A wearable device for monitoring worker health in high temperature environments

Through the dynamic adjustment scheme of hyperspectral images and sweat electrolyte data, the problems of low accuracy and poor robustness of sweat monitoring in the prior art are solved, and high accuracy and effectiveness monitoring of workers' health status are achieved, with a wide range of applications , better robustness.

CN118781554BActive Publication Date: 2024-11-19BEIJING BRANCH BUILDING TECH CO LTD
View PDF 6 Cites 0 Cited by

Patent Information

Application Number
CN202411276878.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-12
Publication Date
2024-11-19
Estimated Expiration
2044-09-12

AI Technical Summary

Technical Problem

The existing sweat monitoring technology cannot accurately reflect the health status of workers in high temperature environments, especially when individual differences are large, the monitoring accuracy of traditional methods is low and the robustness is poor, and real-time monitoring and personalization cannot be achieved. Dynamic adjustment of thresholds.

Method used

Using hyperspectral images, ambient temperature and humidity and sweat electrolyte data, through hyperspectral cameras, sweat sensors, sweat feature extraction modules, threshold dynamic adjustment modules and monitoring result output modules, data is collected and analyzed in real time, personalized monitoring thresholds are dynamically adjusted, and personalized monitoring thresholds are improved. Monitoring accuracy and robustness.

Benefits of technology

It realizes high accuracy and effectiveness monitoring of workers' health status, has a wide range of application and good robustness, and can identify changes in sweat volume and electrolyte concentration in real time, improving data processing speed and safe management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118781554B_ABST
    Figure CN118781554B_ABST
Patent Text Reader

Abstract

The present invention discloses a worker health monitoring system using a wearable device in a high temperature environment, comprising: a hyperspectral camera to collect facial images of workers in real time; a sweat feature extraction module to extract sweat area features to calculate real-time sweat volume; a sweat sensor to collect electrolyte concentration data in workers' sweat in real time; an ambient temperature and humidity sensor to collect ambient temperature and humidity in real time; a threshold dynamic adjustment module to dynamically adjust the sweat volume threshold according to the initial sweat volume threshold, ambient temperature and humidity, and to dynamically adjust the sweat volume threshold according to the relationship between the sweat volume threshold and the rate of change of electrolyte concentration data to obtain a personalized monitoring threshold; and to dynamically compare and judge the real-time sweat volume with the personalized monitoring threshold, and output the monitoring result. The present invention dynamically adjusts the personalized monitoring threshold based on hyperspectral images, ambient temperature and humidity, and sweat electrolyte data, thereby improving the monitoring accuracy and robustness, and having higher adaptability to individual differences.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of building safety technology, and in particular to a worker health monitoring system using a wearable device in a high temperature environment. Background Art

[0002] In the construction industry, workers often need to work outdoors in high-temperature environments. Long-term exposure to such conditions can easily lead to heat stress reactions, such as heat stroke and dehydration. Sweating is an important physiological mechanism for the human body to maintain body temperature balance under high temperatures or strenuous exercise. However, there are significant differences in the amount of sweat and electrolyte loss between each person, which mainly depends on the individual's physiological characteristics and environmental conditions. Traditional sweat monitoring methods often use fixed thresholds to determine whether the sweat volume has reached a warning level or whether the electrolytes are unbalanced. This method ignores individual differences, resulting in low monitoring accuracy, limited scope of application, and poor robustness.

[0003] At present, although there are a variety of sweat volume monitoring technologies, they still have some limitations in practical applications. Especially in complex environments such as construction sites, these technologies may not accurately reflect the health status of workers. For example, the existing sweat volume monitoring technologies mainly include the following:

[0004] Optical monitoring: Use light of a specific wavelength to illuminate the sweat area, and determine the amount of sweat by detecting changes in the intensity of the reflected light. It has high requirements for the light source.

[0005] Conductivity monitoring: Measures the electrolyte content in sweat and indirectly assesses sweat volume, with low reference value.

[0006] Weight method: Estimating sweat volume by the change in body weight before and after weighing, which cannot meet the needs of real-time monitoring.

[0007] Image monitoring: Use ordinary cameras to capture changes in sweat areas, but since sweat and skin are similar in color, it is difficult to distinguish them accurately.

[0008] These methods have their own advantages and disadvantages, but their accuracy and applicability are limited when individual differences are large. In addition, traditional methods often cannot achieve real-time monitoring and dynamic adjustment of personalized thresholds.

[0009] Therefore, how to provide a worker health monitoring system using wearable devices in high temperature environments that can effectively improve monitoring robustness and adapt to individual differences is a technical problem that needs to be urgently solved by those skilled in the art. Summary of the invention

[0010] In response to the above research status and existing problems, the present invention provides a worker health monitoring system using a wearable device in a high temperature environment. The system dynamically adjusts the personalized monitoring threshold based on hyperspectral images, ambient temperature and humidity, and sweat electrolyte data, thereby improving the monitoring accuracy and robustness and adapting to individual differences.

[0011] The present invention provides a wearable device worker health monitoring system in a high temperature environment, comprising: a hyperspectral camera, a sweat sensor, a sweat feature extraction module, a threshold dynamic adjustment module, an environmental temperature and humidity sensor, and a monitoring result output module; wherein,

[0012] The hyperspectral camera is fixed to the worker's head through a fixing device to collect the worker's facial image in real time; the sweat feature extraction module obtains the facial image, extracts the sweat area features, and calculates the real-time sweat amount;

[0013] The sweat sensor is disposed on the body surface of the worker in a contact manner, and is used to collect electrolyte concentration data in the worker's sweat in real time;

[0014] The ambient temperature and humidity sensor is used to collect the ambient temperature and humidity within the set range of the worker's environment in real time;

[0015] The threshold dynamic adjustment module is used to dynamically adjust the sweat volume threshold according to the initial sweat volume threshold, the ambient temperature and the ambient humidity, and to dynamically adjust the sweat volume threshold for a second time according to the relationship between the sweat volume threshold and the rate of change of the electrolyte concentration data, so as to obtain a personalized monitoring threshold;

[0016] The monitoring result output module dynamically compares and judges the real-time sweat amount with the personalized monitoring threshold, and outputs the monitoring result.

[0017] Preferably, the sweat feature extraction module includes an image preprocessing unit, a feature extraction unit, and a sweat area segmentation unit; wherein,

[0018] The image preprocessing unit is used to convert the facial image from the RGB color space to the HSV color space and perform a histogram equalization process;

[0019] The feature extraction unit includes a parallel sweat feature extraction network, which extracts sweat features from the facial image before preprocessing and the facial image after preprocessing to obtain a first sweat feature and a second sweat feature;

[0020] The sweat region segmentation unit includes a parallel sweat region segmentation network, which segments the sweat regions of the first sweat feature and the second sweat feature respectively to obtain a first segmented region and a second segmented region;

[0021] Channel weighting is performed on the first segmented area and the second segmented area, and vectors after the weighting of each channel are point-multiplied and then merged to obtain seed points of the sweat area.

[0022] Preferably, the sweat feature extraction network includes a FSANet network and a SSANet network.

[0023] The FSANet network receives the facial image before / after preprocessing, and uses a convolutional neural network to extract basic features of the sweat area; obtains a multi-scale sweat feature map of the facial image before / after preprocessing; uses a dilated convolution to capture information of different scales of the multi-scale sweat feature map; and for each scale feature map, adjusts the weight of each channel according to the interdependence between channels; and obtains a multi-scale sweat feature map by weighted fusion of the feature maps of each scale;

[0024] The SSANet network receives the feature map of each scale, calculates the importance factor of the feature map of each scale using the spatial attention layer, and assigns weights to the feature map of each scale according to the importance factor; restores the resolution of the feature map to obtain the spatial structure information of the sweat area, and outputs the spatial structure feature map.

[0025] The multi-scale sweat feature map and the spatial structure feature map are convolutionally fused using a convolution layer to obtain a first sweat feature / a second sweat feature.

[0026] Preferably, the sweat volume calculation unit has a built-in region growing algorithm and a contour tracking algorithm; wherein,

[0027] The region growing algorithm expands the sweat region according to the position of the seed point in the facial image; the contour tracing algorithm locates the boundary of the sweat region according to the expanded sweat region and calculates the area of ​​the sweat region; and the real-time sweat amount is calculated according to the area of ​​the sweat region and the preset sweat evaporation rate.

[0028] Preferably, calculating the area of ​​the sweat region comprises the following steps:

[0029] ;

[0030] Where: A area is the total area of ​​the sweating area; A i is the area of ​​the ith pixel; N is the total number of pixels in the sweating area.

[0031] Preferably, the threshold dynamic adjustment module dynamically adjusts the sweat threshold according to the initial sweat threshold, the ambient temperature and the ambient humidity, including:

[0032] ;

[0033] Where: T out,adj is the adjusted sweat threshold; T out is the initial sweating threshold; T env and H env are the current ambient temperature and humidity; T base and H base are the reference ambient temperature and humidity respectively; k1 and k2 are the adjustment coefficients.

[0034] Preferably, the threshold dynamic adjustment module performs secondary dynamic adjustment on the sweat volume threshold according to the relationship between the sweat volume threshold and the change rate of the electrolyte concentration data, including:

[0035] ;

[0036] Where: k3 is the adjustment coefficient, T out,alarm,adj is the personalized monitoring threshold; S rate It is the decreasing rate of electrolyte ion concentration in sweat within the time interval △t.

[0037] Preferably, the electrolyte concentration data includes sodium ion concentration data.

[0038] Preferably, it also includes a positioning module and a wireless transmission module arranged on the worker; wherein,

[0039] The positioning module is used to locate the position information of the worker in real time;

[0040] The wireless transmission module wirelessly sends the monitoring result output by the monitoring result output module and the corresponding position information to the wireless transmission modules of other workers;

[0041] The other workers are co-workers whose location information is less than a specified range from the location information of the worker, and the monitoring result is a monitoring alarm result when the real-time sweat volume is higher than the personalized monitoring threshold.

[0042] Preferably, it also includes a power control module, which is used to receive the position information of the positioning module and automatically determine whether the position information overlaps with the preset working area range. If so, it controls the power of the hyperspectral camera, sweat sensor, sweat feature extraction module, threshold dynamic adjustment module, ambient temperature and humidity sensor and monitoring result output module to turn on and enter the working state.

[0043] The wearable device worker health monitoring system in a high temperature environment proposed by the present invention has the following beneficial effects compared with the prior art:

[0044] The present invention proposes a personalized monitoring threshold dynamic adjustment scheme for high-temperature workers. Hyperspectral images and electrolyte concentration data are used to monitor the health status of workers, and individual difference thresholds are dynamically adjusted based on sweat volume and electrolyte concentration, thereby improving the accuracy and effectiveness of monitoring, having a wide range of applications and good robustness.

[0045] The present invention extracts basic features, multi-scale features and spatial structure features of facial images to determine the seed points of the sweat area, has higher recognition accuracy, calculates the sweat amount based on the seed points, and improves the speed of data processing while ensuring accuracy.

[0046] The present invention strengthens on-site safety management and emergency response through location sharing and wireless transmission technology. The automated power management system can automatically turn on equipment according to actual needs, which can avoid the occurrence of inadequate monitoring due to operator omissions, and is safer. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without creative work.

[0048] Figure 1 is a block diagram of a worker health monitoring system of a wearable device in a high temperature environment provided by an embodiment of the present invention;

[0049] Figure 2 is a feature extraction flow chart of a sweat feature extraction module provided by an embodiment of the present invention;

[0050] Figure 3 This is a schematic diagram of a personalized monitoring threshold dynamic adjustment principle provided by an embodiment of the present invention;

[0051] Figure 4 It is a working principle diagram of the wireless transmission module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] like Figure 1As shown, a wearable device worker health monitoring system in a high temperature environment provided by an embodiment of the present invention includes: a hyperspectral camera, a sweat sensor, a sweat feature extraction module, a threshold dynamic adjustment module, an ambient temperature and humidity sensor, and a monitoring result output module; the hyperspectral camera is fixed to the worker's head through a fixing device, and is used to collect the worker's facial image in real time; the sweat feature extraction module obtains the facial image, extracts the sweat area feature, and calculates the real-time sweat volume; the sweat sensor is arranged on the worker's body surface in a contact manner, and is used to collect the electrolyte concentration data in the worker's sweat in real time; the ambient temperature and humidity sensor is used to collect the ambient temperature and ambient humidity within the set range of the worker's environment in real time; the threshold dynamic adjustment module is used to dynamically adjust the sweat volume threshold according to the initial sweat volume threshold, ambient temperature and ambient humidity, and to dynamically adjust the sweat volume threshold for a second time according to the relationship between the sweat volume threshold and the change rate of the electrolyte concentration data to obtain a personalized monitoring threshold; the monitoring result output module dynamically compares and judges the real-time sweat volume with the personalized monitoring threshold, and outputs the monitoring result.

[0054] This embodiment captures the distribution of facial sweat by real-time facial images, thereby assisting in monitoring the amount of sweat. Combined with sweat samples collected directly from the skin surface, the concentration changes of electrolytes (such as sodium ions) are monitored. At the same time, the temperature and humidity of the surrounding environment are monitored in real time to provide a basis for dynamically adjusting the threshold. Based on the initial sweat threshold and environmental parameters (temperature, humidity), the sweat threshold is dynamically adjusted, and the threshold is further optimized according to the rate of change of electrolyte concentration in sweat to achieve personalized monitoring.

[0055] Hyperspectral images can obtain the reflection or emission spectral information of the target object at multiple continuous wavelengths. By utilizing the absorption characteristics of the electrolyte components in sweat at specific wavelengths, it helps to identify sweating areas and extract the spatial distribution characteristics of sweating areas, such as area and location.

[0056] In one embodiment, the sweat feature extraction module includes an image preprocessing unit, a feature extraction unit, and a sweat area segmentation unit; wherein,

[0057] An image preprocessing unit, used to convert the facial image from RGB color space to HSV color space and process it through histogram equalization;

[0058] The feature extraction unit includes a parallel sweat feature extraction network, which performs sweat feature extraction on the facial image before preprocessing and the facial image after preprocessing, respectively, to obtain a first sweat feature and a second sweat feature, which are used to extract sweat features, such as brightness changes, texture features, etc.;

[0059] The sweat region segmentation unit includes a parallel sweat region segmentation network, which segments the sweat regions of the first sweat feature and the second sweat feature respectively to obtain a first segmented region and a second segmented region;

[0060] Channel weighting is performed on the first segmented area and the second segmented area, and the weighted vectors of each channel are point-multiplied and then merged to obtain the seed point of the sweat area.

[0061] It should be noted that the purpose of preprocessing facial images is to convert the images from the RGB color space to the HSV (hue, saturation, value) color space, which is more consistent with human visual perception. The image contrast can be further enhanced by methods such as histogram equalization to make the sweat area easier to identify.

[0062] In one embodiment, the sweat feature extraction network includes an FSANet network and an SSANet network. Considering that facial images are affected by the illumination angle of ambient light and the scale characteristics of sweat are different, it is necessary to optimize the network to pay attention to different scale features.

[0063] In this embodiment, the FSANet network receives facial images before / after preprocessing, and uses a convolutional neural network to extract basic features of the sweat area; obtains a multi-scale sweat feature map of the facial image before / after preprocessing; uses a dilated convolution to capture information of different scales of the multi-scale sweat feature map; and for the feature map of each scale, adjusts the weight of each channel according to the interdependence between channels; obtains the multi-scale sweat feature map by weighted fusion of the feature maps of each scale; the FSANet network includes four parallel convolution branches, each convolution branch includes a 1×1 convolution kernel and a dilated convolution in series; the 1×1 convolution kernel is used to compress the number of input feature channels, and the dilated convolution is used to further compress the number of feature map channels, so as to increase the receptive field while reducing the increase in parameters and computing resource usage caused by the multi-scale large convolution kernel.

[0064] The SSANet network receives feature maps of each scale, uses the spatial attention layer to calculate the importance factor of the feature map of each scale, and assigns weights to the feature maps of each scale according to the importance factor; restores the resolution of the feature map to obtain the spatial structure information of the sweat area, and outputs the spatial structure feature map. The spatial structure attention network includes a 3×3 convolution kernel, and a horizontal pooling layer and a vertical pooling layer connected to the 3×3 convolution kernel at the same time; the 3×3 convolution kernel is used to compress the number of feature map channels, which is used to reduce the number of parameters and computational complexity, accelerate the training and inference process, avoid overfitting, and maintain important spatial structure information; the horizontal pooling layer and the vertical pooling layer are used to extract information in the horizontal and vertical directions of the feature map, respectively, to capture structural features in different directions, and enhance the model's understanding of direction-sensitive image content.

[0065] The convolution layer is used to perform convolution fusion on the multi-scale sweat feature map and the spatial structure feature map to obtain the first sweat feature / the second sweat feature.

[0066] In this embodiment, the sweat feature extraction network matches the collected spectrum with the standard spectrum of known sweat components to determine the sweating area.

[0067] In one embodiment, the sweat region segmentation unit uses the difference between the second sweat feature and the surrounding skin in the HSV color space, identifies the region where sweat exists through an edge detection method, and sets an appropriate threshold to segment the second segmented region.

[0068] In one embodiment, the sweat volume calculation unit embeds a region growing algorithm and a contour tracking algorithm; wherein,

[0069] The region growing algorithm expands the sweat area according to the position of the seed point in the facial image; the contour tracing algorithm locates the boundary of the sweat area according to the expanded sweat area and calculates the area of ​​the sweat area; the real-time sweat amount is obtained according to the area of ​​the sweat area and the preset sweat evaporation rate.

[0070] In one embodiment, calculating the area of ​​the sweat region includes the following steps:

[0071] ;

[0072] Where: A area is the total area of ​​the sweating area; A i is the area of ​​the ith pixel; N is the total number of pixels in the sweating area.

[0073] According to the total area of ​​sweating area A area Calculate the amount of sweat V out :

[0074] ;

[0075] Where: C is the sweat evaporation rate per unit area.

[0076] In one embodiment, before adjusting the threshold, an initial sweat threshold needs to be determined, and this threshold can be set based on the individual's historical data. Next, the sweat threshold is dynamically adjusted according to the ambient temperature and humidity. The threshold dynamic adjustment module dynamically adjusts the sweat threshold according to the initial sweat threshold, ambient temperature and ambient humidity, including:

[0077] ;

[0078] Where: T out,adjis the adjusted sweat threshold; T out is the initial sweating threshold; T env and H env are the current ambient temperature and humidity; T base and H base are the reference ambient temperature and humidity respectively; k1 and k2 are the adjustment coefficients.

[0079] In this embodiment, the threshold dynamic adjustment module performs secondary dynamic adjustment on the sweat volume threshold according to the relationship between the sweat volume threshold and the change rate of the electrolyte concentration data, including:

[0080] ;

[0081] Where: k3 is the adjustment coefficient, T out,alarm,adj is the personalized monitoring threshold; S rate It is the decreasing rate of electrolyte ion concentration in sweat within the time interval △t.

[0082] In practical applications, the adjustment coefficients are determined based on individual differences and experimental data, and these threshold adjustment formulas can be embedded in the monitoring system's algorithm to make adjustments in real time based on environmental conditions and personal data.

[0083] In one embodiment, the electrolyte concentration data includes sodium ion concentration data. The concentration of sodium in normal sweat is usually higher than that of potassium. If the ratio is unbalanced, there may be an electrolyte imbalance. In a high temperature environment, the concentration of sodium in sweat may decrease due to excessive sweating. Therefore, comparing the electrolyte concentration in sweat with the normal reference range is an effective means of assessing whether there is an electrolyte imbalance.

[0084] In one embodiment, it also includes a positioning module and a wireless transmission module arranged on the worker; wherein,

[0085] The positioning module is used to locate the workers’ location information in real time;

[0086] The wireless transmission module wirelessly transmits the monitoring results and corresponding position information output by the monitoring result output module to the wireless transmission modules of other workers;

[0087] Other workers are co-workers whose location information is less than a specified range from the worker's location information, and the monitoring result is the monitoring alarm result when the real-time sweat volume exceeds the personalized monitoring threshold.

[0088] In one embodiment, it also includes a power control module, which is used to receive the location information of the positioning module and automatically determine whether the location information overlaps with the preset working area range. If so, it controls the power of the hyperspectral camera, sweat sensor, sweat feature extraction module, threshold dynamic adjustment module, ambient temperature and humidity sensor and monitoring result output module to turn on and enter the working state.

[0089] The above is a detailed introduction to a worker health monitoring system of a wearable device in a high temperature environment provided by the present invention. This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0090] In this article, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

Claims

1. A wearable device worker health monitoring system in a high temperature environment, characterized in that: include: Hyperspectral camera, sweat sensor, sweat feature extraction module, threshold dynamic adjustment module, ambient temperature and humidity sensor and monitoring result output module; among them, The hyperspectral camera is fixed to the worker's head through a fixing device to collect the worker's facial image in real time; the sweat feature extraction module obtains the facial image, extracts the sweat area features, and calculates the real-time sweat amount; The sweat sensor is disposed on the body surface of the worker in a contact manner, and is used to collect electrolyte concentration data in the worker's sweat in real time; The ambient temperature and humidity sensor is used to collect the ambient temperature and humidity within the set range of the worker's environment in real time; The threshold dynamic adjustment module is used to dynamically adjust the sweat volume threshold according to the initial sweat volume threshold, the ambient temperature and the ambient humidity, and to dynamically adjust the sweat volume threshold for a second time according to the relationship between the sweat volume threshold and the rate of change of the electrolyte concentration data, so as to obtain a personalized monitoring threshold; The monitoring result output module dynamically compares and judges the real-time sweat amount with the personalized monitoring threshold, and outputs the monitoring result.

2. A wearable device worker health monitoring system in a high temperature environment according to claim 1, characterized in that: The sweat feature extraction module includes an image preprocessing unit, a feature extraction unit, and a sweat area segmentation unit; wherein, The image preprocessing unit is used to convert the facial image from the RGB color space to the HSV color space and perform a histogram equalization process; The feature extraction unit includes a parallel sweat feature extraction network, which extracts sweat features from the facial image before preprocessing and the facial image after preprocessing to obtain a first sweat feature and a second sweat feature; The sweat region segmentation unit includes a parallel sweat region segmentation network, which segments the sweat regions of the first sweat feature and the second sweat feature respectively to obtain a first segmented region and a second segmented region; Channel weighting is performed on the first segmented area and the second segmented area, and vectors after each channel weighting are point-multiplied and then merged to obtain seed points of the sweat area.

3. A wearable device worker health monitoring system in a high temperature environment according to claim 2, characterized in that: The sweat feature extraction network includes an FSANet network and an SSANet network. The FSANet network receives the facial image before / after preprocessing and uses a convolutional neural network to extract basic features of the sweat area; Obtain a multi-scale sweat feature map of the facial image before / after preprocessing; use a dilated convolution to capture information of different scales of the multi-scale sweat feature map; and for each scale of the feature map, adjust the weight of each channel according to the interdependence between the channels; A multi-scale sweat feature map is obtained by weighted fusion of feature maps at each scale; The SSANet network receives the feature map of each scale, calculates the importance factor of the feature map of each scale using the spatial attention layer, and assigns weights to the feature maps of each scale according to the importance factor; Restore the resolution of the feature map to obtain the spatial structure information of the sweat area, and output the spatial structure feature map; The multi-scale sweat feature map and the spatial structure feature map are convolutionally fused using a convolution layer to obtain a first sweat feature / a second sweat feature.

4. The worker health monitoring system using a wearable device in a high temperature environment according to claim 1, characterized in that: The sweat volume calculation unit has a built-in region growing algorithm and a contour tracking algorithm; wherein, The region growing algorithm expands the sweat region according to the position of the seed point in the facial image; the contour tracing algorithm locates the boundary of the sweat region according to the expanded sweat region and calculates the area of ​​the sweat region; and the real-time sweat amount is calculated according to the area of ​​the sweat region and the preset sweat evaporation rate.

5. The worker health monitoring system using a wearable device in a high temperature environment according to claim 4, characterized in that: Calculating the area of ​​the sweat zone includes the following steps: ; Where: A area is the total area of ​​the sweating area; A i is the area of ​​the ith pixel; N is the total number of pixels in the sweating area.

6. The worker health monitoring system using a wearable device in a high temperature environment according to claim 1, characterized in that: The threshold dynamic adjustment module dynamically adjusts the sweat threshold according to the initial sweat threshold, the ambient temperature and the ambient humidity, including: ; Where: T out,adj is the adjusted sweat threshold; T out is the initial sweating threshold; T env and H env are the current ambient temperature and humidity; T base and H base are the reference ambient temperature and humidity respectively; k1 and k2 are the adjustment coefficients.

7. A wearable device worker health monitoring system in a high temperature environment according to claim 6, characterized in that: The threshold dynamic adjustment module performs secondary dynamic adjustment on the sweat volume threshold according to the relationship between the sweat volume threshold and the change rate of the electrolyte concentration data, including: ; Where: k3 is the adjustment coefficient, T out,alarm,adj is the personalized monitoring threshold; S rate It is the decreasing rate of electrolyte ion concentration in sweat within the time interval △t.

8. The worker health monitoring system using a wearable device in a high temperature environment according to claim 1, characterized in that: The electrolyte concentration data includes sodium ion concentration data.

9. The worker health monitoring system using a wearable device in a high temperature environment according to claim 1, characterized in that: It also includes a positioning module and a wireless transmission module arranged on the worker; wherein, The positioning module is used to locate the position information of the worker in real time; The wireless transmission module wirelessly sends the monitoring result output by the monitoring result output module and the corresponding position information to the wireless transmission modules of other workers; The other workers are co-workers whose location information is less than a specified range from the location information of the worker, and the monitoring result is a monitoring alarm result when the real-time sweat volume is higher than the personalized monitoring threshold.

10. A wearable device worker health monitoring system in a high temperature environment according to claim 9, characterized in that: It also includes a power control module, which is used to receive the position information of the positioning module and automatically determine whether the position information overlaps with the preset working area range. If so, it controls the power of the hyperspectral camera, sweat sensor, sweat feature extraction module, threshold dynamic adjustment module, ambient temperature and humidity sensor and monitoring result output module to turn on and enter the working state.

Citation Information

Patent Citations

  • Intelligent terminal based health monitoring method and device

    CN107595263A

  • Physiological parameter prewarning method and system based on wearable equipment

    CN109498031A

  • Motion monitoring method and device, wearable equipment and storage medium

    CN115670444A

  • Head-mounted device for monitoring labor heatstroke and early warning method

    CN117612329A

  • Method and system for touch control of electrical stimulation

    CN118543032A