Procedure for determining the health burden of a user of the procedure

Abstract

The invention relates to a method (100) for reducing the health impact of a user (14) due to their posture (16) and the lighting conditions at the user's (14) workplace (48). Within the scope of the method (100), a partial area of ​​the surface of the user's (14) body (18) is detected by an optical sensor system (28) of a strain gauge (10). Geometric parameters of the detected partial area are then determined. From the determined geometric parameters, a neural network (36) of the strain gauge (10) determines characteristic parameter values ​​of the user's (14) posture (16), which define the user's (14) posture (16).In addition, characteristic parameter values ​​of the electromagnetic radiation (25) incident on the optical sensor system (28) are determined, wherein the characteristic parameter values ​​are particularly suitable for determining the radiation energy of the electromagnetic radiation incident on the optical sensor system (28). The posture (16) and the characteristic parameter values ​​of the electromagnetic radiation (25) are assigned to a stress level that indicates the stress exerted on the user (14) by the posture (16) and the electromagnetic radiation (25).

Description

Background of the invention

[0001] The invention relates to a method for determining the health stress of a user of the method, particularly at the user's workplace. The invention further relates to a stress meter for implementing the method.

[0002] Methods for reducing physical and psychological stress at the workplace are known from the state of the art, particularly in the field of ergonomics.

[0003] DE 10 2022 201 925 A1 relates to a method for determining a user's relaxation level using a camera. The method takes into account the influences of the user's work environment.

[0004] DE 10 2020 207 975 A1 discloses a method for reducing a user's health burden, in which a sensor determines coordinates of selected key points of the user's musculoskeletal system. These coordinates are used to determine the user's posture and derive recommendations for correcting posture.

[0005] US 2016 / 0183687 A1 concerns sensors mounted on furniture, particularly seating, to determine a user's posture. The sensors are used to determine, in particular, forces acting on various parts of the seating, as well as movements and positions of the seating.

[0006] US 2009 / 0324024 A1 discloses a system and a method for correcting a posture of a user of the system, wherein a camera regularly captures an image of the user. The user's face is identified in the respective image and compared with a previously determined reference image of the face in a posture classified as good. If the user repeatedly assumes a posture classified as good, they are given appropriate feedback. In particular, the user is also given feedback if they repeatedly assume a posture classified as bad or consistently over a specified period of time.

[0007] DE 10 2022 202 729 A1 relates to a method for determining a person's posture, in which the posture recognizable in a camera image is characterized using a value. If this value lies within a warning range, an alert signal is issued.

[0008] However, these procedures are carried out with a time delay compared to the time at which the respective stress occurs and without objective and precise information on the stress in question. Object of the invention

[0009] The object of the invention is to provide a method with which the load of a user of the method can be determined promptly and precisely. It is also an object of the invention to provide a device for implementing the method. Description of the invention

[0010] This object is achieved according to the invention by a method for determining the health burden of a user of the method, which comprises the following steps: I. Detecting a surface of a part of the user's body by the optical sensor system in the measuring interval; II. Determining a posture of the user from pixel parameter values of the detected surface by the neural network, wherein the pixels are designed as elements of the optical sensor system, wherein step II is carried out after step I; III. Determining characteristic parameter values of the light received by the optical sensor system in the measurement interval; IV. Classifying the posture of the user and classifying the received light with respect to a load on the user, wherein step IV is performed after steps I - III; V. Outputting the classification of the body posture and the received light, where step V is performed after step IV.

[0011] Advantageously, the user's posture can be determined promptly, precisely, and objectively using the optical sensor and the neural network. This allows the user to receive quick and accurate feedback on how to correct their posture in light of the physical strain.

[0012] Posture classification can be based on the pixel values themselves, e.g., by assigning groups or patterns of pixel values to specific postures. Alternatively or additionally, geometric parameter values, such as curvatures of the user's body or the angle of inclination of the user's body, can be determined to classify the posture. Reference points on the user's body can be used for this purpose. Alternatively or additionally, a user's posture can be classified by comparing it with a reference posture.

[0013] The optical sensor system preferably comprises a camera. Alternatively or additionally, the optical sensor system is configured with a time-of-flight sensor. When recording the surface of the part of the user's body, parameters of pixels of the optical sensor assume specific parameter values. The parameter values can, in particular, be grayscale values of the pixels.

[0014] The neural network is implemented specifically within the strain gauge. Alternatively, the neural network can be located outside the strain gauge and exchange data with the strain gauge via a communication unit.

[0015] For the purposes of this application, the term "light" refers specifically to the range of the electromagnetic spectrum perceptible to humans. The light detected in accordance with step III originates, in particular, from the environment of the user's workplace. The sensors mentioned in the application can be configured to detect other ranges of the electromagnetic spectrum.

[0016] The characteristic parameter values of the light are, in particular, parameter values that are necessary to determine the irradiance of the light with which the light irradiates the optical sensor system. In particular, the characteristic parameter values of the light include the intensity of the light and / or the frequency of the light. Alternatively or additionally, characteristic parameter values of electromagnetic radiation outside the optical spectrum can also be determined.

[0017] The classification of light with regard to user exposure can be achieved using a deviation of the light's characteristic parameter values from reference parameter values. Alternatively or additionally, the light's characteristic parameter values can be assigned to a exposure class, for example.

[0018] The determined classifications are used within the process, in particular, to create a risk analysis. In particular, the precise classifications can be used to determine the causes of stress for a user of the process (such as stressful light exposure or factors that lead to a stressful posture). These causes can be used in the risk analysis.

[0019] The following steps are carried out within the scope of the method according to the invention: VI. Detection of a user's workstation by the optical sensor system in the measurement interval; VII. Determining geometric parameter values of the user’s workplace; VIII. Classify the user's workplace using the geometric parameters in relation to a user's load.

[0020] The geometric parameters of the user's workstation include, in particular, the geometric parameter values of the work surface. The geometric parameters of the work surface include, in particular, the height, width, and length of the work surface. The geometric parameters of the workstation preferably include geometric parameters of the user's work equipment, for example, a computer screen of the user, and / or geometric parameters of the user's movement space at their workstation. This configuration can advantageously reduce the strain on the user caused by the design and location of the work surface.

[0021] In particular, the method can determine whether a user's posture, which leads to a relatively high load on the user, is related to the geometric parameter values of the work surface. This can be used to identify a work surface height that is unfavorable with respect to the load. For example, the work surface is designed as part of a table.

[0022] In a preferred embodiment of the method, the strain gauge has an acoustic sensor system, wherein the following steps are carried out: IX. Detection of sound waves by the acoustic sensor system in the measurement interval; X. Determining characteristic parameter values of sound waves; XI. Classify the sound waves using the characteristic parameter values of the sound waves in relation to a user load.

[0023] The characteristic parameter values of the sound waves include, in particular, parameter values from which the energy transported by the respective sound wave can be determined. These include, among other things, the intensity and / or frequency of the respective sound wave. The sound waves can be classified using reference values for the respective parameters. This embodiment of the method can advantageously reduce the user's exposure to noise. In some embodiments, the strain gauge includes a transmitter for ultrasonic waves, and the acoustic sensor system includes a sensor for detecting the sound waves. Measurements of the ultrasonic waves can be used, in particular, to measure distances and heights (e.g., of the work surface).

[0024] It is advantageous to design the procedure in which the following steps are carried out: XII. Capturing light signals to determine the pulse and / or respiration and / or heart rate variability of the user by the optical sensor system in the measurement interval; XIII. Determining characteristic parameter values of the pulse and / or respiration and / or heart rate variability; XIV. Classifying the pulse and / or respiration and / or heart rate variability using the characteristic parameter values of the pulse and / or respiration and / or heart rate variability in relation to a user load.

[0025] In particular, light reflected by blood vessels (and its spectrum) is detected. From this light and its spectrum, the characteristic parameter values for determining the pulse and / or respiration and / or heart rate variability, in particular the pulse rate or respiration rate, are determined. The sound waves can be classified using reference values for the relevant parameters. This design advantageously allows the user's stress exposure to be displayed and reduced.

[0026] Furthermore, an embodiment of the method is preferred in which the following steps are carried out: XV. Detecting the temperature, humidity and / or material composition of the air at the load cell by means of a respective sensor system of the load cell; XVI. Determination of characteristic parameter values of temperature, humidity and / or the material composition of the air; XVII. Classifying the temperature, humidity and / or material composition using the characteristic parameter values of the temperature, humidity and / or material composition of the air in relation to a user load.

[0027] The classification of temperature, humidity, and / or the material composition of the air can be performed using reference values for the relevant parameters. In particular, the CO2 content in the air constitutes a portion of the material composition of the air. VOC sensors, CO2 sensors, and / or temperature sensors are used, in particular, to determine the properties of the air. This design can advantageously reduce the user's exposure to poor air quality or an unfavorable temperature in the workplace.

[0028] It is advantageous to design the procedure in which the following steps are carried out: XVIII. Recording the length of time the user maintains a posture; XIX. Classifying posture and duration in relation to user load.

[0029] The classification is carried out, in particular, using reference values for the strain for periods of time in which a specific posture is assumed. This design can advantageously reduce strain on the user caused by excessively long periods of time in a posture. In particular, the user can be prompted to adopt a different posture or perform a specific movement sequence as part of the output.

[0030] A further embodiment of the method is characterized in that the optical sensor system includes a camera. The camera is preferably a digital camera. A camera can precisely capture the user's posture.

[0031] In the method according to the invention, the optical sensor system includes a radar. With the help of the radar, distances can be measured very precisely, which enables precise determination of the user's posture and movements.

[0032] A preferred embodiment of the method is one in which the optical sensor system has a time-of-flight sensor, in particular for measuring in three dimensions. A time-of-flight sensor can be used to record the surface of the user's body in a time-saving manner by emitting a light pulse and the light reflected from various points on the surface being detected by the time-of-flight sensor. The time-of-flight sensor can be designed to emit and / or detect electromagnetic radiation in the infrared range. In some embodiments, the optical sensor system has a one-dimensional (1D) time-of-flight sensor for measuring distances in one dimension. This makes it possible, for example, to measure the height of a work surface, in particular by determining a distance to the ceiling.

[0033] A further preferred embodiment of the method is one in which the load cell encodes the data measured or determined during the method, in particular in the form of a graphical encoding, preferably in the form of a QR code. This allows the data to be easily encrypted to ensure data protection. In particular, a risk analysis can be encoded after its creation in the form of a graphical encoding, preferably in the form of a QR code, as part of the determined data.

[0034] In an advantageous embodiment of the method, the load cell transmits the data measured or determined during the method. This allows the data (in particular, including a risk analysis) to be made available promptly for further evaluation by a corresponding central evaluation device.

[0035] It is advantageous to design the method in which the neural network is trained to recognize a body posture according to the following steps: i. Determining target parameters that characterize a test posture of a test person; ii. Initializing weights of the neural network; iii. detecting a surface of a part of the body of the test person in the test posture by the optical sensor system in the measuring interval; iv. Determining pixel parameter values of the detected surface and weights of the pixel parameter values; v. Determining a difference between the weighted parameters and the target parameters; vi. Adjusting the weights to reduce the difference; vii. Repeat steps v to vi until a minimization limit is reached;

[0036] The target parameters can be one or more curvatures of the user's body in the test posture, or angles of inclination of the user's body in the test posture. Specifically, the neural network is trained to assign the pixel parameter values to a specific set of target parameters that characterize the test posture.

[0037] A strain gauge according to the invention is configured to perform the method according to one of the aforementioned embodiments. Such a strain gauge can quickly and objectively determine the strain of a user of the strain gauge.

[0038] Further advantages of the invention will become apparent from the description and the drawings. Likewise, the above-mentioned and further-described features can be used individually or in combinations according to the invention. The embodiments shown and described are not intended to be exhaustive, but rather are exemplary in nature for describing the invention. Detailed description of the invention and drawing Fig. 1 shows schematically a side view of a strain gauge according to the invention at a workstation of a user of the strain gauge. Fig. 2 schematically shows a method according to the invention for reducing the health burden on the user.

[0039] Fig. 1 schematically shows a side view of a strain gauge 10 according to the invention, which is positioned on a workstation 48 in the form of a table 12, at which a user 14 of the strain gauge 10 is located. The user 14 is bent over the table 12 in a posture 16, so that the body 18 of the user 14, in particular the upper body as part of the body 18 of the user 14, has a curvature on its surface. For classifying the posture 16, for example, reference points 20a, 20, 20c of the body 18 of the user 14 can be used. In the Fig. In the example shown in Figure 1, an inclination angle NW is formed between two rays S1, S2, which pass through such reference points 20a, 20b, 20c of the body 18, due to the body posture 16. For example, the first reference point 20a and the second reference point 20b are located on the spine of the user 14, and the third reference point 20c is located on the face, in particular on the nose of the user 14.

[0040] In the embodiment shown, the strain gauge 10 emits light waves 24, in particular a light pulse, via an optional light emission unit 22. The light waves 24 are reflected at least by the body 18 (in particular the upper body) of the user 14 and, in the form of reflected light waves 25, reach a sensor device 26 of the strain gauge 10, which has an optical sensor system 28. The light emission unit 22 and the sensor device 26 are preferably integrated into one component (not shown). In the embodiment shown, the optical sensor system 28 has a camera 30. Alternatively or additionally, the optical sensor system is configured with a time-of-flight sensor 32. In some embodiments, the optical sensor system 28 is configured with a radar 50. The optical sensor system 28 serves to record the body 18 of the user 14.In this case, parameters of pixels (not shown) of the optical sensor system 28 assume specific parameter values. The parameter values can, in particular, be grayscale values of the pixels. The image of the body 18 of the user 14 can also be captured using light that originates from the environment of the strain gauge 10 and is reflected by the body 18 of the user 14. This depends in particular on the optical sensor system 28 used. The image of the body 18 of the user 14 is captured in a measurement interval for determining the posture 16 of the user 14.

[0041] A neural network 36 is implemented in a control unit 34 of the strain gauge 10, which determines the posture 16 of the user 14 from the pixel parameter values. The control unit 34 and the optical sensor system 28 are also configured to determine characteristic parameter values of the light received by the optical sensor system 28 during the measurement interval. The characteristic parameter values are, in particular, parameter values necessary to determine the irradiance of the light with which the light irradiates the optical sensor system 28. In particular, the characteristic parameter values of the light include the intensity of the light and / or the frequency of the light.

[0042] The control unit 34 is further configured to classify the body posture 16 of the user 14 and the light 25 received by the optical sensor system 28 with respect to a load on the user 14. The classification of the body posture 16 is carried out in particular by a comparison with a reference body posture (not shown), preferably by determining differences between the positions of the reference points of the body 18 of the user 14 in its body posture 16 and the positions of the reference points of the body 18 of the user 14 in the reference body posture. An alternative or additional possibility for classifying the body posture 16 of the user 14 consists in determining values of geometric parameters of the body posture 16 of the user 14, such as curvatures of the body 18 or the aforementioned angle of inclination NW, and assigning these values to a load class.An alternative or additional way of classifying the posture 16 is to use the pixel parameter values themselves to assign the posture 16 to a load value.

[0043] In addition, the control unit 34 is configured to classify light 52 irradiated by the optical sensor system during the measurement interval with respect to the exposure of the user 14. The light 52 originates, in particular, from the environment of the workstation 48 of the user 14. The classification of the light can be performed using a deviation of the characteristic parameter values of the light from reference parameter values of the light. Alternatively or additionally, the characteristic parameter values of the light can be assigned, for example, to a exposure class.

[0044] An output unit of the load meter 10 serves to output the load of the user 14 due to his posture 16 and the light irradiated by the optical sensor system 28. Thus, the user 14 receives feedback about his load.

[0045] The control unit 34 is preferably configured to encode the data measured or determined within the scope of the method, in particular in the form of a graphical-image coding, preferably in the form of a QR code. The measured or determined data regarding the load of the user 14 can be transmitted to a central data device (not shown) by an optional transmission unit 38, wherein the transmission unit 38 is configured in particular for sending and receiving data. The data is transmitted in particular in the form of graphical-image coding.

[0046] In some embodiments, the strain gauge 10 is configured to detect geometric parameters of the workstation 48, in particular a work surface 40, here in particular the table surface. The geometric parameters include, in particular, the width, length, and height of the work surface 40. For this purpose, the strain gauge is preferably positioned at locations such that the optical sensor system 28 can receive light for measuring the corresponding distances (length, width, height). For this purpose, the light emission unit 22 preferably emits light pulses, which are reflected by the work surface 40 and then detected by the optical sensor system 28. In the aforementioned embodiments, the control unit 34 is configured to classify the determined geometric parameters of the work surface 40 with respect to a load of the user 14, for example, using the deviation from reference values of the geometric parameters of the work surface 40.The optical sensor system 28 may comprise a one-dimensional time-of-flight sensor 32a, which is configured to measure distances, preferably heights, of the work surface 40 by transmitting electromagnetic waves 24a (particularly in the vertical direction). Alternatively or additionally, the optical sensor system 28 may comprise a radar 50.

[0047] In the embodiment shown, the strain gauge 10 has an acoustic sensor system 42. The acoustic sensor system 42 can detect sound waves 44 that impinge on the acoustic sensor system 42 during the measurement interval. The control unit 34 is designed to determine characteristic parameter values of the sound waves 44 detected by the acoustic sensor system 42. The characteristic parameter values relate in particular to values of parameters with which the power of the sound waves 44 impinging on the acoustic sensor system 42 can be determined. In particular, the characteristic parameter values include the respective intensity and / or frequency of the sound waves 44.In the embodiment shown, the control unit 34 is configured to classify the characteristic parameters of the sound waves 44 with respect to a load on the user 14, for example, using the deviation from reference values of the characteristic parameters of the sound waves 44.

[0048] In some embodiments, the optical sensor system 28 is designed to detect light signals for determining a pulse and / or respiration and / or heart rate variability of the user in the measurement interval. In particular, the optical sensor system 28 is designed to detect light 25 reflected by blood vessels (in particular its spectrum). In these embodiments, the control unit 34 is designed to determine characteristic parameter values of the pulse and / or respiration and / or heart rate variability. The characteristic parameter values include, in particular, the pulse frequency or the oxygen content of the blood. In these embodiments, the control unit 34 is further designed to classify the characteristic parameter values of the blood pulse and / or the oxygen saturation in relation to a load of the user 14, for example, using the deviation from reference values of the pulse orof respiration and / or heart rate variability.

[0049] In the embodiment shown, the load cell 10 has a further sensor system 46, which is designed to detect the temperature, the humidity and / or the material composition of the air at the load cell 10. In this case, components 47 of the air in Fig. 1 is schematically indicated by filled circles. In the illustrated embodiment of the stress meter 10, the control unit 34 is configured to classify the temperature, humidity, and / or the material composition of the air in relation to the stress of the user 14.

[0050] In the embodiment shown, the load meter 10, in particular the control unit 34 of the load meter 10, is configured to record the duration of time during which the user 14 assumes a posture 16. Furthermore, in the embodiment shown, the control unit 34 is configured to classify the posture 16 and the duration of time with respect to the load on the user 14. The classification is carried out, in particular, using reference values of the load for durations during which a specific posture 16 is assumed.

[0051] Fig. 2 schematically shows a method 100 according to the invention for reducing the health burden of a user 14 (see Fig.1) during a measurement interval by a strain gauge 10 with an optical sensor system 28 and a neural network 36. Within the scope of the method 100, in a first step 102, the surface of a part of the body 18 of the user 14 is first detected by the optical sensor system 28 during the measurement interval. In a second step 104, the body posture 16 of the user 14 is determined from pixel parameter values of the detected surface by the neural network 36. The pixels are designed as elements of the optical sensor system 28. In a third step 106, characteristic parameter values of the light received by the optical sensor system 28 during the measurement interval are determined. The characteristic parameters whose values are determined are suitable for calculating an irradiance of the light incident on the optical sensor system 28.In a fourth step 108, the posture 16 and the received light are classified with respect to a load of the user 14. In a fifth step 110, the classification of the posture 16 and the received light is output to inform the user 14 about his load.

[0052] The first step 102 and the second step 104 follow one another. The third step 106 can be performed in parallel with the first step 102 and the second step 104. The fourth and fifth steps 108, 110 follow one another and after the first through third steps 102-106.

[0053] Taking all the figures of the drawing together, the invention relates to a method 100 for reducing the health stress of a user 14 caused by their posture 16 and the lighting conditions at the user's 14 workplace 48. Within the scope of the method 100, a partial area of the surface of the user's 14 body 18 is detected by an optical sensor system 28 of a strain gauge 10. Geometric parameters of the detected partial area are then determined. From the determined geometric parameters, a neural network 36 of the strain gauge 10 determines characteristic parameter values of the user's 14 posture 16, which define the user's 14 posture 16.In addition, characteristic parameter values of the electromagnetic radiation 25 incident on the optical sensor system 28 are determined, wherein the characteristic parameter values are particularly suitable for determining the radiation energy of the electromagnetic radiation incident on the optical sensor system 28. The body posture 16 and the characteristic parameter values of the electromagnetic radiation 25 are assigned to a stress level that indicates the stress placed on the user 14 by the body posture 16 and the electromagnetic radiation 25. List of reference symbols 10 strain gauges 12 table 14 users 16 Posture 18 bodies 20a-20c Reference points 22 Light emission unit 24 emitted light waves 25 reflected light waves 26 Sensor device 28 optical sensor system 30 Camera 32 TOF sensors 34 Control unit 36 neural network 38 Transmission unit 40 work surface 42 acoustic sensor system 44 sound waves 46 additional sensor system 47 air components 48 workplaces 50 radars 52 Light irradiating the optical sensor system 100 inventive method 102 -110 steps of the procedure S1, S2 rays NW inclination angle

Claims

[1] Method (100) for determining the health stress of a user (14) during a measurement interval by means of a stress meter (10) with an optical sensor system (28) and a neural network (36), wherein the optical sensor system (28) has a radar (50), comprising the steps: I. detecting a surface of a part of the body (18) of the user (14) by the optical sensor system (28) in the measuring interval; II. Determining a posture (16) of the user (14) from pixel parameter values of the detected surface by the neural network (36), wherein the pixels are designed as elements of the optical sensor system (28), wherein step II is carried out after step I; III. Determining characteristic parameter values of the light (25) received by the optical sensor system (28) in the measurement interval; IV. Classifying the posture (16) of the user (14) and classifying the received light (25) with respect to a load on the user (14), wherein step IV is carried out after steps I - III; V. Outputting the classification of the posture (16) and the received light (25), wherein step V is performed after step IV; VI. Detecting a workstation (48) of the user (14) by the optical sensor system (28) in the measuring interval; VII. Determining geometric parameters of the workstation (48) of the user (14); VIII. Classifying the workstation (48) of the user (14) using the geometric parameters with respect to a load on the user (14). [2] Method according to one of the preceding claims, in which the strain gauge (10) has an acoustic sensor system (42) and the following steps are carried out: IX. Detecting sound waves (44) by the acoustic sensor system (42) in the measuring interval; X. Determining characteristic parameter values of the sound waves (44); XI. Classifying the sound waves (44) using the characteristic parameter values of the sound waves (44) with respect to a load on the user (14). [3] Method according to one of the preceding claims, in which the following steps are carried out: XII. Detecting light signals (25) for determining the pulse and / or respiration and / or heart rate variability of the user (14) by the optical sensor system (28) in the measuring interval; XIII. Determining characteristic parameter values of the pulse and / or respiration and / or heart rate variability; XIV. Classifying the pulse and / or respiration and / or heart rate variability using the characteristic parameter values of the pulse and / or respiration and / or heart rate variability in relation to a user load (14). [4] Method according to one of the preceding claims, in which the following steps are carried out: XV. Detecting the temperature, the humidity and / or the material composition of the air at the load cell by a respective sensor system (46) of the load cell (10); XVI. Determination of characteristic parameter values of temperature, humidity and / or the material composition of the air; XVII. Classifying the temperature, humidity and / or material composition using the characteristic parameter values of the temperature, humidity and / or material composition of the air in relation to a load on the user (14). [5] Method according to one of the preceding claims, in which the following steps are carried out: XVIII. Recording the length of time during which the user (14) assumes a posture (16); XIX. Classifying posture (16) and duration of user load (14). [6] Method according to one of the preceding claims, wherein the optical sensor system (28) comprises a camera (30). [7] Method according to one of the preceding claims, wherein the optical sensor system (28) comprises a time-of-flight sensor (32, 32a), in particular for measuring in three dimensions. [8] Method according to one of the preceding claims, wherein the load cell (10) encodes the data measured or determined within the scope of the method, in particular in the form of a graphic-pictorial coding, preferably in the form of a QR code. [9] Method according to one of the preceding claims, wherein the strain gauge (10) sends the data measured or determined within the scope of the method. [10] Method according to one of the preceding claims, wherein the neural network (36) is trained to recognize a body posture (16) according to the following steps: i. Determining target parameters that characterize a test posture (16) of a test person; ii. Initializing weights of the neural network (36); iii. detecting a surface of a part of the body (18) of the test person in the test posture (16) by the optical sensor system in the measuring interval; iv. Determining pixel parameter values of the detected surface and weights of the pixel parameter values; v. Determining a difference between the weighted parameters and the target parameters; vi. Adjusting the weights to reduce the difference; vii. Repeat steps v to vi until a minimization limit is reached; [11] Strain gauge (10) for carrying out the method (100) according to one of claims 1 to 10.

Images