured to perform an adjustment step of automatically adjusting a helmet wearing parameter according to the pressure distribution map and the valid subject data by using an adaptive adjustment mechanism to adapt the helmet to the head of the subject in a case where the closeness pattern is helmet wearing being not close; a verification unit configured to perform a verification step of re-acquiring subject perception data according to the automatically adjusted helmet wearing parameter, and verifying an adjustment effect according to the subject perception data, the subject perception data comprising perception data, physiological data, electroencephalogram data and pressure distribution data of the subject, the adjustment effect being adjustment invalid or adjustment valid, the adjustment invalid indicating that the helmet wearing is still not close, and the adjustment valid indicating that the helmet wearing is close; a repeating unit configured to repeatedly perform the acquisition step, the generation step, the pattern recognition step, the adjustment step and the verification step at least once until the helmet wearing is close in a case where the adjustment effect is adjustment invalid.

[0013] According to still another aspect of the present application, a computer readable storage medium is provided, which comprises a stored program, wherein the computer readable storage medium controls a device in which the computer readable storage medium is located to perform any of the methods when the program is run.

[0014] According to still another aspect of the present application, a computer program product is provided, which comprises computer instructions, which implement any of the methods when executed by a processor.

[0015] According to the technical solution of the present application, in the helmet wearing tightness test method, first, an obtaining step is performed to obtain effective subject data of a subject wearing a helmet in different simulated scenarios, wherein the effective subject data comprises effective eye tracking data, head micro-motion data and electroencephalogram data after data processing, the eye tracking data at least comprises a line-of-sight movement trajectory, a gaze point and a blink frequency of the subject, and the head micro-motion data at least comprises head linear acceleration, speed and angular speed data of the subject; then, a generating step is performed to monitor pressure distribution of a contact surface between the helmet and the head of the subject by using 3D pressure mapping technology, and a pressure distribution map is generated; then, a pattern recognition step is performed to input the pressure distribution map and the effective subject data into a pattern recognition model to determine the tightness of the helmet wearing, and obtain a tightness pattern of the helmet wearing of the subject, wherein the tightness pattern is divided into helmet wearing tightness or helmet wearing not tight; then, an adjusting step is performed to automatically adjust helmet wearing parameters by using an adaptive adjustment mechanism according to the pressure distribution map and the effective subject data in the case that the tightness pattern is the helmet wearing not tight, so as to adapt the helmet to the head of the subject; then, a verifying step is performed to re-obtain subject feeling data according to the automatically adjusted helmet wearing parameters, and verify the adjustment effect according to the subject feeling data, wherein the subject feeling data comprises feeling data, physiological data, electroencephalogram data and pressure distribution data of the subject, the adjustment effect is adjustment invalid or adjustment valid, the adjustment invalid means that the helmet wearing is still not tight, and the adjustment valid means that the helmet wearing is tight; finally, in the case that the adjustment effect is the adjustment invalid, the obtaining step, the generating step, the pattern recognition step, the adjusting step and the verifying step are repeatedly executed at least once until the helmet wearing is tight. According to the present application, eye tracking, electroencephalogram acquisition and 3D pressure mapping technology are combined, and a deep learning pattern recognition algorithm is developed, so as to accurately test and adaptively adjust the tightness of the helmet wearing, re-collect data of the subject according to the automatically adjusted helmet wearing parameters, verify the adjustment effect and analyze until the helmet wearing is tight, thereby improving the comfort and protection effect of the helmet. The present application solves the problem that the test result of the helmet in the prior art deviates from the actual use, and it is difficult to accurately evaluate the fitting degree of the helmet and the head. Attached Figure Description

[0016] Figure 1 A hardware structure block diagram of a mobile terminal for performing a helmet wearing tightness test method according to an embodiment of this application is shown;

[0017] Figure 2 A schematic flowchart of a helmet fit tightness testing method according to an embodiment of this application is shown;

[0018] Figure 3 A structural block diagram of a helmet fit tightness testing device provided according to an embodiment of this application is shown.

[0019] The above figures include the following reference numerals:

[0020] 102. Processor; 104. Memory; 106. Transmission device; 108. Input / output device. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] As described in the background section, existing helmet fit testing methods often overlook the diversity of actual usage scenarios, such as weather conditions and movement states, leading to discrepancies between test results and actual usage. To address the problem that discrepancies between helmet test results and actual usage make it difficult to accurately assess the fit between the helmet and the head, embodiments of this application provide a helmet fit fit testing method, a helmet fit fit testing device, a computer-readable storage medium, and a computer program product.

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0026] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for performing a helmet-wearing tightness test according to an embodiment of the present invention. Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0027] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the helmet fit tightness test method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. Specific examples of the aforementioned networks may include wireless networks provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0028] This embodiment provides a helmet wearing tightness test method that runs on a mobile terminal, computer terminal or similar computing device. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although the logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0029] Figure 2 This is a schematic flowchart of a helmet fit tightness test method according to an embodiment of this application. Figure 2 As shown, the method includes the following steps:

[0030] Step S201, acquisition step, acquire effective subject data of subjects wearing helmets in different simulated scenarios. The effective subject data includes effective eye-tracking data, head micro-movement data and EEG signals after data processing. The eye-tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data.

[0031] Specifically, eye-tracking devices were used to record the subjects' gaze movement trajectories, fixation points, and blink counts. Sensors were used to collect data on the subjects' head linear acceleration, velocity, and angular velocity. Electroencephalography (EEG) was used to record the subjects' brain signals. The collected data underwent preprocessing to obtain the aforementioned valid subject data, including filtering and artifact removal, to ensure data validity. The goal was to ensure that the collected data accurately reflected the subjects' real-world state in the simulated scenario.

[0032] Step S202, generation step: using 3D pressure mapping technology to monitor the pressure distribution inside the helmet and the contact surface with the subject's head, and generate a pressure distribution map.

[0033] Specifically, 3D pressure mapping technology is used to monitor the pressure distribution inside the helmet and on the contact surface with the subject's head. A pressure distribution map is generated based on the monitoring results, showing the pressure points where the helmet contacts the head. Providing detailed pressure information about the helmet's contact with the head helps identify issues with the helmet's fit.

[0034] Step S203, pattern recognition step: input the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing and obtain the helmet wearing tightness pattern of the subject. The tightness pattern is divided into tight helmet wearing or loose helmet wearing.

[0035] Specifically, pressure distribution maps and valid subject data are input into a pattern recognition model. The model analyzes the data, determines the tightness of the helmet fit, and outputs a pattern indicating whether the helmet is worn tightly or loosely. Pattern recognition technology quickly determines the helmet's fit status, providing a basis for subsequent adjustments.

[0036] Step S204, adjustment step: In the case where the helmet is not worn tightly in the above-mentioned tightness mode, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism based on the above-mentioned pressure distribution map and the above-mentioned valid subject data, so that the helmet can adapt to the head of the subject.

[0037] Specifically, when a helmet is detected as not fitting snugly, an adaptive adjustment mechanism automatically adjusts the helmet's fitting parameters. Based on these parameters, the helmet's position or tightness is automatically adjusted to fit the subject's head. This automated adjustment process reduces human intervention, improving efficiency and accuracy. This allows the helmet to better fit the subject's head, enhancing comfort and functionality.

[0038] Step S205, verification step: Based on the automatically adjusted helmet wearing parameters, reacquire the subject's perception data, and verify the adjustment effect based on the subject's perception data. The subject's perception data includes the subject's perception data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, and effective adjustment means that the helmet is worn tightly.

[0039] Specifically, based on the automatically adjusted helmet wearing parameters, the subjects' sensory data, physiological data, electroencephalogram (EEG) signal data, and pressure distribution data are collected again. The re-collected data is analyzed to verify the adjustment effect and determine whether the helmet fits snugly. Ensuring the adjusted helmet fit meets requirements improves the accuracy and reliability of the simulation test. If the adjustment is ineffective, further adjustments can be made promptly.

[0040] Step S206: If the above adjustment is ineffective, repeat the above acquisition step, the above generation step, the above pattern recognition step, the above adjustment step, and the above verification step at least once until the helmet is securely worn.

[0041] Specifically, if the verification results show that the adjustment is ineffective, the acquisition, generation, pattern recognition, adjustment, and verification steps are repeated until the helmet fits snugly. This iterative process ensures a snug fit until a satisfactory result is achieved. This improves the system's robustness, ensuring optimal helmet fit for different subjects and in different simulated scenarios. The entire process aims to ensure a snug fit in simulated scenarios, thereby improving subject comfort and the accuracy of test data. Through automated and intelligent adjustment mechanisms, this process can adapt to different subject head characteristics, achieving personalized helmet fit adjustments.

[0042] In this embodiment, firstly, an acquisition step is performed to acquire valid subject data when the subject wears the helmet in different simulated scenarios. This valid subject data includes processed eye-tracking data, head micro-motion data, and electroencephalogram (EEG) signals. The eye-tracking data includes at least the subject's gaze trajectory, fixation point, and blink count. The head micro-motion data includes at least the subject's head linear acceleration, velocity, and angular velocity data. Next, a generation step is performed to monitor the pressure distribution inside the helmet and on the contact surface with the subject's head using 3D pressure mapping technology, generating a pressure distribution map. Following this, a pattern recognition step is performed to input the pressure distribution map and the valid subject data into a pattern recognition model to determine the helmet's tightness, obtaining the subject's helmet tightness pattern, which is categorized as either a tight or loose helmet fit. Finally, an adjustment step is performed... The aforementioned tightness mode involves automatically adjusting the helmet wearing parameters using an adaptive adjustment mechanism based on the pressure distribution map and valid subject data when the helmet is not worn tightly, so that the helmet fits the subject's head. Then, in the verification step, the subject's sensory data is reacquired based on the automatically adjusted helmet wearing parameters, and the adjustment effect is verified based on this data. The subject's sensory data includes the subject's sensory data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment indicates that the helmet is still not worn tightly, while effective adjustment indicates that the helmet is worn tightly. Finally, if the adjustment effect is ineffective, the acquisition step, generation step, pattern recognition step, adjustment step, and verification step are repeated at least once until the helmet is worn tightly. This application achieves accurate testing and adaptive adjustment of helmet wearing tightness by combining eye-tracking, EEG signal acquisition, and 3D pressure mapping technology, along with a developed deep learning pattern recognition algorithm. Based on the automatically adjusted helmet wearing parameters, subject data is reacquired, the adjustment effect is verified, and analysis is performed until the helmet is worn tightly, thereby improving the helmet's comfort and protective effect. This application addresses the problem in the prior art where discrepancies exist between helmet test results and actual usage, making it difficult to accurately assess the fit between the helmet and the head.

[0043] To enable those skilled in the art to better understand the technical solution of this application, the implementation process of the helmet wearing tightness test method of this application will be described in detail below with reference to specific embodiments.

[0044] To comprehensively assess the subjects' behavior and physiological responses in a virtual environment, in one optional implementation, step S201 includes:

[0045] Step S2011: If the preparation conditions are met, start the 4D environment simulator and randomly generate different simulation scenarios. The preparation conditions are that the subject cleans his head and wears a helmet as instructed and sits in the simulation environment to prepare for different scenario tests.

[0046] Step S2012: Based on the above simulated scenario, virtual instructions indicating the helmet wearing status are superimposed in the field of vision of the subject in real time using AR technology, so that the subject can complete the corresponding actions according to the virtual instructions and obtain raw eye-tracking data and raw head micro-movement data in real time.

[0047] Step S2013: Perform data cleaning on the above-mentioned original eye-tracking data and the above-mentioned original head micro-motion data to obtain valid eye-tracking data and valid head micro-motion data.

[0048] Step S2014: Real-time monitoring of the raw EEG signals of the subjects using a high-resolution EEG signal acquisition system combined with signal processing technology;

[0049] Step S2015: Perform preprocessing operations on the above-mentioned raw EEG signals to obtain intermediate EEG signals, and use Fourier transform to extract characteristic signals from the above-mentioned intermediate EEG signals to obtain effective EEG signals. The above preprocessing operations include at least artifact removal operations.

[0050] In the above embodiment, the subject first needs to clean their head to reduce signal interference and wear a helmet as instructed, which is typically a head-mounted display (HMD) with sensors. The subject sits in a simulated environment, ready to undergo testing in different scenarios. A 4D environment simulator is activated, which can generate a virtual environment with multiple sensory stimuli, including visual, auditory, and tactile stimuli. Different simulated scenarios are randomly generated, which may include different environments, tasks, or challenges to test the subject's responses. By simulating real or fictional environments, experimental conditions can be controlled, reducing interference from external variables. Randomly generated scenarios increase the diversity and unpredictability of the test, helping to collect more comprehensive data. Using AR technology, virtual instructions are overlaid in real time in the subject's field of vision. These instructions may include arrows, cursors, or other visual cues. The subject performs corresponding actions based on these virtual instructions, such as moving their head or eye-tracking. Simultaneously, the system collects the subject's eye-tracking data and head micro-movement data in real time. AR technology provides an intuitive interactive method, enabling the subject to interact with the virtual environment more naturally. The collected raw eye-tracking data and head micro-movement data are cleaned to remove noise and outliers. Algorithms identify and eliminate invalid data, retaining only valid data for subsequent analysis. Data cleaning ensures the accuracy and reliability of the data, providing a high-quality data foundation for later analysis. Valid data can more accurately reflect the subjects' behavior and responses. A high-resolution EEG signal acquisition system is used to monitor the subjects' brain activity in real time. This system may include multiple electrodes distributed on the subject's scalp to capture electrical activity in different brain regions. EEG signal monitoring can provide direct information about the subjects' cognitive state and emotional responses. The raw EEG signals are preprocessed, including artifact removal operations such as removing eye movement artifacts and electromyography artifacts. The preprocessed signals undergo Fourier transform to extract frequency components, i.e., characteristic signals. These characteristic signals can reflect specific patterns of brain activity, such as alpha waves and beta waves. Preprocessing improves the signal-to-noise ratio of the EEG signals, making them easier to analyze. By integrating multiple technologies and data, a comprehensive assessment of subjects' behavior and physiological responses in a virtual environment can be achieved, which is of great significance for the development of psychology, cognitive science, neuroscience, and virtual reality technology.

[0051] To improve the quality of EEG signals, in one optional implementation, step S2015 includes:

[0052] Step S20151: Perform the artifact removal operation on the original EEG signal according to the artifact removal formula to obtain the intermediate EEG signal. The artifact removal formula is as follows: Where Artifact = |e in (t)-μ e |>ασ e ein (t) represents the original EEG signal at time t, μ e σ is the average value of the above raw EEG signals. e The standard deviation of the above raw EEG signal is given, α is the threshold coefficient, Artifact is a Boolean value, and e is the standard deviation of the original EEG signal. clean (t) represents the intermediate EEG signal after artifact removal, e in (t-1) represents the artifact-free original EEG signal value at time t-1, e in (t+1) represents the original EEG signal value without artifacts at time t+1;

[0053] Step S20152: Feature extraction is performed on the aforementioned intermediate EEG signal according to the Fourier transform formula to obtain the aforementioned EEG signal. The aforementioned Fourier transform formula is: Where F(ω) is the EEG signal obtained by Fourier transform, f(t) is the time domain signal, representing the function of the intermediate EEG signal, t is time, ω is angular frequency, and j is the imaginary unit.

[0054] In the above embodiments, the mean and standard deviation are calculated: the mean μ of the raw EEG signal is calculated. e and standard deviation σ e For each time point t, calculate |e in (t)-μ e |, if this absolute value is greater than ασ e If an artifact is found at a given point, it is considered to exist, i.e., Artifact = True; otherwise, Artifact = False. Artifact removal is then performed according to the above formula. Artifacts in the EEG signal are removed; these artifacts may be caused by eye movement, electromyography (EMG) interference, power supply interference, etc., which can interfere with signal analysis and interpretation. By replacing or smoothing artifact points, the quality of the EEG signal is improved, noise caused by artifacts is reduced, and the signal becomes clearer, which is helpful for subsequent feature extraction and pattern recognition. The Fourier transform formula is applied to the artifact-removed intermediate EEG signal to obtain the Fourier transform signal, i.e., the amplitude and phase information at different frequencies. The time-domain signal is converted into a frequency-domain signal, facilitating the analysis of the signal's frequency components. Features such as power spectral density and frequency band energy are extracted from the spectrum; these features are crucial for understanding the physiological meaning of EEG signals and for pattern recognition. By extracting key frequency components, the dimensionality of the data can be reduced, simplifying subsequent data processing and analysis. This improves the quality of the EEG signal and extracts features that are helpful for analysis and recognition.

[0055] To help improve the comfort and safety of the helmet, in one optional implementation, step S202 includes:

[0056] Step S2021: Real-time pressure data is collected on the contact surface between the helmet and the head using a 3D pressure mapping sensor, and the 3D pressure mapping sensor is fixed inside the helmet.

[0057] Step S2022: Convert the pressure data into the pressure distribution map according to the pressure mapping formula. The pressure mapping formula is as follows: Where P(x,y) is the pressure value at coordinate (x,y), (x,y) is the coordinate position on the contact surface inside the helmet, and w i p is the weighting factor for the i-th of the aforementioned 3D pressure mapping sensors. i Let i be the pressure value measured by the i-th 3D pressure mapping sensor mentioned above, where i is the variable index and N is the total number of the 3D pressure mapping sensors mentioned above. The expression for the pressure value at each coordinate point in the pressure distribution map is as follows: Where p′(x,y) is the interpolated pressure value, (x1,y1) and (x2,y2) are the coordinates of adjacent points with known pressure values, and P(x1,y1), P(x2,y1), P(x1,y2) and P(x2,y2) are the aforementioned known pressure values ​​at adjacent points.

[0058] In the above embodiment, 3D pressure mapping sensors are uniformly distributed and fixed on the contact surface inside the helmet. These sensors need to be able to detect minute pressure changes. The sensors monitor the pressure on the contact surface inside the helmet in real time and convert it into electrical signals. The electrical signals are transmitted to a data processing unit via wires or wirelessly to obtain pressure data. Real-time understanding of the pressure distribution inside the helmet is crucial for adjusting the helmet design to improve wearing comfort. Using a given pressure mapping formula, the pressure data collected from each 3D pressure mapping sensor is converted into coordinate points on a coordinate plane. Based on the known pressure values ​​and coordinate positions, an interpolation formula is used to calculate the pressure value at each point on the entire contact surface. The interpolation formula uses the known pressure values ​​of four adjacent points to estimate the pressure value at any point. A continuous pressure distribution map is generated based on all coordinate point positions and the corresponding interpolated pressure values. By analyzing the pressure distribution map, areas with excessively high or low pressure can be identified, allowing for optimization of the helmet design, such as adjusting the thickness or stiffness of the lining material. Real-time monitoring and analysis of the pressure distribution inside the helmet helps improve helmet comfort and safety, and also provides important data support for helmet design and manufacturing.

[0059] In order to effectively predict the helmet's tightness pattern, in one optional implementation, step S203 includes:

[0060] Step S2031: The pressure distribution map and the valid subject data are input into the pattern recognition model. The convolutional layer in the pattern recognition model extracts spatial features from the pressure distribution map and the valid subject data to generate a feature map. Dimensionality reduction pooling is performed in the pooling layer of the pattern recognition model. The fully connected layer in the pattern recognition model converts the feature map into a one-dimensional feature vector. The output layer of the pattern recognition model performs feature processing on the one-dimensional feature vector to output the compactness pattern. The formula for generating the feature map is as follows: y(n,i,j) is the value of the nth feature map in the output at position (i,j), where n, i, and j are variable indices, W(n,k,l) ​​is the weight of the corresponding nth convolutional kernel at position (k,l), x(n,i+k,j+l) is the value of the input data x at position (i+k,j+l), b(n) is the bias coefficient of the corresponding nth convolutional kernel, σ is the activation function, and x is the input data, representing the above pressure distribution map and the above effective subject data.

[0061] In the above embodiments, training the deep learning model includes training the model using a convolutional neural network, extracting spatial features of the input data using convolution to generate a feature map, introducing nonlinearity using an activation function, reducing the data dimensionality and performing pooling operations in a pooling layer, and converting the feature map into a one-dimensional feature vector in a fully connected layer. The formula for generating the feature map is: The loss function formula in the iterative optimization process is: in, For loss function, Let y represent the predicted value, M represent the true value, and i represent the number of samples. The overall process is as follows: First, the stress distribution map and valid subject data are input into the pattern recognition model. Then, convolutional layers in the model are used to extract spatial features. After the convolution operation, an activation function (such as ReLU, Sigmoid, or Tanh) is usually applied to introduce non-linearity and enhance the model's expressive power. Pooling layers: After the convolutional layers, pooling layers (such as max pooling or average pooling) are used for dimensionality reduction and extraction of important features, reducing computational cost. Fully connected layers transform the feature map into a one-dimensional feature vector, preparing it for classification or regression tasks. The output layer processes the one-dimensional feature vector, outputting a compactness pattern. This allows the model to automatically learn features from the input data that distinguish different categories or prediction targets, and to make effective predictions on new data.

[0062] To improve the comfort and usability of the helmet, in one optional embodiment, step S204 includes:

[0063] Step S2041: Feature extraction is performed on the pressure distribution map and the effective subject data to obtain corresponding feature data. The feature data corresponding to the pressure distribution map includes at least the pressure peak, the pressure average, and the uniformity of the pressure distribution. The feature data corresponding to the effective subject data includes at least the fixation point position, fixation duration, blink frequency, head movement speed, acceleration, and movement trajectory.

[0064] Step S2042: Perform feature fusion on all the above feature data to obtain a feature vector;

[0065] Step S2043: Based on the aforementioned feature vector, adjust the helmet wearing parameters according to the adaptive adjustment mechanism formula. The adaptive adjustment mechanism formula includes a pressure adjustment formula and a helmet position adjustment formula. The pressure adjustment formula is ΔP = K. p ·(P desired -P current ), where ΔP is the pressure regulation amount, K p P is the proportional adjustment coefficient. desired For the desired pressure distribution, P current Based on the currently monitored pressure distribution, the above helmet position adjustment formula is: θ=θ0+Δθ, where θ is the new position of the helmet, θ0 is the initial position of the helmet, and Δθ is the adjustment amount of the helmet.

[0066] In the above embodiments, the pressure peak is obtained by analyzing the pressure distribution map and finding the point with the highest pressure value. The average pressure is obtained by calculating the average of all pressure values ​​in the pressure distribution map. The uniformity of the pressure distribution is evaluated by statistically analyzing the standard deviation or variance of the pressure values ​​in the pressure distribution map. The fixation point position is obtained by acquiring the subject's gaze focus position using an eye-tracking device. The fixation duration is obtained by recording the length of time the subject maintains fixation at a specific fixation point. The blink frequency is obtained by recording the number of blinks and the time interval between blinks using an eye-tracking device. The speed of head movement is obtained by measuring the speed of head movement using a head motion tracking device. The acceleration of head movement is obtained by measuring the acceleration of head movement using an accelerometer. The movement trajectory is obtained by recording the path of head movement using a tracking device. Key information representing helmet wearing comfort and the subject's physiological state is extracted from the raw data, providing a basis for subsequent helmet adjustments. All extracted feature data are integrated to form a feature vector. Feature fusion combines multiple feature data into a unified data structure for subsequent algorithm processing. This helps reduce data dimensionality, improve algorithm efficiency, and may improve the model's generalization ability. The adjustment amount mentioned above can be a step size set according to the feature vector, or it can be obtained through fine calculation using other algorithms. By adjusting the pressure distribution within the helmet to more closely approximate the desired pressure distribution, the helmet's comfort and stability are improved. Adjusting the helmet's position allows for a better fit to the user's head, reducing discomfort and enhancing the user experience. Based on feature vectors, an adaptive adjustment mechanism is used to adjust the helmet's wearing parameters. Through this process, the system can automatically adjust the helmet's wearing parameters to adapt to the physiological and psychological needs of different users, improving the helmet's comfort and effectiveness.

[0067] To evaluate the effect of the adjustment, in one optional implementation, step S205 includes:

[0068] Step S2051: Substitute the aforementioned perception data from the subject's perception data into the first analysis formula to obtain the average perception data score. The first analysis formula is: in, S is the average score of the above perception data. i The above-mentioned feeling data is scored for the i-th subject, where i is the variable index and Z is the total number of subjects;

[0069] Step S2052: Substitute the physiological data from the subject's perceived data into the second analytical formula to obtain heart rate variability. The second analytical formula is as follows: Where HRV is heart rate variability, σ RR The standard deviation of heart rate This represents the average heart rate.

[0070] Step S2053: Substitute the aforementioned EEG signal data from the subject's perceived data into the third analysis formula to obtain the power of the alpha wave in the EEG signal. The third analysis formula is as follows: Among them, P α Let P(f) be the power of the α wave mentioned above, and P(f) be the power spectral density of the EEG signal. α,high and f α,low These are the upper and lower limits of the frequency range of the aforementioned alpha wave, respectively;

[0071] Step S2054: Substitute the aforementioned EEG signal data from the subject's perception data into the fourth analysis formula to obtain the aforementioned pressure distribution uniformity index. The aforementioned fourth analysis formula is: Where UI is the pressure distribution uniformity index mentioned above, and P j Let j be the pressure value detected by the j-th 3D pressure mapping sensor, where j is the variable index. V is the average of the pressure values ​​detected by all the aforementioned 3D pressure mapping sensors, where V is the total number of the aforementioned 3D pressure mapping sensors.

[0072] Step S2055: If the tightness condition is met, the adjustment effect is determined to be effective. The tightness condition is that the average value of the above-mentioned feeling data score is greater than the first standard value, the above-mentioned heart rate variability is greater than or equal to the second set threshold and less than or equal to the third standard value, the power of the above-mentioned alpha wave is less than or equal to the fourth standard value, and the above-mentioned pressure distribution uniformity index is less than the fifth standard value.

[0073] Step S2056: If the tightness condition is not met, the above adjustment effect is determined to be invalid.

[0074] In the above embodiments, all subjects' feeling data scores are collected. These scores may be based on questionnaires, self-reports, or other forms of subjective evaluation. Then, a first analytical formula is used to calculate the average, providing a quantitative overall trend in the feeling data scores, helping to understand the subjects' overall feelings about the adjustment. The average can serve as one of the benchmark indicators for evaluating the effectiveness of the adjustment. Physiological data, particularly heart rate data, is collected from the subjects. A second analytical formula is used to calculate heart rate variability (HRV), an important indicator of autonomic nervous system activity that reflects the subject's stress level and emotional state. Analyzing HRV helps to understand the impact of the adjustment on the subject's physiological state. Electroencephalogram (EEG) signal data is collected from the subjects, and a third analytical formula is used to calculate the power of alpha waves. Alpha wave power is an important indicator of brain relaxation and is related to attention and relaxation levels. Analyzing alpha wave power helps to understand the impact of the adjustment on the subject's brain activity. Pressure values ​​detected by the subjects using a 3D pressure mapping sensor are collected, and a fourth analytical formula is used to calculate the pressure distribution uniformity index. The pressure distribution uniformity index reflects the pressure distribution on the contact surface and is related to comfort and health risk, helping to understand the impact of the adjustment on the subject's body pressure distribution. The results obtained from the first to fourth analytical formulas are compared with preset standard values ​​to determine whether the tightness condition is met. This is crucial for comprehensively evaluating the adjustment effect, as it combines data from multiple dimensions, including subjective feelings, physiological responses, brain activity, and body stress distribution. If the tightness condition is met, the adjustment is considered effective; otherwise, it is considered ineffective. This helps decision-makers understand the actual effect of the adjustment measures and optimize them accordingly. If the adjustment is deemed ineffective, it can prompt researchers or decision-makers to reconsider the adjustment strategy and find more effective solutions. By comprehensively analyzing the subjects' subjective feelings, physiological responses, brain activity, and body stress distribution, a comprehensive and scientific basis is provided for evaluating the adjustment effect.

[0075] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0076] This application also provides a helmet fit tightness testing device. It should be noted that the helmet fit tightness testing device of this application embodiment can be used to execute the helmet fit tightness testing method provided in this application embodiment. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0077] The helmet fit tightness testing device provided in the embodiments of this application is described below.

[0078] Figure 3 This is a structural block diagram of a helmet fit tightness testing device according to an embodiment of this application. Figure 3 As shown, the device includes:

[0079] The acquisition unit 10 is used to perform the acquisition step to acquire valid subject data when the subject wears a helmet in different simulated scenarios. The valid subject data includes valid eye-tracking data, head micro-movement data and electroencephalogram (EEG) signals after data processing. The eye-tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data.

[0080] Specifically, eye-tracking devices were used to record the subjects' gaze movement trajectories, fixation points, and blink counts. Sensors were used to collect data on the subjects' head linear acceleration, velocity, and angular velocity. Electroencephalography (EEG) was used to record the subjects' brain signals. The collected data underwent preprocessing to obtain the aforementioned valid subject data, including filtering and artifact removal, to ensure data validity. The goal was to ensure that the collected data accurately reflected the subjects' real-world state in the simulated scenario.

[0081] The generation unit 20 is used to perform the generation step, which monitors the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generates a pressure distribution map.

[0082] Specifically, 3D pressure mapping technology is used to monitor the pressure distribution inside the helmet and on the contact surface with the subject's head. A pressure distribution map is generated based on the monitoring results, showing the pressure points where the helmet contacts the head. Providing detailed pressure information about the helmet's contact with the head helps identify issues with the helmet's fit.

[0083] The pattern recognition unit 30 is used to perform the pattern recognition step, inputting the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing, and obtaining the helmet wearing tightness pattern of the subject. The tightness pattern is divided into tight helmet wearing or loose helmet wearing.

[0084] Specifically, pressure distribution maps and valid subject data are input into a pattern recognition model. The model analyzes the data, determines the tightness of the helmet fit, and outputs a pattern indicating whether the helmet is worn tightly or loosely. Pattern recognition technology quickly determines the helmet's fit status, providing a basis for subsequent adjustments.

[0085] The adjustment unit 40 is used to perform the adjustment steps. When the above-mentioned tightness mode is that the helmet is not worn tightly, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism based on the pressure distribution map and the above-mentioned effective subject data, so that the helmet fits the subject's head.

[0086] Specifically, when a helmet is detected as not fitting snugly, an adaptive adjustment mechanism automatically adjusts the helmet's fitting parameters. Based on these parameters, the helmet's position or tightness is automatically adjusted to fit the subject's head. This automated adjustment process reduces human intervention, improving efficiency and accuracy. This allows the helmet to better fit the subject's head, enhancing comfort and functionality.

[0087] The verification unit 50 is used to perform the verification steps. Based on the automatically adjusted helmet wearing parameters, it re-acquires the subject's feeling data and verifies the adjustment effect based on the subject's feeling data. The subject's feeling data includes the subject's feeling data, physiological data, electroencephalogram (EEG) signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, while effective adjustment means that the helmet is worn tightly.

[0088] Specifically, based on the automatically adjusted helmet wearing parameters, the subjects' sensory data, physiological data, electroencephalogram (EEG) signal data, and pressure distribution data are collected again. The re-collected data is analyzed to verify the adjustment effect and determine whether the helmet fits snugly. Ensuring the adjusted helmet fit meets requirements improves the accuracy and reliability of the simulation test. If the adjustment is ineffective, further adjustments can be made promptly.

[0089] The repeating unit 60 is used to sequentially repeat the above acquisition step, the above generation step, the above pattern recognition step, the above adjustment step, and the above verification step at least once until the helmet is worn securely when the above adjustment effect is invalid.

[0090] Specifically, if the verification results show that the adjustment is ineffective, the acquisition, generation, pattern recognition, adjustment, and verification steps are repeated until the helmet fits snugly. This iterative process ensures a snug fit until a satisfactory result is achieved. This improves the system's robustness, ensuring optimal helmet fit for different subjects and in different simulated scenarios. The entire process aims to ensure a snug fit in simulated scenarios, thereby improving subject comfort and the accuracy of test data. Through automated and intelligent adjustment mechanisms, this process can adapt to different subject head characteristics, achieving personalized helmet fit adjustments.

[0091] In this embodiment, the acquisition unit is used to perform the acquisition step, acquiring effective subject data when the subject wears the helmet in different simulated scenarios. The effective subject data includes effective eye-tracking data, head micro-movement data, and electroencephalogram (EEG) signals after data processing. The eye-tracking data includes at least the subject's gaze movement trajectory, fixation point, and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity, and angular velocity data. The generation unit is used to perform the generation step, monitoring the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology to generate a pressure distribution map. The pattern recognition unit is used to perform the pattern recognition step, inputting the pressure distribution map and the effective subject data into a pattern recognition model to determine the helmet's tightness and obtain the subject's helmet tightness pattern, which is divided into a tight helmet fit or a loose helmet fit. The adjustment unit is used to... The adjustment step involves automatically adjusting the helmet wearing parameters using an adaptive adjustment mechanism based on the pressure distribution map and valid subject data when the helmet is not worn tightly, in order to make the helmet fit the subject's head. The verification unit performs the verification step by re-acquiring the subject's sensory data based on the automatically adjusted helmet wearing parameters and verifying the adjustment effect based on this data. The sensory data includes the subject's sensory data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment indicates that the helmet is still not worn tightly, while effective adjustment indicates that the helmet is worn tightly. The repetition unit, in the case where the adjustment effect is ineffective, sequentially repeats the acquisition step, generation step, pattern recognition step, adjustment step, and verification step at least once until the helmet is worn tightly. This application combines eye-tracking, electroencephalogram (EEG) signal acquisition, and 3D pressure mapping technologies with a developed deep learning pattern recognition algorithm to achieve precise testing and adaptive adjustment of helmet fit. Based on the automatically adjusted helmet wearing parameters, the application re-collects subject data, verifies the adjustment effect, and analyzes it until the helmet fits snugly, thereby improving helmet comfort and protective effectiveness. This application solves the problem in existing technologies where discrepancies between helmet test results and actual usage make it difficult to accurately assess the helmet's fit to the head.

[0092] To comprehensively assess the subjects' behavior and physiological responses in a virtual environment, in one optional implementation, the acquisition unit includes:

[0093] The simulation module is activated to start the 4D environment simulator when the preparation conditions are met, and to randomly generate different simulation scenarios. The preparation conditions are that the subject cleans his head and wears a helmet as instructed and sits in the simulation environment to prepare for different scenario tests.

[0094] The virtual instruction module is used to overlay virtual instructions indicating the helmet wearing status in real time in the field of vision of the subject using AR technology according to the above-mentioned simulated scenario, so that the subject can complete the corresponding actions according to the virtual instructions and obtain raw eye-tracking data and raw head micro-movement data in real time.

[0095] The data cleaning module is used to perform data cleaning operations on the above-mentioned raw eye-tracking data and the above-mentioned raw head micro-motion data to obtain valid eye-tracking data and valid head micro-motion data.

[0096] The monitoring module is used to monitor the raw EEG signals of the subjects in real time using a high-resolution EEG signal acquisition system combined with signal processing technology.

[0097] The preprocessing module is used to preprocess the raw EEG signal to obtain an intermediate EEG signal, and to extract the characteristic signal from the intermediate EEG signal using Fourier transform to obtain the effective EEG signal. The preprocessing operation includes at least artifact removal.

[0098] In the above embodiment, the subject first needs to clean their head to reduce signal interference and wear a helmet as instructed, which is typically a head-mounted display (HMD) with sensors. The subject sits in a simulated environment, ready to undergo testing in different scenarios. A 4D environment simulator is activated, which can generate a virtual environment with multiple sensory stimuli, including visual, auditory, and tactile stimuli. Different simulated scenarios are randomly generated, which may include different environments, tasks, or challenges to test the subject's responses. By simulating real or fictional environments, experimental conditions can be controlled, reducing interference from external variables. Randomly generated scenarios increase the diversity and unpredictability of the test, helping to collect more comprehensive data. Using AR technology, virtual instructions are overlaid in real time in the subject's field of vision. These instructions may include arrows, cursors, or other visual cues. The subject performs corresponding actions based on these virtual instructions, such as moving their head or eye-tracking. Simultaneously, the system collects the subject's eye-tracking data and head micro-movement data in real time. AR technology provides an intuitive interactive method, enabling the subject to interact with the virtual environment more naturally. The collected raw eye-tracking data and head micro-movement data are cleaned to remove noise and outliers. Algorithms identify and eliminate invalid data, retaining only valid data for subsequent analysis. Data cleaning ensures the accuracy and reliability of the data, providing a high-quality data foundation for later analysis. Valid data can more accurately reflect the subjects' behavior and responses. A high-resolution EEG signal acquisition system is used to monitor the subjects' brain activity in real time. This system may include multiple electrodes distributed on the subject's scalp to capture electrical activity in different brain regions. EEG signal monitoring can provide direct information about the subjects' cognitive state and emotional responses. The raw EEG signals are preprocessed, including artifact removal operations such as removing eye movement artifacts and electromyography artifacts. The preprocessed signals undergo Fourier transform to extract frequency components, i.e., characteristic signals. These characteristic signals can reflect specific patterns of brain activity, such as alpha waves and beta waves. Preprocessing improves the signal-to-noise ratio of the EEG signals, making them easier to analyze. By integrating multiple technologies and data, a comprehensive assessment of subjects' behavior and physiological responses in a virtual environment can be achieved, which is of great significance for the development of psychology, cognitive science, neuroscience, and virtual reality technology.

[0099] To improve the quality of EEG signals, in one optional implementation, the preprocessing module includes:

[0100] The artifact removal submodule is used to perform the artifact removal operation on the original EEG signal according to the artifact removal formula to obtain the intermediate EEG signal. The artifact removal formula is as follows: Where Artifact = |e in (t)-μ e |>ασ e ein (t) represents the original EEG signal at time t, μ e σ is the average value of the above raw EEG signals. e The standard deviation of the above raw EEG signal is given, α is the threshold coefficient, Artifact is a Boolean value, and e is the standard deviation of the original EEG signal. clean (t) represents the intermediate EEG signal after artifact removal, e in (t-1) represents the artifact-free original EEG signal value at time t-1, e in (t+1) represents the original EEG signal value without artifacts at time t+1;

[0101] The feature extraction submodule is used to extract features from the aforementioned intermediate EEG signals according to the Fourier transform formula, thereby obtaining the aforementioned EEG signals. The aforementioned Fourier transform formula is: Where F(ω) is the EEG signal obtained by Fourier transform, f(t) is the time domain signal, representing the function of the intermediate EEG signal, t is time, ω is angular frequency, and j is the imaginary unit.

[0102] In the above embodiments, the mean and standard deviation are calculated: the mean μ of the raw EEG signal is calculated. e and standard deviation σ e For each time point t, calculate |e in (t)-μ e |, if this absolute value is greater than ασ e If an artifact is found at a given point, it is considered to exist, i.e., Artifact = True; otherwise, Artifact = False. Artifact removal is then performed according to the above formula. Artifacts in the EEG signal are removed; these artifacts may be caused by eye movement, electromyography (EMG) interference, power supply interference, etc., which can interfere with signal analysis and interpretation. By replacing or smoothing artifact points, the quality of the EEG signal is improved, noise caused by artifacts is reduced, and the signal becomes clearer, which is helpful for subsequent feature extraction and pattern recognition. The Fourier transform formula is applied to the artifact-removed intermediate EEG signal to obtain the Fourier transform signal, i.e., the amplitude and phase information at different frequencies. The time-domain signal is converted into a frequency-domain signal, facilitating the analysis of the signal's frequency components. Features such as power spectral density and frequency band energy are extracted from the spectrum; these features are crucial for understanding the physiological meaning of EEG signals and for pattern recognition. By extracting key frequency components, the dimensionality of the data can be reduced, simplifying subsequent data processing and analysis. This improves the quality of the EEG signal and extracts features that are helpful for analysis and recognition.

[0103] To help improve the comfort and safety of helmets, in one optional embodiment, the above-mentioned generating unit includes:

[0104] The acquisition module is used to acquire pressure data on the contact surface between the helmet and the head in real time through a 3D pressure mapping sensor, and the 3D pressure mapping sensor is fixed inside the helmet.

[0105] The conversion module is used to convert the pressure data into the pressure distribution map according to the pressure mapping formula, which is as follows: Where P(x,y) is the pressure value at coordinate (x,y), (x,y) is the coordinate position on the contact surface inside the helmet, and w i p is the weighting factor for the i-th of the aforementioned 3D pressure mapping sensors. i Let i be the pressure value measured by the i-th 3D pressure mapping sensor mentioned above, where i is the variable index and N is the total number of the 3D pressure mapping sensors mentioned above. The expression for the pressure value at each coordinate point in the pressure distribution map is as follows: Where p′(x,y) is the interpolated pressure value, (x1,y1) and (x2,y2) are the coordinates of adjacent points with known pressure values, and P(x1,y1), P(x2,y1), P(x1,y2) and P(x2,y2) are the aforementioned known pressure values ​​at adjacent points.

[0106] In the above embodiment, 3D pressure mapping sensors are uniformly distributed and fixed on the contact surface inside the helmet. These sensors need to be able to detect minute pressure changes. The sensors monitor the pressure on the contact surface inside the helmet in real time and convert it into electrical signals. The electrical signals are transmitted to a data processing unit via wires or wirelessly to obtain pressure data. Real-time understanding of the pressure distribution inside the helmet is crucial for adjusting the helmet design to improve wearing comfort. Using a given pressure mapping formula, the pressure data collected from each 3D pressure mapping sensor is converted into coordinate points on a coordinate plane. Based on the known pressure values ​​and coordinate positions, an interpolation formula is used to calculate the pressure value at each point on the entire contact surface. The interpolation formula uses the known pressure values ​​of four adjacent points to estimate the pressure value at any point. A continuous pressure distribution map is generated based on all coordinate point positions and the corresponding interpolated pressure values. By analyzing the pressure distribution map, areas with excessively high or low pressure can be identified, allowing for optimization of the helmet design, such as adjusting the thickness or stiffness of the lining material. Real-time monitoring and analysis of the pressure distribution inside the helmet helps improve helmet comfort and safety, and also provides important data support for helmet design and manufacturing.

[0107] To effectively predict the tightness pattern of the helmet, in one optional implementation, the pattern recognition unit includes:

[0108] The input / output module is used to input the pressure distribution map and the valid subject data into the pattern recognition model. The convolutional layer in the pattern recognition model extracts spatial features from the pressure distribution map and the valid subject data to generate a feature map. Dimensionality reduction pooling is performed in the pooling layer of the pattern recognition model. The fully connected layer in the pattern recognition model transforms the feature map into a one-dimensional feature vector. The output layer of the pattern recognition model performs feature processing on the one-dimensional feature vector to output the compactness pattern. The formula for generating the feature map is as follows: y(n,i,j) is the value of the nth feature map in the output at position (i,j), where n, i, and j are variable indices, W(n,k,l) ​​is the weight of the corresponding nth convolutional kernel at position (k,l), x(n,i+k,j+l) is the value of the input data x at position (i+k,j+l), b(n) is the bias coefficient of the corresponding nth convolutional kernel, σ is the activation function, and x is the input data, representing the above pressure distribution map and the above effective subject data.

[0109] In the above embodiments, training the deep learning model includes training the model using a convolutional neural network, extracting spatial features of the input data using convolution to generate a feature map, introducing nonlinearity using an activation function, reducing the data dimensionality and performing pooling operations in a pooling layer, and converting the feature map into a one-dimensional feature vector in a fully connected layer. The formula for generating the feature map is: The loss function formula in the iterative optimization process is: in, For loss function, Let y represent the predicted value, M represent the true value, and i represent the number of samples. The overall process is as follows: First, the stress distribution map and valid subject data are input into the pattern recognition model. Then, convolutional layers in the model are used to extract spatial features. After the convolution operation, an activation function (such as ReLU, Sigmoid, or Tanh) is usually applied to introduce non-linearity and enhance the model's expressive power. Pooling layers: After the convolutional layers, pooling layers (such as max pooling or average pooling) are used for dimensionality reduction and extraction of important features, reducing computational cost. Fully connected layers transform the feature map into a one-dimensional feature vector, preparing it for classification or regression tasks. The output layer processes the one-dimensional feature vector, outputting a compactness pattern. This allows the model to automatically learn features from the input data that distinguish different categories or prediction targets, and to make effective predictions on new data.

[0110] To improve the comfort and usability of the helmet, in one optional embodiment, the adjustment unit includes:

[0111] The feature extraction module is used to extract features from both the pressure distribution map and the valid subject data to obtain corresponding feature data. The feature data corresponding to the pressure distribution map includes at least the pressure peak, the pressure average, and the uniformity of the pressure distribution. The feature data corresponding to the valid subject data includes at least the fixation point position, fixation duration, blink frequency, head movement speed, acceleration, and movement trajectory.

[0112] The feature fusion module is used to fuse all the above feature data to obtain a feature vector;

[0113] The adjustment module is used to adjust the helmet wearing parameters based on the aforementioned feature vectors according to the adaptive adjustment mechanism formula. The adaptive adjustment mechanism formula includes a pressure adjustment formula and a helmet position adjustment formula. The pressure adjustment formula is ΔP = K. p ·(P desired -P current ), where ΔP is the pressure regulation amount, K p P is the proportional adjustment coefficient. desired For the desired pressure distribution, P current Based on the currently monitored pressure distribution, the above helmet position adjustment formula is: θ=θ0+Δθ, where θ is the new position of the helmet, θ0 is the initial position of the helmet, and Δθ is the adjustment amount of the helmet.

[0114] In the above embodiments, the pressure peak is obtained by analyzing the pressure distribution map and finding the point with the highest pressure value. The average pressure is obtained by calculating the average of all pressure values ​​in the pressure distribution map. The uniformity of the pressure distribution is evaluated by statistically analyzing the standard deviation or variance of the pressure values ​​in the pressure distribution map. The fixation point position is obtained by acquiring the subject's gaze focus position using an eye-tracking device. The fixation duration is obtained by recording the length of time the subject maintains fixation at a specific fixation point. The blink frequency is obtained by recording the number of blinks and the time interval between blinks using an eye-tracking device. The speed of head movement is obtained by measuring the speed of head movement using a head motion tracking device. The acceleration of head movement is obtained by measuring the acceleration of head movement using an accelerometer. The movement trajectory is obtained by recording the path of head movement using a tracking device. Key information representing helmet wearing comfort and the subject's physiological state is extracted from the raw data, providing a basis for subsequent helmet adjustments. All extracted feature data are integrated to form a feature vector. Feature fusion combines multiple feature data into a unified data structure for subsequent algorithm processing. This helps reduce data dimensionality, improve algorithm efficiency, and may improve the model's generalization ability. The adjustment amount mentioned above can be a step size set according to the feature vector, or it can be obtained through fine calculation using other algorithms. By adjusting the pressure distribution within the helmet to more closely approximate the desired pressure distribution, the helmet's comfort and stability are improved. Adjusting the helmet's position allows for a better fit to the user's head, reducing discomfort and enhancing the user experience. Based on feature vectors, an adaptive adjustment mechanism is used to adjust the helmet's wearing parameters. Through this process, the system can automatically adjust the helmet's wearing parameters to adapt to the physiological and psychological needs of different users, improving the helmet's comfort and effectiveness.

[0115] To evaluate the adjustment effect, in one optional implementation, the verification unit includes:

[0116] The first calculation module is used to input the aforementioned subject perception data into a first analytical formula to obtain the average perception data score. The first analytical formula is: in, S is the average score of the above perception data. i The above-mentioned feeling data is scored for the i-th subject, where i is the variable index and Z is the total number of subjects;

[0117] The second calculation module is used to input the physiological data from the subject's sensory data into the second analytical formula to obtain heart rate variability. The second analytical formula is as follows: Where HRV is heart rate variability, σ RR The standard deviation of heart rate This represents the average heart rate.

[0118] The third calculation module is used to substitute the aforementioned EEG signal data from the subject's perceived data into the third analysis formula to obtain the power of the alpha wave in the EEG signal. The third analysis formula is as follows: Among them, P α Let P(f) be the power of the α wave mentioned above, and P(f) be the power spectral density of the EEG signal. α,high and f α,low These are the upper and lower limits of the frequency range of the aforementioned alpha wave, respectively;

[0119] The fourth calculation module is used to substitute the aforementioned EEG signal data from the subject's perception data into the fourth analysis formula to obtain the aforementioned pressure distribution uniformity index. The aforementioned fourth analysis formula is: Where UI is the pressure distribution uniformity index mentioned above, and P j Let j be the pressure value detected by the j-th 3D pressure mapping sensor, where j is the variable index. V is the average of the pressure values ​​detected by all the aforementioned 3D pressure mapping sensors, where V is the total number of the aforementioned 3D pressure mapping sensors.

[0120] The first determination module is used to determine that the above adjustment effect is effective when the tightness condition is met. The tightness condition is that the average value of the above feeling data score is greater than the first standard value, the above heart rate variability is greater than or equal to the second set threshold and less than or equal to the third standard value, the power of the above alpha wave is less than or equal to the fourth standard value, and the pressure distribution uniformity index is less than the fifth standard value.

[0121] The second determination module is used to determine that the above adjustment is invalid if the tightness condition is not met.

[0122] In the above embodiments, all subjects' feeling data scores are collected. These scores may be based on questionnaires, self-reports, or other forms of subjective evaluation. Then, a first analytical formula is used to calculate the average, providing a quantitative overall trend in the feeling data scores, helping to understand the subjects' overall feelings about the adjustment. The average can serve as one of the benchmark indicators for evaluating the effectiveness of the adjustment. Physiological data, particularly heart rate data, is collected from the subjects. A second analytical formula is used to calculate heart rate variability (HRV), an important indicator of autonomic nervous system activity that reflects the subject's stress level and emotional state. Analyzing HRV helps to understand the impact of the adjustment on the subject's physiological state. Electroencephalogram (EEG) signal data is collected from the subjects, and a third analytical formula is used to calculate the power of alpha waves. Alpha wave power is an important indicator of brain relaxation and is related to attention and relaxation levels. Analyzing alpha wave power helps to understand the impact of the adjustment on the subject's brain activity. Pressure values ​​detected by the subjects using a 3D pressure mapping sensor are collected, and a fourth analytical formula is used to calculate the pressure distribution uniformity index. The pressure distribution uniformity index reflects the pressure distribution on the contact surface and is related to comfort and health risk, helping to understand the impact of the adjustment on the subject's body pressure distribution. The results obtained from the first to fourth analytical formulas are compared with preset standard values ​​to determine whether the tightness condition is met. This is crucial for comprehensively evaluating the adjustment effect, as it combines data from multiple dimensions, including subjective feelings, physiological responses, brain activity, and body stress distribution. If the tightness condition is met, the adjustment is considered effective; otherwise, it is considered ineffective. This helps decision-makers understand the actual effect of the adjustment measures and optimize them accordingly. If the adjustment is deemed ineffective, it can prompt researchers or decision-makers to reconsider the adjustment strategy and find more effective solutions. By comprehensively analyzing the subjects' subjective feelings, physiological responses, brain activity, and body stress distribution, a comprehensive and scientific basis is provided for evaluating the adjustment effect.

[0123] The aforementioned helmet fit tightness testing device includes a processor and a memory. The acquisition unit, generation unit, and pattern recognition unit are all stored as program units in the memory, and the processor executes these program units to achieve their respective functions. All of the above modules reside in the same processor; alternatively, the modules may be located in different processors in any combination.

[0124] The processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and adjusting kernel parameters can address the problem in existing technologies where discrepancies between helmet test results and actual usage make it difficult to accurately assess the fit between the helmet and the head.

[0125] The memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0126] This invention provides a computer-readable storage medium including a stored program, wherein when the program is executed, it controls the device containing the computer-readable storage medium to perform the helmet wearing tightness test method.

[0127] Specifically, the methods for testing the tightness of helmet fit include:

[0128] Step S201, acquisition step, acquire effective subject data when the subject wears the helmet in different simulated scenarios. The effective subject data includes effective eye tracking data, head micro-movement data and EEG signals after data processing. The eye tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data.

[0129] Step S202, generation step: monitor the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generate a pressure distribution map;

[0130] Step S203, pattern recognition step: input the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing and obtain the helmet wearing tightness pattern of the subject. The tightness pattern is divided into tight helmet wearing or loose helmet wearing.

[0131] Step S204, adjustment step: when the above tightness mode is that the helmet is not worn tightly, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism based on the above pressure distribution map and the above effective subject data, so that the helmet adapts to the head of the subject.

[0132] Step S205, verification step: Based on the automatically adjusted helmet wearing parameters, reacquire the subject's feeling data, and verify the adjustment effect based on the subject's feeling data. The subject's feeling data includes the subject's feeling data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, and effective adjustment means that the helmet is worn tightly.

[0133] Step S206: If the above adjustment is ineffective, repeat the above acquisition step, the above generation step, the above pattern recognition step, the above adjustment step, and the above verification step at least once until the helmet is securely worn.

[0134] This invention provides a processor for running a program, wherein the program executes the helmet wearing tightness test method.

[0135] Specifically, the methods for testing the tightness of helmet fit include:

[0136] Step S201, acquisition step, acquire effective subject data when the subject wears the helmet in different simulated scenarios. The effective subject data includes effective eye tracking data, head micro-movement data and EEG signals after data processing. The eye tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data.

[0137] Step S202, generation step: monitor the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generate a pressure distribution map;

[0138] Step S203, pattern recognition step: input the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing and obtain the helmet wearing tightness pattern of the subject. The tightness pattern is divided into tight helmet wearing or loose helmet wearing.

[0139] Step S204, adjustment step: when the above tightness mode is that the helmet is not worn tightly, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism based on the above pressure distribution map and the above effective subject data, so that the helmet adapts to the head of the subject.

[0140] Step S205, verification step: Based on the automatically adjusted helmet wearing parameters, reacquire the subject's feeling data, and verify the adjustment effect based on the subject's feeling data. The subject's feeling data includes the subject's feeling data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, and effective adjustment means that the helmet is worn tightly.

[0141] Step S206: If the above adjustment is ineffective, repeat the above acquisition step, the above generation step, the above pattern recognition step, the above adjustment step, and the above verification step at least once until the helmet is securely worn.

[0142] This invention provides a helmet, which includes a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it performs at least the following steps:

[0143] Step S201, acquisition step, acquire effective subject data when the subject wears the helmet in different simulated scenarios. The effective subject data includes effective eye tracking data, head micro-movement data and EEG signals after data processing. The eye tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data.

[0144] Step S202, generation step: monitor the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generate a pressure distribution map;

[0145] Step S203, pattern recognition step: input the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing and obtain the helmet wearing tightness pattern of the subject. The tightness pattern is divided into tight helmet wearing or loose helmet wearing.

[0146] Step S204, adjustment step: when the above tightness mode is that the helmet is not worn tightly, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism based on the above pressure distribution map and the above effective subject data, so that the helmet adapts to the head of the subject.

[0147] Step S205, verification step: Based on the automatically adjusted helmet wearing parameters, reacquire the subject's feeling data, and verify the adjustment effect based on the subject's feeling data. The subject's feeling data includes the subject's feeling data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, and effective adjustment means that the helmet is worn tightly.

[0148] Step S206: If the above adjustment is ineffective, repeat the above acquisition step, the above generation step, the above pattern recognition step, the above adjustment step, and the above verification step at least once until the helmet is securely worn.

[0149] This application also provides a computer program product, which, when executed on a data processing device, is suitable for executing an initialization program having at least the following method steps:

[0150] Step S201, acquisition step, acquire effective subject data when the subject wears the helmet in different simulated scenarios. The effective subject data includes effective eye tracking data, head micro-movement data and EEG signals after data processing. The eye tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data.

[0151] Step S202, generation step: monitor the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generate a pressure distribution map;

[0152] Step S203, pattern recognition step: input the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing and obtain the helmet wearing tightness pattern of the subject. The tightness pattern is divided into tight helmet wearing or loose helmet wearing.

[0153] Step S204, adjustment step: when the above tightness mode is that the helmet is not worn tightly, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism based on the above pressure distribution map and the above effective subject data, so that the helmet adapts to the head of the subject.

[0154] Step S205, verification step: Based on the automatically adjusted helmet wearing parameters, reacquire the subject's feeling data, and verify the adjustment effect based on the subject's feeling data. The subject's feeling data includes the subject's feeling data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, and effective adjustment means that the helmet is worn tightly.

[0155] Step S206: If the above adjustment is ineffective, repeat the above acquisition step, the above generation step, the above pattern recognition step, the above adjustment step, and the above verification step at least once until the helmet is securely worn.

[0156] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0157] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0158] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0159] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0160] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process.Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0161] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0162] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0163] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0164] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0165] As can be seen from the above description, the embodiments of this application achieve the following technical effects:

[0166] 1) The helmet wearing tightness testing method of this application includes, firstly, an acquisition step, acquiring effective subject data of subjects wearing helmets in different simulated scenarios. The effective subject data includes processed eye-tracking data, head micro-movement data, and electroencephalogram (EEG) signals. The eye-tracking data includes at least the subject's gaze trajectory, fixation point, and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity, and angular velocity data. Then, a generation step involves monitoring the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology to generate a pressure distribution map. Next, a pattern recognition step inputs the pressure distribution map and the effective subject data into a pattern recognition model to determine the helmet wearing tightness, obtaining the subject's helmet wearing tightness pattern, which is categorized as either a tight helmet fit or a loose helmet fit. Finally, adjustments are made... The entire process involves the following steps: If the helmet is not fitted snugly according to the aforementioned fit pattern, an adaptive adjustment mechanism automatically adjusts the helmet fitting parameters based on the pressure distribution map and valid subject data to adapt the helmet to the subject's head. Next, a verification step involves re-acquiring the subject's sensory data based on the automatically adjusted helmet fitting parameters and verifying the adjustment effect. The subject's sensory data includes sensory data, physiological data, EEG signal data, and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment indicates the helmet is still not fitted snugly, while effective adjustment indicates a snug fit. Finally, if the adjustment is ineffective, the acquisition, generation, pattern recognition, adjustment, and verification steps are repeated at least once until the helmet fits snugly. This application combines eye-tracking, EEG signal acquisition, and 3D pressure mapping technologies, along with a developed deep learning pattern recognition algorithm, to achieve accurate testing and adaptive adjustment of helmet fit snugness. By re-collecting subject data based on the automatically adjusted helmet fitting parameters, verifying the adjustment effect, and analyzing it until the helmet fits snugly, the comfort and protective effect of the helmet are improved. This application addresses the problem in the prior art where discrepancies exist between helmet test results and actual usage, making it difficult to accurately assess the fit between the helmet and the head.

[0167] 2) The helmet wearing tightness testing device of this application includes an acquisition unit for performing an acquisition step, acquiring valid subject data of subjects wearing helmets in different simulated scenarios. The valid subject data includes processed eye-tracking data, head micro-movement data, and electroencephalogram (EEG) signals. The eye-tracking data includes at least the subject's gaze trajectory, fixation point, and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity, and angular velocity data. A generation unit is used to perform a generation step, monitoring the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology to generate a pressure distribution map. A pattern recognition unit is used to perform a pattern recognition step, inputting the pressure distribution map and the valid subject data into a pattern recognition model to determine the helmet wearing tightness and obtain the subject's helmet wearing tightness pattern, which is divided into helmet wearing tightness or helmet wearing looseness. The system comprises the following units: an adjustment unit for adjusting the helmet wearing parameters to fit the subject's head, based on the pressure distribution map and valid subject data, when the helmet is not fitted snugly; a verification unit for performing a verification step, which re-acquires the subject's sensory data based on the automatically adjusted helmet wearing parameters and verifies the adjustment effect based on the sensory data, including the subject's sensory data, physiological data, EEG signal data, and pressure distribution data; and a verification unit for repeating the acquisition, generation, pattern recognition, adjustment, and verification steps at least once when the adjustment is deemed ineffective. This application combines eye-tracking, electroencephalogram (EEG) signal acquisition, and 3D pressure mapping technologies with a developed deep learning pattern recognition algorithm to achieve precise testing and adaptive adjustment of helmet fit. Based on the automatically adjusted helmet wearing parameters, the application re-collects subject data, verifies the adjustment effect, and analyzes it until the helmet fits snugly, thereby improving helmet comfort and protective effectiveness. This application solves the problem in existing technologies where discrepancies between helmet test results and actual usage make it difficult to accurately assess the helmet's fit to the head.

[0168] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for testing the tightness of helmet fitting, characterized in that, include: The acquisition step involves acquiring valid subject data when subjects wear helmets in different simulated scenarios. The valid subject data includes valid eye-tracking data, head micro-movement data, and electroencephalogram (EEG) signals after data processing. The eye-tracking data includes at least the subject's gaze movement trajectory, fixation point, and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity, and angular velocity data. The generation step involves monitoring the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generating a pressure distribution map. The pattern recognition step involves inputting the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing and obtain the subject's helmet wearing tightness pattern, which is divided into helmet wearing tightness or helmet wearing looseness. The adjustment steps involve automatically adjusting the helmet wearing parameters using an adaptive adjustment mechanism based on the pressure distribution map and the valid subject data when the helmet is not worn tightly in the tightness mode, so that the helmet fits the subject's head. The verification step involves re-acquiring the subject's sensory data based on the automatically adjusted helmet wearing parameters, and verifying the adjustment effect based on the subject's sensory data. The subject's sensory data includes the subject's sensory data, physiological data, electroencephalogram (EEG) signal data, and pressure distribution data. The adjustment effect is defined as either ineffective or effective. An ineffective adjustment indicates that the helmet is still not worn tightly, while an effective adjustment indicates that the helmet is worn tightly. If the adjustment is ineffective, the acquisition step, generation step, pattern recognition step, adjustment step, and verification step are repeated at least once until the helmet is securely fitted.

2. The method according to claim 1, characterized in that, Obtain valid subject data when subjects wear helmets in different simulated scenarios, including: Once the preparation conditions are met, the 4D environment simulator is started and different simulation scenarios are randomly generated. The preparation conditions are that the subject cleans his / her head and wears a helmet as instructed and sits in the simulation environment to prepare for different scenario tests. Based on the simulated scenario, AR technology is used to overlay virtual instructions indicating the helmet wearing status in real time into the subject's field of vision, so that the subject can complete the corresponding actions according to the virtual instructions and obtain raw eye-tracking data and raw head micro-movement data in real time. The raw eye-tracking data and the raw head micro-motion data are cleaned to obtain valid eye-tracking data and valid head micro-motion data; The raw EEG signals of the subjects were monitored in real time using a high-resolution EEG signal acquisition system combined with signal processing technology; The original EEG signal is preprocessed to obtain an intermediate EEG signal, and the characteristic signals in the intermediate EEG signal are extracted using Fourier transform to obtain the effective EEG signal. The preprocessing operation includes at least artifact removal.

3. The method according to claim 2, characterized in that, The raw EEG signal is preprocessed to obtain an intermediate EEG signal, and Fourier transform is used to extract characteristic signals from the intermediate EEG signal to obtain the effective EEG signal, including: The original EEG signal is subjected to the artifact removal operation according to the artifact removal formula to obtain the intermediate EEG signal. The artifact removal formula is as follows: Where Artifact = |e in (t)-μ e |>ασ e e in (t) represents the original EEG signal at time t, μ e σ is the average value of the original EEG signal. e The standard deviation of the original EEG signal is given, α is the threshold coefficient, Artifact is a Boolean value, and e is the threshold value. clean (t) represents the intermediate EEG signal after artifact removal, e in (t-1) represents the artifact-free original EEG signal value at time t-1, e in (t+1) represents the original EEG signal value without artifacts at time t+1; Feature extraction is performed on the intermediate EEG signal according to the Fourier transform formula to obtain the EEG signal. The Fourier transform formula is as follows: Wherein, F(ω) is the EEG signal obtained by Fourier transform, f(t) is the time-domain signal, representing a function of the intermediate EEG signal, t is time, ω is angular frequency, and j is the imaginary unit.

4. The method according to claim 1, characterized in that, The pressure distribution inside the helmet and on the subject's head contact surface is monitored using 3D pressure mapping technology to generate a pressure distribution map, including: The pressure data between the helmet and the head contact surface is collected in real time by a 3D pressure mapping sensor, and the 3D pressure mapping sensor is fixed inside the helmet. All pressure data are converted into the pressure distribution map according to the pressure mapping formula, which is as follows: Where P(x,y) is the pressure value at coordinate (x,y), (x,y) is the coordinate position on the contact surface inside the helmet, and w i p is the weighting factor for the i-th 3D pressure mapping sensor. i Let i be the pressure value measured by the i-th 3D pressure mapping sensor, i be the variable index, and N be the total number of 3D pressure mapping sensors. The expression for the pressure value at each coordinate point in the pressure distribution map is: Where p′(x,y) is the interpolated pressure value, (x1,y1) and (x2,y2) are the coordinates of adjacent points of known pressure value, and P(x1,y1), P(x2,y1), P(x1,y2) and P(x2,y2) are the known pressure values ​​at adjacent points.

5. The method according to claim 1, characterized in that, The pressure distribution map and the valid subject data are input into a pattern recognition model to obtain the helmet wearing tightness pattern of the subject, including: The pressure distribution map and the valid subject data are input into the pattern recognition model. The convolutional layer in the pattern recognition model extracts the spatial features of the pressure distribution map and the valid subject data to generate a feature map. Dimensionality reduction pooling is then performed in the pooling layer of the pattern recognition model. The fully connected layer in the pattern recognition model transforms the feature map into a one-dimensional feature vector. Finally, the output layer of the pattern recognition model performs feature processing on the one-dimensional feature vector to output the compactness pattern. The formula for generating the feature map is as follows: y(n,i,j) is the value of the nth feature map at position (i,j), where n, i, and j are variable indices, W(n,k,l) ​​is the weight of the corresponding nth convolutional kernel at position (k,l), x(n,i+k,j+l) is the value of the input data x at position (i+k,j+l), b(n) is the bias coefficient of the corresponding nth convolutional kernel, σ is the activation function, and x is the input data, representing the pressure distribution map and the effective subject data.

6. The method according to claim 1, characterized in that, Based on the pressure distribution map and the valid subject data, the helmet wearing parameters are automatically adjusted using an adaptive adjustment mechanism, including: Feature extraction is performed on both the pressure distribution map and the valid subject data to obtain corresponding feature data. The feature data corresponding to the pressure distribution map includes at least the pressure peak, the pressure average, and the uniformity of the pressure distribution. The feature data corresponding to the valid subject data includes at least the fixation point position, fixation duration, blink frequency, head movement speed, acceleration, and movement trajectory. All the aforementioned feature data are fused to obtain a feature vector; Based on the feature vector, the helmet wearing parameters are adjusted according to an adaptive adjustment mechanism formula. This adaptive adjustment mechanism formula includes a pressure adjustment formula and a helmet position adjustment formula. The pressure adjustment formula is ΔP = K. p ·(P desired -P current ), where ΔP is the pressure regulation amount, K p P is the proportional adjustment coefficient. desired For the desired pressure distribution, P current Based on the currently monitored pressure distribution, the helmet position adjustment formula is: θ = θ0 + Δθ, where θ is the new position of the helmet, θ0 is the initial position of the helmet, and Δθ is the adjustment amount of the helmet.

7. The method according to claim 1, characterized in that, The effectiveness of the adjustment was verified based on the subject's perception data, including: Substituting the perception data from the subject's perception data into the first analytical formula, we obtain the average perception data score. The first analytical formula is: in, S is the average score of the perceived data. i The score is given to the perception data of the i-th subject, where i is the variable index and Z is the total number of subjects; Substituting the physiological data from the subject's sensory data into the second analytical formula, heart rate variability is obtained. The second analytical formula is: Where HRV is heart rate variability, σ RR The standard deviation of heart rate This represents the average heart rate. Substituting the EEG signal data from the subject's perceived data into the third analysis formula, the power of the alpha wave in the EEG signal is obtained. The third analysis formula is as follows: Among them, P α Let P(f) be the power of the α wave, and P(f) be the power spectral density of the EEG signal. α,high and f α,low These are the upper and lower limits of the frequency range of the α wave, respectively; Substituting the EEG signal data from the subject's perception data into the fourth analytical formula, the pressure distribution uniformity index is obtained. The fourth analytical formula is: Where UI is the pressure distribution uniformity index, and P j Let j be the pressure value detected by the j-th 3D pressure mapping sensor, where j is the variable index. V is the average of the pressure values ​​detected by all the 3D pressure mapping sensors, where V is the total number of the 3D pressure mapping sensors. If the tightness condition is met, the adjustment effect is determined to be effective. The tightness condition is that the average value of the feeling data score is greater than the first standard value, the heart rate variability is greater than or equal to the second set threshold and less than or equal to the third standard value, the power of the alpha wave is less than or equal to the fourth standard value, and the pressure distribution uniformity index is less than the fifth standard value. If the tightness condition is not met, the adjustment is deemed invalid.

8. A helmet fitting tightness testing device, characterized in that, The device includes: The acquisition unit is used to perform the acquisition step and acquire effective subject data when the subject wears a helmet in different simulated scenarios. The effective subject data includes effective eye-tracking data, head micro-movement data and electroencephalogram (EEG) signals after data processing. The eye-tracking data includes at least the subject's gaze movement trajectory, fixation point and blink count. The head micro-movement data includes at least the subject's head linear acceleration, velocity and angular velocity data. The generation unit is used to perform the generation step, which monitors the pressure distribution inside the helmet and the contact surface with the subject's head using 3D pressure mapping technology, and generates a pressure distribution map. The pattern recognition unit is used to perform the pattern recognition step, inputting the pressure distribution map and the valid subject data into the pattern recognition model to determine the tightness of the helmet wearing, and obtain the subject's helmet wearing tightness pattern, which is divided into helmet wearing tight or helmet wearing loose; An adjustment unit is used to adjust the steps, in the case where the tightness mode is that the helmet is not worn tightly, automatically adjusting the helmet wearing parameters according to the pressure distribution map and the effective subject data using an adaptive adjustment mechanism, so that the helmet fits the subject's head. The verification unit is used to perform the verification steps, re-acquire the subject's feeling data according to the automatically adjusted helmet wearing parameters, and verify the adjustment effect according to the subject's feeling data. The subject's feeling data includes the subject's feeling data, physiological data, electroencephalogram signal data and pressure distribution data. The adjustment effect is either ineffective or effective. Ineffective adjustment means that the helmet is still not worn tightly, and effective adjustment means that the helmet is worn tightly. The repeating unit is used to sequentially repeat the acquisition step, the generation step, the pattern recognition step, the adjustment step, and the verification step at least once until the helmet is worn securely when the adjustment effect is deemed ineffective.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device on which the computer-readable storage medium is located to perform the method according to any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images