Lifting table height adjusting method based on sitting posture recognition

Abstract

The invention discloses a lifting table height adjusting method based on sitting posture recognition. The method comprises the following steps: data acquisition and preprocessing; performing feature extraction on the collected data; carrying out sitting posture monitoring and analysis; identifying a sitting posture health state; setting a lifting table adjusting mechanism; identifying the health state level of the day; and health trend analysis. According to the method, through automatic sitting posture recognition and height adjustment, the user can be helped to realize convenient and continuous health management, so that health problems related to sedentariness are effectively prevented and reduced; through the computer vision technology and the man-machine interaction technology, the sitting posture state of the user is recognized, the specific health level of the sitting posture is further judged, refined health evaluation can provide more personalized health suggestions and intervention measures for the user, and the health level of the user is improved; and addition and subtraction processing is performed according to factors such as the health level, the rising and falling time and the frequency, so that too frequent or untimely adjustment is avoided, and the office experience and health of the user are improved.

Description

Technical Field

[0001] The invention belongs to the field of office furniture applications, and in particular relates to a method for adjusting the height of a lifting table based on sitting posture recognition. Background Art

[0002] In modern society, office furniture, especially height-adjustable tables, have become an indispensable part of people's daily lives. With the advancement of science and technology, people have begun to explore how to apply computer vision technology and human-computer interaction technology to the design and improvement of office furniture to improve people's office experience and health level. Among them, sitting posture recognition technology, as an emerging technical means, can effectively assess people's health status by capturing and analyzing people's sitting postures, and provide a basis for the automatic adjustment of height-adjustable tables.

[0003] In the existing technology, some methods for adjusting the height of lifting tables based on sitting posture recognition have been proposed. These methods usually collect the user's sitting posture data through a camera or other sensor devices, and then input these data into a computer vision system for analysis and processing to identify whether the user's sitting posture meets the health standard. If a bad sitting posture is identified, the height of the table can be adjusted by controlling the motor of the lifting table to drive it to rise and fall, thereby prompting the user to adopt a healthier sitting posture. However, the existing technical solutions have the following technical defects:

[0004] Defect 1: The existing solutions mainly identify basic sitting postures, such as whether the sitting posture is correct or hunched, but cannot make accurate judgments on the specific health level of the sitting posture;

[0005] Defect 2: After identifying the user's bad sitting posture, the existing solution usually simply adjusts the height of the lifting table without considering factors such as the frequency and time of adjustment;

[0006] Defect three: When processing sitting posture data, the correlation and trend between data are not considered. Summary of the invention

[0007] In order to make up for the shortcomings of the prior art, the present invention provides a height adjustment method for a lifting table based on sitting posture recognition. By introducing data conversion and machine learning algorithms, the sitting posture health level of users at different stages can be accurately identified, and the lifting time and frequency of the lifting table can be intelligently adjusted according to the user's sitting posture change trend and health level, thereby achieving more scientific and humane adjustment.

[0008] The technical problem solved by the present invention can be achieved through the following specific technical solutions:

[0009] The method for adjusting the height of a lifting table based on sitting posture recognition comprises the following steps:

[0010] S1. Data collection and preprocessing;

[0011] S2. Extract features from the collected data;

[0012] S3. Perform sitting posture monitoring and analysis;

[0013] S4. Identify the sitting posture health status;

[0014] S5. Set up the lifting table adjustment mechanism;

[0015] S6. Identify the health status level of the day;

[0016] S7. Health trend analysis.

[0017] Furthermore, the specific process of step S1 is: collecting data through radar sensors and pressure sensors, and fusing the acquired basic information; the radar sensor is used to capture information between the human body and the lifting table; the pressure sensor is used to obtain pressure distribution information on the seat.

[0018] Furthermore, in step S2, the angle features of the body are extracted from the radar sensor data, and the statistical values ​​of the sitting posture features within a period of time are calculated as features.

[0019] Furthermore, in step S3, the lifting reminder interval is y minutes, and the sitting posture is monitored at all times within the y time interval. If the same bad sitting posture is monitored for 5 consecutive minutes, it is judged as a bad sitting posture. The same bad sitting posture can be monitored multiple times at different time periods within the y time interval, and the maximum threshold is 10 minutes for the same bad sitting posture.

[0020] Furthermore, in step S4, the corresponding healthy sitting posture level is matched according to the user's bad sitting posture index, and the reminder interval is dynamically adjusted; the healthy sitting posture level is divided into four levels: A healthy, B general, C unhealthy, and D very unhealthy. The specific matching rules are as follows:

[0021] When the bad sitting posture index is 0 in the y time period, it is judged as "A health" level;

[0022] When the bad sitting posture index is 1-2, it is initially judged as "B General" level, and then the duration of these 1-2 bad sitting posture characteristics is judged. If any of them reaches the preset maximum threshold and lasts for more than 10 minutes, the health level of this time period is "C Unhealthy";

[0023] When the bad sitting posture index is between 3 and 4, it is judged as "C unhealthy" level. At the same time, it is judged whether 3 or more of the 3-4 bad sitting posture characteristics have reached the maximum duration threshold. If so, it is upgraded to "D very unhealthy" level;

[0024] If the bad sitting posture index is ≥5, it will be directly judged as "D very unhealthy" level.

[0025] Furthermore, in step S5, the sitting posture level determination time points are y-15, y-10, and y-5. If the health levels B, C, and D are met, a reminder can be given.

[0026] Furthermore, in step S6, at the end of the day, the healthy sitting posture level determined in each time period is summarized and counted, and the number of time periods in the four levels A, B, C, and D on that day is calculated, and recorded as TA, TB, TC, and TD respectively.

[0027] Considering the differences in the potential effects of different levels of healthy sitting posture on the body, each level is assigned a corresponding weight; the weight of the A healthy level is set to 0, indicating no adverse effects; the weight of the B general level is set to 1; the weight of the C unhealthy level is set to 3, and the weight of the D very unhealthy level is set to 5; the weighted average method is used to calculate the comprehensive health score of the day, and the calculation formula is: S = TB × 1 + TC × 3 + TD × 5 / TA + TB + TC + TD;

[0028] Set the health level for the day based on the S value:

[0029] If S≤1, the health level for the day is determined to be "A health", indicating that the overall sitting posture is good;

[0030] If 1<S≤2, the health level for the day is set as "B General", which means that there is a certain bad sitting posture;

[0031] If 2<S≤3, the health level for that day is determined to be "C unhealthy", indicating that bad sitting postures are more frequent and some of them are more serious, and the body may have been under a certain load;

[0032] If S>3, the health level for that day is determined to be "D Very Unhealthy", which means that the sitting posture on that day is worrying and will greatly damage health in the long run;

[0033] Based on the health level of the day, get the next reminder interval.

[0034] Furthermore, in step S7, the user's daily and weekly fatigue cycles are located based on the association between historical sitting posture and daily health level, and the reminder interval is actively shortened at that time.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] The method of the present invention can help users achieve convenient and continuous health management through automatic sitting posture recognition and height adjustment, thereby effectively preventing and reducing health problems related to long-term sitting, such as cervical spondylosis, lumbar muscle strain, etc.; through computer vision technology and human-computer interaction technology, the user's sitting posture state is identified, and the specific health level of the sitting posture is further determined. The refined health assessment can provide users with more personalized health suggestions and intervention measures to improve the user's health level; and addition and subtraction processing is performed according to factors such as health level, lifting time and frequency, so as to avoid too frequent or untimely adjustments, thereby improving the user's office experience and health. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a flow chart of the method of the present invention;

[0038] Figure 2 The method for identifying and judging the sitting posture health level of the present invention;

[0039] Figure 3 This is a flow chart of the sitting posture health level triggering lifting reminder in the present invention. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0041] The height adjustment system of the lifting table of the present invention comprises a lifting table, a sedentary reminder box, a mobile phone APP, a pressure sensor, etc. The radar sensor is located at the bottom of the lifting table, and the pressure sensor is located at the bottom of the office chair.

[0042] like Figure 1 As shown, a method for adjusting the height of a lifting table based on sitting posture recognition comprises the following steps:

[0043] Step 1: Data collection and preprocessing.

[0044] Data is collected through radar sensors and pressure sensors. The radar sensor captures the relative position and movement between the human body and the lifting table, such as whether the user's body is close to the table, whether he or she stands up, etc. The pressure sensor can obtain the pressure distribution on the seat, such as whether the user sits evenly on the seat, whether he or she crosses his or her legs, leans on his or her body, etc., which leads to uneven pressure distribution. These sensor data are the basic information source for sitting posture recognition. The basic information obtained is fused and processed. For example, when the pressure sensor detects that the pressure on one side of the seat suddenly decreases, and the radar sensor detects that the human body leans on that side, it can more accurately determine that the user may have a bad sideways sitting posture.

[0045] Due to occasional sensor failures or external interference, erroneous data points are eliminated by setting a reasonable data range or using a filtering algorithm to ensure the quality of data for subsequent analysis. The collected data is normalized or standardized so that data with different features have the same scale. For example, the distance data collected by the radar sensor and the pressure data collected by the pressure sensor are normalized so that they are both within a unified range of 0-1 or similar. In this way, in subsequent feature extraction and model training, each feature can equally affect the analysis results, avoiding over- or under-emphasis on certain features due to different data dimensions.

[0046] Step 2: Data feature extraction.

[0047] First, define features based on the key elements of sitting posture. For example, extract body angle features from radar sensor data, such as the angle between the back and the vertical direction to determine whether the person is hunched (the angle between the back and the vertical direction in a normal sitting posture is close to 0 degrees, and the angle will increase when the person is hunched), and the angle between the body and the table in the horizontal direction to determine whether the body is leaning forward (the angle is small in a normal sitting posture, and increases when the person is leaning forward); extract pressure distribution features from pressure sensor data, such as the pressure difference between the left and right sides to determine whether the person is crossing his legs (when crossing his legs, the pressure on one side will be significantly reduced, resulting in an increase in the pressure difference).

[0048] Then, the statistical values ​​of the sitting posture characteristics within a period of time are calculated as features. The average value and standard deviation of the hunchback angle, the maximum value and minimum value of the body forward leaning angle, etc. are calculated within a monitoring period (such as y minutes). These statistical features can reflect the stability and degree of change of the user's sitting posture within the period. For example, if the average value of the hunchback angle is large and the standard deviation is also large, it means that the user not only hunches frequently, but also has a large fluctuation in the degree of hunchback.

[0049] Step 3: Sitting posture monitoring and analysis.

[0050] The interval between lifting and lowering reminders is y minutes. The sitting posture is monitored at all times within the y time interval. If the same bad sitting posture is monitored for 5 consecutive minutes, it is judged as a bad sitting posture. The same bad sitting posture can be monitored multiple times at different time periods within the y time interval. The maximum threshold is 10 minutes for the same bad sitting posture. This can not only avoid frequent reminders causing trouble to users, but also ensure that users maintain good sitting habits for a long time.

[0051] Step 4: Identify your sitting posture health status.

[0052] According to the user's bad sitting posture index, the corresponding healthy sitting posture level is matched (as shown in Table 1), and the reminder interval is dynamically adjusted. Specifically, the healthy sitting posture level is divided into four levels: A healthy, B general, C unhealthy, and D very unhealthy. Figure 2 The specific matching rules are as follows:

[0053] When the bad sitting posture index is 0 in the y time period, it is judged as "A health" level;

[0054] When the bad sitting posture index is 1-2, it is first preliminarily judged as "B General" level, and then the duration of these 1-2 bad sitting posture characteristics is judged. If any of them reaches the preset maximum threshold, for example, for the bad sitting posture of "hunching over" lasting for more than 10 minutes, the health level of this time period is "C Unhealthy";

[0055] When the bad sitting posture index is between 3 and 4, it is judged as "C unhealthy" level. At the same time, the system determines whether 3 or more of the 3-4 bad sitting posture characteristics have reached the maximum duration threshold. If so, it is upgraded to "D very unhealthy" level;

[0056] If the bad sitting posture index is ≥5, it will be directly judged as "D very unhealthy" level.

[0057] Step 5: Set up the desk adjustment mechanism.

[0058] like Figure 3 As shown, the time points for judging the sitting posture level are y-15, y-10, and y-5. If the health levels B, C, and D are met, a reminder can be given. For example, the system sets a reminder interval of 60 minutes, with a start time to an end time of 8:00-9:00. When the time reaches 8:45, the system counts the sitting posture health level. If the level is B, the initial judgment reminder time is 8:55; if the level is C, the initial judgment reminder time is 8:50; if the level is D, the reminder is triggered at this moment; if the level is A, the initial judgment reminder time is 9:00, and the sitting posture health status is continuously monitored. When the time reaches 8:50, the sitting posture health level is monitored and judged again. If the level is C or D, a reminder is triggered at this moment; if the level is B, the reminder time is 8:50.

[0059] For example: according to the sitting posture level of the user at each stage, the sit and stand reminder time and interval of the lift table are determined. For example, the user's sitting start time is 9:00. If the system sets a 40-minute sit-stand reminder, when the time difference reaches 15 minutes, that is, 9:25, the system collects the user's sitting posture level. If it is C, it is reminded according to y-10, that is, 40-10=30, that is, the lift reminder is triggered at 9:30.

[0060] Table 1: Healthy sitting posture level table

[0061] grade index Reminder duration (reminder interval y minutes) AHealth 0 0 BGeneral 1-2 y-5 C. Unhealthy 3-4 y-10 D Very unhealthy ≥5 y-15

[0062] Step 6: Identify your health status level for the day.

[0063] At the end of the day, the system summarizes and counts the healthy sitting posture levels determined in each time period, and calculates the number of time periods at the four levels A, B, C, and D on that day, which are recorded as TA, TB, TC, and TD respectively.

[0064] Considering the differences in the potential effects of different healthy sitting posture levels on the body, each level is assigned a corresponding weight. As shown in Table 2, the weight of the A healthy level is set to 0 (indicating no adverse effects), the weight of the B general level is set to 1, the weight of the C unhealthy level is set to 3, and the weight of the D very unhealthy level is set to 5. This shows that as the degree of poor sitting posture increases, the negative impact on health increases.

[0065] The comprehensive health score for the day is calculated by the weighted average method, and the calculation formula is: S = TB×1+TC×3+TD×5 / TA+TB+TC+TD.

[0066] Set the health level for the day based on the S value:

[0067] If S≤1, the health level for the day is determined to be "A health", indicating that the overall sitting posture is good and a bad sitting posture has little impact on the working status of the day.

[0068] If 1<S≤2, it is "B General", which means that there are certain bad sitting postures, but they are still controllable overall and have not caused significant cumulative impact on health.

[0069] If 2<S≤3, it is "C unhealthy", indicating that bad sitting postures were frequent and some of them were severe, and the body may have been under a certain load.

[0070] If S>3 means "D is very unhealthy", it means your sitting posture on that day is worrying and it will greatly damage your health in the long run.

[0071] Based on the health level of the day, the next reminder interval is obtained. For example, if the initial reminder interval is y minutes and the health level is C, the next initial reminder interval is y-10.

[0072] Table 2: Health level table

[0073]

[0074] Step 7: Health trend analysis.

[0075] According to the correlation between historical sitting posture and daily health level, the user's daily (as shown in Table 3) and weekly fatigue cycles (as shown in Table 4) are located, and the reminder interval is actively shortened. That is, the user actively adjusts the y value (sitting and standing reminder time of the lifting desk) in the mobile phone APP.

[0076] Table 3: User daily table

[0077] time 8:00-9:00 9:23-10:13 10:31-11:15 14:21-15:16 16:11-17:01 8:00-17:01 grade A B A C B B

[0078] Table 4: Weekly fatigue period table

[0079] November on Monday Tuesday Wednesday Thursday Friday Week 1 A A B B A Week 2 B B A A C Week 3 C A A B C Week 4 B B D A B

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for adjusting the height of a lifting table based on sitting posture recognition, characterized in that: The following steps are involved: S1. Data collection and preprocessing; S2. Extract features from the collected data; S3. Perform sitting posture monitoring and analysis; S4. Identify the sitting posture health status; S5. Set up the lifting table adjustment mechanism; S6. Identify the health status level of the day; S7. Health trend analysis.

2. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 1, characterized in that: The specific process of step S1 is: collecting data through radar sensors and pressure sensors, and fusing the acquired basic information; the radar sensor is used to capture information between the human body and the lifting table; the pressure sensor is used to obtain pressure distribution information on the seat.

3. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 1, characterized in that: In step S2, the angle features of the body are extracted from the radar sensor data, and the statistical values ​​of the sitting posture features over a period of time are calculated as features.

4. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 1, characterized in that: In step S3, the lifting reminder interval is y minutes, and the sitting posture is monitored at all times within the y time interval. If the same bad sitting posture is monitored for 5 consecutive minutes, it is judged as a bad sitting posture. The same bad sitting posture can be monitored multiple times at different time periods within the y time interval. The maximum threshold is 10 minutes for the same bad sitting posture.

5. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 1, characterized in that: In step S4, the corresponding healthy sitting posture level is matched according to the user's bad sitting posture index, and the reminder interval is dynamically adjusted; the healthy sitting posture level is divided into four levels: A healthy, B general, C unhealthy, and D very unhealthy. The specific matching rules are as follows: When the bad sitting posture index is 0 in the y time period, it is judged as "A health" level; When the bad sitting posture index is 1-2, it is initially judged as "B General" level, and then the duration of these 1-2 bad sitting posture characteristics is judged. If any of them reaches the preset maximum threshold and lasts for more than 10 minutes, the health level of this time period is "C Unhealthy"; When the bad sitting posture index is between 3 and 4, it is judged as "C unhealthy" level. At the same time, it is judged whether 3 or more of the 3-4 bad sitting posture characteristics have reached the maximum duration threshold. If so, it is upgraded to "D very unhealthy" level; If the bad sitting posture index is ≥5, it will be directly judged as "D very unhealthy" level.

6. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 5, characterized in that: In step S5, the sitting posture level is determined at time points y-15, y-10, and y-5. If the health levels B, C, and D are met, a reminder can be given.

7. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 1, characterized in that: In step S6, at the end of the day, the healthy sitting posture level determined in each time period is summarized and counted, and the number of time periods in the four levels A, B, C, and D on that day is calculated, and recorded as TA, TB, TC, and TD respectively. Considering the differences in the potential effects of different levels of healthy sitting posture on the body, each level is assigned a corresponding weight; the weight of the A healthy level is set to 0, indicating no adverse effects; the weight of the B general level is set to 1; the weight of the C unhealthy level is set to 3, and the weight of the D very unhealthy level is set to 5; the weighted average method is used to calculate the comprehensive health score of the day, and the calculation formula is: S = TB × 1 + TC × 3 + TD × 5 / TA + TB + TC + TD; Set the health level for the day based on the S value: If S≤1, the health level for the day is determined to be "A health", indicating that the overall sitting posture is good; If 1<S≤2, the health level for the day is set as "B General", which means that there are certain bad sitting postures; If 2<S≤3, the health level for that day is determined to be "C unhealthy", indicating that bad sitting postures are more frequent and some of them are more serious, and the body may have been under a certain load; If S>3, the health level for that day is determined to be "D Very Unhealthy", which means that the sitting posture on that day is worrying and will greatly damage health in the long run; Based on the health level of the day, get the next reminder interval.

8. The method for adjusting the height of a lifting table based on sitting posture recognition according to claim 1, characterized in that: In step S7, the user's daily and weekly fatigue cycles are located based on the association between historical sitting posture and daily health level, and the reminder interval is actively shortened at that time.

Images