WO2018113582A1 - 一种坐姿识别系统和坐姿识别方法 - Google Patents

一种坐姿识别系统和坐姿识别方法 Download PDF

Info

Publication number
WO2018113582A1
WO2018113582A1 PCT/CN2017/116162 CN2017116162W WO2018113582A1 WO 2018113582 A1 WO2018113582 A1 WO 2018113582A1 CN 2017116162 W CN2017116162 W CN 2017116162W WO 2018113582 A1 WO2018113582 A1 WO 2018113582A1
Authority
WO
WIPO (PCT)
Prior art keywords
posture
alarm
unit
sitting posture
sitting
Prior art date
Application number
PCT/CN2017/116162
Other languages
English (en)
French (fr)
Inventor
姜海涛
范晓鸣
Original Assignee
欧普照明股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 欧普照明股份有限公司 filed Critical 欧普照明股份有限公司
Publication of WO2018113582A1 publication Critical patent/WO2018113582A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A61B5/1116—Determining posture transitions
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/0002—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network
    • A61B5/0004—Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network characterised by the type of physiological signal transmitted
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Detecting, measuring or recording devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor, mobility of a limb
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
    • A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/74—Details of notification to user or communication with user or patient ; user input means
    • A61B5/746—Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms
    • G—PHYSICS
    • G08—SIGNALLING
    • G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/18—Status alarms
    • G08B21/24—Reminder alarms, e.g. anti-loss alarms

Definitions

  • the invention relates to a sitting posture recognition system and a sitting posture recognition method.
  • Another way to identify is to use a three-axis accelerometer to detect sitting posture, but the current algorithm is too simple, often due to too sensitive and false positives (not wrong sitting posture false positives, such as just pulling the chair) Or missed (it is wrong sitting but not detected). Therefore, there is a need for a sitting recognition system that can fully consider various possible false sitting postures and user habits, and minimize false positives and false negatives.
  • An object of the present invention is to solve the above problems and to provide a sitting posture recognition system and a sitting posture recognition method which can accurately recognize an incorrect sitting posture.
  • the technical solution adopted is to provide a sitting posture recognition system, and the sitting posture recognition system includes:
  • the sensing unit is worn on the human body when in use, and outputs a sensing signal
  • a posture computing unit receiving the sensing signal from the sensing unit, calculating a posture feature signal, and outputting the posture feature signal when the human body is relatively stationary;
  • a storage unit for storing standard posture data
  • the posture determining unit receives the posture feature signal from the posture computing unit, reads the standard posture data in the storage unit, determines a sitting posture, and outputs a sitting posture correcting information;
  • the alarm judging unit receives the posture correctness information of the posture judging unit, determines whether an alarm is needed, and outputs an alarm signal when an alarm is required;
  • the alarm unit receives the alarm signal and alarms.
  • the sensing unit is an acceleration sensor.
  • the sensing unit is a three-axis acceleration sensor.
  • warning information sent by the alarm unit is a combination of one or more of sound, vibration, color, and brightness.
  • the sensing unit, the posture computing unit, the storage unit, the posture determining unit, the alarm determining unit, and the alarm unit are physically connected together.
  • the sensing unit, the posture computing unit, the storage unit, the posture determining unit, the alarm determining unit, and the alarm unit are respectively disposed in two or more separate portions, and the separated portions transmit information by wireless communication.
  • the wireless communication mode is ZigBee, Bluetooth or WiFi.
  • the present invention also provides a sitting posture recognizing method applied to the above lighting control system, wherein the sitting posture recognizing method includes a posture judging process and an alarm judging process, and the posture judging process is cyclically executed in the sitting posture recognizing process when an error occurs.
  • the alarm determination process is executed during the sitting posture to determine whether an alarm is required, whether or not the alarm is performed, and the posture judgment flow is continued after the execution of the alarm determination flow is completed.
  • the posture determination process includes the following steps:
  • Step A1 The sensing unit reads data, and the subsequent step is step A2;
  • Step A2 determining whether the sensing unit is in a relatively static state, if step A3 is performed, not returning to a stationary state, then returning to step A1;
  • Step A3 The posture computing unit calculates the posture feature signal, and the subsequent step is step A4;
  • Step A4 calculating the difference between the posture feature signal and the standard posture data, the subsequent step is step A5;
  • Step A5 It is judged whether the difference between the posture feature signal and the standard posture data exceeds a threshold value, and if yes, the alarm determination flow is executed, and if the threshold value is not exceeded, the step A1 is returned again.
  • the alarm judgment process includes the following steps:
  • Step B1 determining whether the detection is an incorrect sitting posture twice, if yes, executing step B3, if otherwise, performing step B2;
  • Step B2 the timer is cleared and starts timing, and the subsequent step is step B8;
  • Step B3 the timer accumulates to continue counting, and the subsequent step is step B4;
  • Step B4 determining whether the timer duration exceeds the threshold, if step B5 is performed, if not, executing step B8;
  • Step B5 alarm unit alarm, the subsequent step is step B6;
  • Step B6 delay, the subsequent step is step B7;
  • Step B7 The timer is cleared, and the subsequent step is step B8;
  • Step B8 The alarm judgment process ends, and the posture judgment process is subsequently performed.
  • the alarm judgment process includes the following steps:
  • Step C1 determining whether the detection is an incorrect sitting posture twice, if yes, executing step C3, if otherwise, performing step C2;
  • Step C2 the counter starts counting from zero, and the subsequent step is step C8;
  • Step C3 the counter is accumulated, and the subsequent step is step C4;
  • Step C4 determining whether the value of the counter exceeds the threshold, if step C5 is performed, if not, executing step C8;
  • Step C5 The alarm unit alarms, and the subsequent steps are step C6;
  • Step C6 delay, the subsequent step is step C7;
  • Step C7 the counter is cleared, and the subsequent step is step C8;
  • Step C8 The alarm judgment process ends, and the posture judgment process is subsequently executed.
  • the posture determination process further includes an initialization process that measures and calculates standard posture data.
  • the initialization process includes the following steps:
  • Step D1 the sensing unit reads data, and the subsequent step is step D2;
  • Step D2 determining whether the sensing unit is in a relatively static state, if step D3 is performed, not returning to a stationary state, then returning to step D1;
  • Step D3 judging whether the measured value of the sensing unit meets the standard sitting posture range, if step D4 is performed, if not, a reminder is issued and the process returns to step D1;
  • Step D4 Calculate the standard posture data according to the current degree of the sensor, and subsequently perform the posture judgment process.
  • the technical solution provided by the invention does not directly alarm when judging the wrong sitting posture, but performs alarm by combining further alarm judgments, combined with the distribution of multiple error information, thus avoiding the situation of false negatives and false reports. Improve the accuracy of false sitting posture alarms.
  • FIG. 1 is a schematic structural view of a sitting posture recognition system of the present invention
  • FIG. 2 is a flow chart of a sitting posture recognition method of the present invention
  • Figure 3 is a flow chart showing the flow of alarm determination in the second embodiment of the present invention.
  • FIG. 1 is a diagram of a preferred embodiment of a sitting posture recognition system according to the present invention.
  • the posture recognition system includes: a sensing unit, a posture calculation unit, a storage unit, a posture determination unit, an alarm determination unit, and an alarm unit.
  • the sensing unit is used to sense the state of the user.
  • the sensing unit needs to be worn on the human body, in particular, the upper body, such as the neckline, the head or the back of the ear, so that the human body It can be monitored when the posture changes.
  • the sensing unit monitors the human body to generate the sensing signal S1 in real time.
  • the sensing unit is a three-axis acceleration sensor, and in other preferred embodiments, other sensors such as a tilt sensor and a position sensor may be used. Or the same is an acceleration sensor, but a single-axis or two-axis acceleration sensor is used, and the data of each direction is obtained by the operation of the subsequent posture arithmetic unit.
  • the posture computing unit obtains the posture data by calculating the data obtained by the sensing unit, and the signal input of the posture computing unit is the sensing signal S1 from the sensing unit, and the output is the posture characteristic signal S2.
  • the posture characteristic signal is S2 is the angle of the human body forward, backward and left and right directions.
  • the posture computing unit needs to determine whether the sensing unit is in a relatively static state. If the human body is in a moving state, it indicates that the reading and writing state has not yet entered, and at this time, the posture characteristic signal S2 is not outputted outward. The output of the signal is only performed when the person is in a relatively static state.
  • the standard posture data S3 in the storage unit may be preset in the system when the system is shipped, or may be the forward tilt, the backward tilt, and the left of the human body obtained by the gesture computing unit according to a standard reading and writing posture when the system is running.
  • the angles of the right four directions are used as reference values, written and stored in the storage unit.
  • the posture judging unit has two input information, one is the posture characteristic signal S2 transmitted from the posture computing unit, and the other is the standard posture data S3 in the storage unit, and the calculation is performed.
  • the deviation value between the posture characteristic signal S2 and the standard posture data S3 is the difference between the angles of the human body forward tilt, the backward tilt, and the left and right directions in the embodiment, and any one of the angle changes exceeds the threshold value
  • the sitting posture correct information S4 is output.
  • the judging unit judges that it is an erroneous sitting posture, which causes a lot of false alarms, so that an alarm judging unit is added to the system of the embodiment.
  • the alarm judging unit takes the sitting posture error information S4 of the posture judging unit as an input, determines whether an alarm is required, and outputs an alarm signal S5 when an alarm is required.
  • the judgment basis of the alarm judging unit is mainly to synthesize the sitting posture correcting information S4 repeatedly, and judge the continuity of the error state.
  • the posture error of one time cannot trigger the alarm, and the alarm must be continued for a certain period of time.
  • the specific alarm judging algorithm will be behind us. Detailed introduction.
  • the alarm unit receives the alarm signal S5 sent by the alarm judgment unit and issues an alarm.
  • the way of alarm can be So it makes a sound, produces vibration, changes color, flashes light, or changes in light brightness.
  • a plurality of combinations can be used for alarming, such as sound plus vibration, brightness of the red warning light.
  • the change combines sound, or vibration with a color-changing LED that blinks continuously.
  • each unit only represents the transfer of data, and does not mean the actual physical connection, and each unit may be provided in an integrated manner or in a separate manner.
  • the sensing unit, the posture computing unit, the storage unit, the posture determining unit, the alarm judging unit, and the alarm unit are all disposed in a small device that can be connected to the neckline, and the data is transmitted through the connection between each other. .
  • the sensing unit, the posture computing unit, and the alarm unit may be fabricated as wearing parts, and the storage unit, the posture determining unit, and the alarm determining unit are disposed on a fixed body, such as a desk lamp, such that It can guarantee the power supply problem of the computing components that consume more power.
  • the sensing unit is separately disposed on the human body, the posture computing unit, the storage unit, the posture determining unit, and the alarm determining unit are disposed in a single component, and the alarm unit is an app disposed on the handheld mobile device.
  • the wireless communication method can be WiFi, Zigbee or Bluetooth.
  • FIG. 2 is a flowchart of a preferred embodiment of the present invention, which includes three processes of an initialization process, a posture judgment process, and an alarm judgment process, wherein the posture judgment is performed.
  • the process and the alarm judging process are the basic processes in the sitting posture recognizing method of the present invention.
  • the posture judging unit cyclically executes the posture judging process, and the posture judging unit sends the sitting posture to the alarm judging unit only when the wrong sitting posture occurs.
  • the error information S4 so that the alarm determination unit executes the alarm determination flow, determines whether an alarm is required, whether or not the alarm is issued, and after the execution of the alarm determination flow is completed, the posture determination unit continues to execute the posture determination flow.
  • the posture judgment process is completed by the posture judgment unit, and specifically includes the following steps:
  • Step A1 The sensing unit reads the data, and the subsequent step is step A2;
  • Step A2 judging whether the sensing unit is in a relatively static state, in the present invention, our main purpose is to identify whether the sitting posture is correct, and the sitting posture must be a relatively static concept, if the person is in motion, such as wearing a sensor Walking, or bending over and picking up things are not considered to be sitting, so if people are in motion, we don't judge sitting, but wait for them to calm down and judge. Therefore, if it is judged in this step that the sensor is relatively stationary, step A3 is performed, and if it is not in a relatively stationary state, returning to step A1 is continued until the person is stationary to maintain a certain posture.
  • the judgment about the relative static is judged by comparing the measured values of the two adjacent sensing units according to the amount of change. If the overall change is small, it can be determined to be stationary, and the specific amount of change is related to the sensitivity of the sensor.
  • Step A3 The posture operation unit calculates the posture feature signal S2, and the subsequent step is step A4;
  • Step A4 calculating the difference between the posture feature signal S2 and the standard posture data S3, the subsequent step is step A5;
  • Step A5 It is judged whether the difference between the posture characteristic signal S2 and the standard posture data S3 exceeds the threshold. If yes, the alarm determination process is subsequently executed. If the threshold value is not exceeded, the process returns to step A1 again to continue to monitor the sitting posture.
  • the alarm judgment process is completed by the alarm judgment unit, and specifically includes the following steps:
  • Step B1 It is determined whether the detection is an incorrect sitting posture twice, if yes, step B3 is performed, and if otherwise, step B2 is performed.
  • the judgment about the two consecutive errors is mainly to avoid false positives. Sometimes, in the reading and writing, only the sitting posture is not correct, and may return to the correct posture immediately. If the alarm will make the alarm too frequent every time, our main idea is It is the alarm that is issued when there are many sitting posture errors in succession. Anything that enters step B1 means that the test is an incorrect sitting posture. Therefore, the judgment of the two consecutive tests is to judge whether the previous test result is the correct sitting posture or the wrong sitting posture. If it is the correct sitting posture, it indicates that this time is the first mistake. Execute B2 to start recording the duration of the error.
  • Step B2 This step is a timer-on action. When the first error occurs, we clear a timer and start timing. The next step is step B8.
  • Step B3 the timer accumulates to continue counting, and the subsequent step is step B4;
  • Step B4 It is determined whether the timer duration exceeds the threshold, if step B5 is performed, and if not, step B8 is performed.
  • This threshold indicates how long the error state lasts. We will alarm. If the error state lasts longer than this threshold, the long side performs alarm step B5.
  • the specific value of the threshold can be set for different detection scenarios.
  • Step B5 The alarm unit alarms, and the subsequent step is step B6.
  • Step B6 is a delay step of letting the alarm last for a while, and the subsequent step is step B7.
  • Step B7 The timer is cleared, and the subsequent step is step B8.
  • the timer needs to be cleared. Otherwise, when the detected person does not adjust the sitting posture, the next error detection occurs when step B4 is reached, and the duration of the timer must exceed the threshold. It will directly alarm, so the alarm will be too frequent, and it will not reach the alarm when we want the state to last, and the two alarms need to be separated by a certain time.
  • Step B8 The alarm judgment process ends, and the posture judgment process is subsequently performed. Regardless of whether the previous process is an alarm or not, step B8 is the end of the alarm determination process, and each branch finally goes to this step to end the alarm determination process.
  • the alarm judging process does not adopt the timing method as in the previous embodiment, but adopts the counting method, that is, only when the error posture is The number of consecutive occurrences reaches the set value before the alarm, and the correct posture in the middle is cleared and recounted.
  • the specific process is shown in Figure 3, including the following steps:
  • Step C1 determining whether the detection is an incorrect sitting posture twice, if yes, executing step C3, if otherwise, performing step C2;
  • Step C2 the counter starts counting from zero, and the subsequent step is step C8;
  • Step C3 the counter is accumulated, and the subsequent step is step C4;
  • Step C4 determining whether the value of the counter exceeds the threshold, if step C5 is performed, if not, executing step C8;
  • Step C5 The alarm unit alarms, and the subsequent steps are step C6;
  • Step C6 delay, the subsequent step is step C7;
  • Step C7 the counter is cleared, and the subsequent step is step C8;
  • Step C8 The alarm judgment process ends, and the posture judgment process is subsequently executed.
  • the posture characteristic signal S2 is calculated after the reading of the sensing unit, and the standard posture is obtained.
  • the data S3 may be a value set in advance, which is stored in the storage unit. In fact, because each person's body shape and habits are different, each person's sitting posture is not the same. It is not suitable to use the unified standard posture data S3. In order to solve this problem, we add an initialization process before the gesture judgment process, which is used to measure and calculate the standard posture data S3.
  • the design of this process is such that, when starting to use, the user is first reminded to reach a standard sitting posture, and then the sensing unit data in this posture is detected, and the standard posture data S3 is calculated and stored in the storage unit. Subsequent judgments are based on this value.
  • the initialization process includes the following steps:
  • Step D1 The sensing unit reads the data, and the subsequent step is step D2.
  • Step D2 It is judged whether the sensing unit is in a relatively static state. If the step D3 is performed, the process returns to step D1 instead of the relatively stationary state. Here we must allow the user to enter the read and write state of this static state before they can be measured. If the person has not stood still and is still preparing for work, it will not be measured.
  • Step D3 judging whether the measured value of the sensing unit meets the standard sitting posture range, if step D4 is performed, if not, a reminder is sent and the step D1 is returned.
  • the standard posture data S3 is calculated through the initialization process, we must also ensure that the measured value is indeed a standard sitting posture. Therefore, we still have a judgment in obtaining the standard posture data S3, and a value is set in advance. The value cannot deviate too much from this value. If the measured value is within the allowable range, the value is recorded as the standard posture data S3. When the range is exceeded, the sensor reading is continued after the alarm is issued. In this embodiment, the voice reminder is used. , please use the standard In the other way, the brightness can be changed. Lights up to enter the official read and write state.
  • Step D4 Calculate the standard posture data S3 according to the current degree of the sensor, and store it in the storage unit, the initialization process ends, and then enter the normal working state, and the posture judgment process is executed.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Biophysics (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Veterinary Medicine (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Physiology (AREA)
  • Dentistry (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Business, Economics & Management (AREA)
  • Emergency Management (AREA)
  • General Physics & Mathematics (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
  • Emergency Alarm Devices (AREA)

Abstract

一种坐姿识别系统和坐姿识别方法。系统包括：传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元。在人处于相对静止状态后开始进行坐姿识别，并在判断出错坐姿时并不直接报警，而是通过进一步的报警判断，结合多个错误信息的分布情况来进行报警，这样就避免了漏报、错报的情况，提高了错误坐姿报警的准确率。

Description

一种坐姿识别系统和坐姿识别方法 技术领域
本发明涉及一种坐姿识别系统和坐姿识别方法。
背景技术
当今，已经进入了一个信息爆炸的时代，无论是学生还是成人，每天都需要进行大量的伏案工作、学习。随之而来的是驼背、颈椎病、肩周炎、干眼症、近视眼等症状，给人们带来越来越多的困扰。这些都是由于坐姿不正确引起的，因此及时纠正错误坐姿并保持正确的坐姿是一个亟待解决的问题。
目前市面上销售的坐姿矫正设备，大多是纯物理式的矫正设备，或者使用红外测距传感器或超声波测距传感器，通过测量人脸到传感器的距离来判断坐姿，而学生和成人的身高不同、测距仪放置的位置、阻挡物的存在都会影响设备的正确判断，因而识别率不高。
而另一种识别方式是利用三轴加速度传感器来检测坐姿，但是目前所采用的算法过于简单，经常由于过于灵敏而产生误报(不是错误坐姿误报为错误坐姿，如只是拉一下椅子)，或者漏报(是错误坐姿但是没检测出来)。因此，需要一种可以充分考虑各种可能的错误坐姿和用户使用习惯，尽量减少误报和漏报的坐姿识别系统。
发明内容
本发明的目的是为了解决上述问题，提供一种可以准确识别错误坐姿的坐姿识别系统和坐姿识别方法。
本发明为实现上述功能，所采用的技术方案是提供一种坐姿识别系统，所述坐姿识别系统包括：
传感单元，使用时所述传感单元佩戴于人体上，输出传感信号；
姿势运算单元，从所述传感单元接收所述传感信号，运算获得姿势特征信号，并在人体处于相对静止时输出所述姿势特征信号；
存储单元，用于存储标准姿势数据；
姿势判断单元，从所述姿势运算单元接收所述姿势特征信号，并读取所述存储单元中的所述标准姿势数据，判断坐姿，输出坐姿正误信息；
报警判断单元，接收所述姿势判断单元的所述坐姿正误信息，判断是否需要报警，在需要报警时输出报警信号；
报警单元，接收所述报警信号，并报警。
进一步的，所述传感单元为加速度传感器。
进一步的，所述传感单元为三轴加速度传感器。
进一步的，所述报警单元发出的警示信息为声音、振动、色彩、亮度中的一种或多种的组合。
进一步的，所述传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元之间物理连接为一体。
进一步的，所述传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元分别设置在两个以上分离的部分，所述分离的部分之间通过无线通信方式传递信息。
进一步的，所述无线通信方式为ZigBee、蓝牙或WiFi。
本发明还提供一种应用于上述的照明控制系统的一种坐姿识别方法，所述坐姿识别方法包括姿势判断流程和报警判断流程，在坐姿识别过程中循环执行所述姿势判断流程，当出现错误坐姿时执行所述报警判断流程，判断是否需要报警，无论是否报警，在所述报警判断流程执行完毕后继续执行所述姿势判断流程。
进一步的，所述姿势判断流程包括如下步骤：
步骤A1：所述传感单元读取数据，后续步骤为步骤A2；
步骤A2：判断所述传感单元是否处于相对静止状态，如是执行步骤A3，不是相对静止状态则返回执行步骤A1；
步骤A3：姿势运算单元计算姿势特征信号，后续步骤为步骤A4；
步骤A4：计算所述姿势特征信号和标准姿势数据之间的差值，后续步骤为步骤A5；
步骤A5：判断所述姿势特征信号和标准姿势数据之间的差值是否超过阈值，如是则执行所述报警判断流程，没有超过阈值则返回再次执行所述步骤A1。
进一步的，所述报警判断流程包括如下步骤：
步骤B1：判断是否连续两次检测为错误坐姿，如是则执行步骤B3，如否则执行步骤B2；
步骤B2：计时器清零并开始计时，后续步骤为步骤B8；
步骤B3：计时器累计继续计时，后续步骤为步骤B4；
步骤B4：判断计时器时长是否超过阈值，如是执行步骤B5，如否执行步骤B8；
步骤B5：报警单元报警，后续步骤为步骤B6；
步骤B6：延时，后续步骤为步骤B7；
步骤B7：计时器清零，后续步骤为步骤B8；
步骤B8：报警判断流程结束，后续执行姿势判断流程。
进一步的，所述报警判断流程包括如下步骤：
步骤C1：判断是否连续两次检测为错误坐姿，如是则执行步骤C3，如否则执行步骤C2；
步骤C2：计数器从零开始计数，后续步骤为步骤C8；
步骤C3：计数器累加，后续步骤为步骤C4；
步骤C4：判断计数器的值是否超过阈值，如是执行步骤C5，如否执行步骤C8；
步骤C5：报警单元报警，后续步骤为步骤C6；
步骤C6：延时，后续步骤为步骤C7；
步骤C7：计数器清零，后续步骤为步骤C8；
步骤C8：报警判断流程结束，后续执行姿势判断流程。
进一步的，所述姿势判断流程之前还包括初始化流程，所述初始化流程测量并计算标准姿势数据。
进一步的，所述初始化流程包括如下步骤：
步骤D1：所述传感单元读取数据，后续步骤为步骤D2；
步骤D2：判断所述传感单元是否处于相对静止状态，如是执行步骤D3，不是相对静止状态则返回执行步骤D1；
步骤D3：判断所述传感单元测得值是否符合标准坐姿范围，如是执行步骤D4，不符合则发出提醒并返回执行步骤D1；
步骤D4：根据传感器当前度数计算所述标准姿势数据，后续执行姿势判断流程。
本发明所提供的技术方案在判断出错误坐姿时并不直接报警，而是通过进一步的报警判断，结合多个错误信息的分布情况来进行报警，这样就避免了漏报、错报的情况，提高了错误坐姿报警的准确率。
附图说明
图1是本发明坐姿识别系统的结构示意图；
图2是本发明坐姿识别方法的流程图；
图3是本发明第二实施例中报警判断流程的流程图。
具体实施方式
以下结合附图和具体实施例对本发明提出的一种坐姿识别系统、坐姿识别方法作进一步详细的说明。
请参考图1，图1所示是本发明提出的一种坐姿识别系统的一个较佳实施例的结 构示意图，该坐姿识别系统包括：传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元。
其中传感单元是用来感测使用者的状态的，在需要监测坐姿时，传感单元需要佩戴于人体上，特别是应当佩戴于上身，如领口、头部或者耳后，这样当人体的姿势发生改变时可被其监测到。传感单元实时监测人体产生传感信号S1。在本实施例中传感单元为三轴加速度传感器，在其他较佳实施例中也可以采用倾斜传感器、位置传感器等其他传感器。或者同样为加速度传感器，但是采用单轴或双轴加速度传感器，而通过后续的姿势运算单元的运算来获得各方向数据。
姿势运算单元，是将传感单元获得的数据通过运算获得姿势数据，姿势运算单元的信号输入是来自传感单元的传感信号S1，输出为姿势特征信号S2，在本实施例中姿势特征信号S2为人体前倾、后仰和左、右4个方向的角度。在本实施例中，姿势运算单元需判断传感单元是否处于相对静止状态，如人体是处于运动状态的，那么就表示还未进入读写状态，这时是不向外输出姿势特征信号S2的，只有在人处于相对静止状态时才进行信号的输出。
为了判断坐姿是否正确，我们必须要有一个基准值，通过将姿势特征信号S2和这个基准值进行比较来识别错误坐姿，而存储单元就是用来存储这个基准值的，这里我们称为标准姿势数据S3。存储单元中的标准姿势数据S3可以是在系统出厂时就预先设置好，也可以是在系统运行时，由姿势运算单元根据一个标准的读写姿态而获得的人体前倾、后仰和左、右4个方向的角度来作为基准值，写入并存放在存储单元中。
判断坐姿的运算是通过姿势判断单元来完成的，因此姿势判断单元的输入信息有两个，一是从姿势运算单元传来的姿势特征信号S2，二是存储单元中的标准姿势数据S3，计算姿势特征信号S2和标准姿势数据S3之间的偏差值，在本实施例中为人体前倾、后仰和左、右4个方向的角度的差值，当其中任意一个角度变化超过阈值即判断为坐姿错误，输出坐姿正误信息S4。
在实际使用的过程中由于传感单元是实时监测的，而人在读写过程中并不是一动不动的，会有很多小动作，如附身拿个橡皮、伸一下懒腰，这些动作都会使得姿势判断单元判断为错误坐姿，这样会产生很多的误报，因此在本实施例的系统中增加了报警判断单元。报警判断单元以姿势判断单元的坐姿正误信息S4为输入，判断是否需要报警，在需要报警时输出报警信号S5。报警判断单元的判断依据主要是综合多次传来的坐姿正误信息S4，判断错误状态的连续性，一次的姿势错误无法触发报警，必须持续一定时间才会报警，具体报警判断算法我们将在后面详细介绍。
报警单元，接收报警判断单元发出的报警信号S5，并发出报警。报警的方式可 以是发出声音、产生振动、改变色彩、灯光闪烁或灯光亮度变化。在本实施例中我们选择上述报警方式中的一种——声音警报来进行报警，在其他较佳实施例中可采用多种方式组合来进行报警，如声音加振动、以红色警示灯的亮度变化结合声音、或振动配以可变色的LED不停闪烁等。
在图1所示的坐姿识别系统中，其中各单元的连线仅代表数据的传递，并不意味着实际的物理连接，各个单元可以为一体式设置，也可以采用分离设置的方式。如在本实施例中传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元都设置在一个可别在领口上的小型设备之中，彼此之间通过连线传递数据。而在另外一些较佳实施例中，可以将传感单元、姿势运算单元、报警单元制作为佩戴部件，而存储单元、姿势判断单元、报警判断单元设置在一个固定体上，如台灯上，这样可以保证耗电较多的运算部件的供电问题。而在另外的方案中，传感单元单独设置于人体，姿势运算单元、存储单元、姿势判断单元、报警判断单元设置在一个单独部件内，而报警单元为设置于手持式移动设备的app。在上述两个实施例中，这些分离部分之间仍然需要进行数据传递，他们之间通过无线信号来进行通讯。无线通信方式可以为WiFi、Zigbee或蓝牙。
下面就具体实施例的流程图对本发明的坐姿识别方法进行说明，图2为本发明一个较佳实施例的流程图，其包括初始化流程、姿势判断流程、报警判断流程三个流程，其中姿势判断流程和报警判断流程，是本发明坐姿识别方法中的基本流程，在坐姿识别过程中由姿势判断单元循环执行姿势判断流程，只有当出现错误坐姿时，姿势判断单元才会向报警判断单元发送坐姿正误信息S4，从而报警判断单元执行报警判断流程，判断是否需要报警，无论是否报警，在所述报警判断流程执行完毕后，姿势判断单元将会继续执行姿势判断流程。
姿势判断流程由姿势判断单元来完成，具体包括如下步骤：
步骤A1：传感单元读取数据，后续步骤为步骤A2；
步骤A2：判断所述传感单元是否处于相对静止状态，在本发明中我们主要的目的是为了识别坐姿是否正确，而坐姿必须是一个相对静止的概念，如果人处于运动状态，如佩戴着传感器走动，或弯腰捡东西这些情况都不是坐姿需要考虑的，因此如果人处于运动状态我们不进行坐姿判断，而是等待其静下来再来判断。因此在本步骤中如果判断传感器是相对静止的那么执行步骤A3，如果不是相对静止状态则返回继续执行步骤A1，直至人静止下来可以保持一定的姿势。关于相对静止的判断是通过比较相邻两次的传感单元测量值，依据其变化量来判断的，如总体变化很小，可以判定为静止，具体的变化量和传感器的灵敏度有关。
步骤A3：姿势运算单元计算姿势特征信号S2，其后续步骤为步骤A4；
步骤A4：计算所述姿势特征信号S2和标准姿势数据S3之间的差值，后续步骤为步骤A5；
步骤A5：判断姿势特征信号S2和标准姿势数据S3之间的差值是否超过阈值，如是则后续执行报警判断流程，没有超过阈值则返回再次执行步骤A1，继续监测坐姿。
而报警判断流程由报警判断单元来完成，具体包括如下步骤：
步骤B1：判断是否连续两次检测为错误坐姿，如是则执行步骤B3，如否则执行步骤B2。关于连续两次错误的判断主要是为了避免误报，有时在读写中仅是一时的坐姿不正，可能马上就会回归正确姿势，如果每一次都报警会使得报警过于频繁，因此我们的主要思路是在连续多次的坐姿错误时才进行报警。凡是进入到步骤B1的都是表示本次检测是错误坐姿，因此这个连续两次检测的判断就是判断前一次的检测结果是正确坐姿还是错误坐姿，如果是正确坐姿则表示本次是首次错误，执行B2开始记录错误的持续时间，如果前一次为错误坐姿则执行B3继续计时。当两次错误中间夹有一次正确时，在正确姿势其后的那次错误时我们仍然认为他是首次错误，重新对计时器清零重新开始计时，这样我们可以保证错误一定要持续足够的时间才会报警。
步骤B2：这个步骤是一个计时器开启的动作，在首次错误出现时，我们对一个计时器清零并开始计时,其后续步骤为步骤B8。
步骤B3：计时器累计继续计时，后续步骤为步骤B4；
步骤B4：判断计时器时长是否超过阈值，如是执行步骤B5，如否执行步骤B8。这个阈值就表示错误状态持续多长时间我们会进行报警，小于这个时长不报警，如果错误状态持续时间超过这个阈值时长侧执行报警步骤B5。阈值的具体值可以针对不同的检测场景来设置。
步骤B5：报警单元报警，后续步骤为步骤B6。步骤B6是一个延时步骤，就是让报警持续一段时间，其后续步骤为步骤B7。
步骤B7：计时器清零，后续步骤为步骤B8。当一次报警完成之后，我们要重新开始检测流程，这个时候需要将计时器清零，否则当被检测者没有调整坐姿时，下一个错误检测出现到达步骤B4时，计时器的时长必定超过阈值，就会直接报警，这样报警就过于频繁了，达不到我们希望的当状态持续时才报警，且两次报警需间隔一定时间的效果。
步骤B8：报警判断流程结束，后续执行姿势判断流程。不管前面的流程是报警还是不报警，步骤B8是报警判断流程的终点，各分支最后都走到这一步骤，结束报警判断流程。
在本发明另一个较佳实施例中，我们称为第二实施例中，报警判断流程并不像上一个实施例中一样采用计时的方式，而是采用计次的方式，即只有当错误姿态连续出现次数达到设定的值才报警，中间有正确姿势的则清零重新计次，具体流程如图3所示，包括如下步骤：
步骤C1：判断是否连续两次检测为错误坐姿，如是则执行步骤C3，如否则执行步骤C2；
步骤C2：计数器从零开始计数，后续步骤为步骤C8；
步骤C3：计数器累加，后续步骤为步骤C4；
步骤C4：判断计数器的值是否超过阈值，如是执行步骤C5，如否执行步骤C8；
步骤C5：报警单元报警，后续步骤为步骤C6；
步骤C6：延时，后续步骤为步骤C7；
步骤C7：计数器清零，后续步骤为步骤C8；
步骤C8：报警判断流程结束，后续执行姿势判断流程。
如前所述我们判断坐姿是否正确是通过判断姿势特征信号S2和标准姿势数据S3之间的差值是否超过阈值来实现的，姿势特征信号S2是传感单元读数后计算获得的，而标准姿势数据S3可以是预先设置的一个值，其存放在存储单元之中内。而实际上每一个人因为身形、习惯不同，每一个人的坐姿都不尽相同，采用统一的标准姿势数据S3可能并不是很适合。为了解决这个问题，我们在姿势判断流程之前加入一个初始化流程，这个初始化流程用来测量并计算标准姿势数据S3。这个流程的设计是这样的，在开始使用时，先通过提醒让使用者达到一个标准的坐姿，然后检测在这个姿态下的传感单元数据，并以此计算出标准姿势数据S3存入存储单元，后续的判断都以这个值为基础。
具体的说，初始化流程包括如下步骤：
步骤D1：传感单元读取数据，后续步骤为步骤D2。
步骤D2：判断所述传感单元是否处于相对静止状态，如是执行步骤D3，不是相对静止状态则返回执行步骤D1。这里我们一定要使用者进入读写这种静止状态时才可以测量，如人还没有静止下来还处于准备工作时不进行测定。
步骤D3：判断所述传感单元测得值是否符合标准坐姿范围，如是执行步骤D4，不符合则发出提醒并返回执行步骤D1。虽然标准姿势数据S3是通过初始化流程才计算出来，但是我们也要保证这个测得值确实是一个标准的坐姿，因此在获得标准姿势数据S3我们还是有一个判断的，预先设置一个值，测得值相对于这个值不能偏离太多，如测得值在可允许的范围内才记录这个值作为标准姿势数据S3，超出范围了就发出提醒后继续进行传感器读数，在本实施例中采用语音提醒，提示请采用标 准坐姿并保持，在其他实施例中也可以采用亮度变化等方式，如未能测得正确坐姿灯处于一个较暗的状态作为提示，只有姿势正确了，并记录下了这个正确数据，灯光变亮表示进入正式读写状态。
步骤D4：根据传感器当前度数计算标准姿势数据S3，并存入存储单元，初始化流程结束，后续进入正常工作状态，执行姿势判断流程。
上文对本发明优选实施例的描述是为了说明和描述，并非想要把本发明穷尽或局限于所公开的具体形式，显然，可能做出许多修改和变化，这些修改和变化可能对于本领域技术人员来说是显然的，应当包括在由所附权利要求书定义的本发明的范围之内。

Claims (13)

  1. 一种坐姿识别系统，所述坐姿识别系统包括：
    传感单元，使用时所述传感单元佩戴于人体上，输出传感信号；
    姿势运算单元，从所述传感单元接收所述传感信号，运算获得姿势特征信号，并在人体处于相对静止时输出所述姿势特征信号；
    存储单元，用于存储标准姿势数据；
    姿势判断单元，从所述姿势运算单元接收所述姿势特征信号，并读取所述存储单元中的所述标准姿势数据，判断坐姿，输出坐姿正误信息；
    报警判断单元，接收所述姿势判断单元的所述坐姿正误信息，判断是否需要报警，在需要报警时输出报警信号；
    报警单元，接收所述报警信号，并报警。
  2. 根据权利要求1所述的坐姿识别系统，其中所述传感单元为加速度传感器。
  3. 根据权利要求2所述的坐姿识别系统，其中所述传感单元为三轴加速度传感器。
  4. 根据权利要求1所述的坐姿识别系统，其中所述报警单元发出的警示信息为声音、振动、色彩、亮度中的一种或多种的组合。
  5. 根据权利要求1所述的坐姿识别系统，其中所述传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元之间物理连接为一体。
  6. 根据权利要求1所述的坐姿识别系统，其中所述传感单元、姿势运算单元、存储单元、姿势判断单元、报警判断单元、报警单元分别设置在两个以上分离的部分，所述分离的部分之间通过无线通信方式传递信息。
  7. 根据权利要求6所述的坐姿识别系统，其中所述无线通信方式为ZigBee、蓝牙或WiFi。
  8. 一种坐姿识别方法，应用于如权利要求1-7所述的坐姿识别系统的，所述坐姿识别方法包括姿势判断流程和报警判断流程，在坐姿识别过程中循环执行所述姿势判断流程，当出现错误坐姿时执行所述报警判断流程，判断是否需要报警，无论是否报警，在所述报警判断流程执行完毕后继续执行所述姿势判断流程。
  9. 根据权利要求8所述的坐姿识别方法，其中所述姿势判断流程包括如下步骤：
    步骤A1：所述传感单元读取数据，后续步骤为步骤A2；
    步骤A2：判断所述传感单元是否处于相对静止状态，如是执行步骤A3，不是相对静止状态则返回执行步骤A1；
    步骤A3：姿势运算单元计算姿势特征信号，后续步骤为步骤A4；
    步骤A4：计算所述姿势特征信号和标准姿势数据之间的差值，后续步骤为步骤A5；
    步骤A5：判断所述姿势特征信号和标准姿势数据之间的差值是否超过阈值，如是则执行所述报警判断流程，没有超过阈值则返回再次执行所述步骤A1。
  10. 根据权利要求8所述的坐姿识别方法，其中所述报警判断流程包括如下步骤：
    步骤B1：判断是否连续两次检测为错误坐姿，如是则执行步骤B3，如否则执行步骤B2；
    步骤B2：计时器清零并开始计时，后续步骤为步骤B8；
    步骤B3：计时器累计继续计时，后续步骤为步骤B4；
    步骤B4：判断计时器时长是否超过阈值，如是执行步骤B5，如否执行步骤B8；
    步骤B5：报警单元报警，后续步骤为步骤B6；
    步骤B6：延时，后续步骤为步骤B7；
    步骤B7：计时器清零，后续步骤为步骤B8；
    步骤B8：报警判断流程结束，后续执行姿势判断流程。
  11. 根据权利要求8所述的坐姿识别方法，其中所述报警判断流程包括如下步骤：
    步骤C1：判断是否连续两次检测为错误坐姿，如是则执行步骤C3，如否则执行步骤C2；
    步骤C2：计数器从零开始计数，后续步骤为步骤C8；
    步骤C3：计数器累加，后续步骤为步骤C4；
    步骤C4：判断计数器的值是否超过阈值，如是执行步骤C5，如否执行步骤C8；
    步骤C5：报警单元报警，后续步骤为步骤C6；
    步骤C6：延时，后续步骤为步骤C7；
    步骤C7：计数器清零，后续步骤为步骤C8；
    步骤C8：报警判断流程结束，后续执行姿势判断流程。
  12. 根据权利要求8、9、10或11所述的坐姿识别方法，其中所述姿势判断流程之前还包括初始化流程，所述初始化流程测量并计算标准姿势数据。
  13. 根据权利要求12所述的坐姿控制方法，其中所述初始化流程包括如下步骤：
    步骤D1：所述传感单元读取数据，后续步骤为步骤D2；
    步骤D2：判断所述传感单元是否处于相对静止状态，如是执行步骤D3，不是相对静止状态则返回执行步骤D1；
    步骤D3：判断所述传感单元测得值是否符合标准坐姿范围，如是执行步骤D4，不符合则发出提醒并返回执行步骤D1；
    步骤D4：根据传感器当前度数计算所述标准姿势数据，后续执行姿势判断流程。
PCT/CN2017/116162 2016-12-22 2017-12-14 一种坐姿识别系统和坐姿识别方法 WO2018113582A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611195020.4A CN108209929B (zh) 2016-12-22 2016-12-22 一种坐姿识别系统和坐姿识别方法
CN201611195020.4 2016-12-22

Publications (1)

Publication Number Publication Date
WO2018113582A1 true WO2018113582A1 (zh) 2018-06-28

Family

ID=62624424

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/116162 WO2018113582A1 (zh) 2016-12-22 2017-12-14 一种坐姿识别系统和坐姿识别方法

Country Status (2)

Country Link
CN (1) CN108209929B (zh)
WO (1) WO2018113582A1 (zh)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111062244A (zh) * 2019-10-31 2020-04-24 河南科技大学 一种基于机器视觉的坐姿提醒方法
CN112733626A (zh) * 2020-12-28 2021-04-30 歌尔光学科技有限公司 基于手环的坐姿提醒方法、手环及计算机可读存储介质
CN113312938A (zh) * 2020-02-26 2021-08-27 北京君正集成电路股份有限公司 一种坐姿检测防检测器前无目标时产生误报的方法及系统
CN113313917A (zh) * 2020-02-26 2021-08-27 北京君正集成电路股份有限公司 一种解决坐姿检测中检测器前无目标时产生误报的方法
CN113378762A (zh) * 2021-06-25 2021-09-10 宁波星巡智能科技有限公司 坐姿智能监测方法、装置、设备及存储介质
CN113591522A (zh) * 2020-04-30 2021-11-02 百度在线网络技术（北京）有限公司 图像处理方法、装置及存储介质

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2020039674A (ja) * 2018-09-12 2020-03-19 トヨタ紡織株式会社 シート及び姿勢推定システム
CN109102682A (zh) * 2018-09-28 2018-12-28 朱晓 坐姿控制系统
CN109584521A (zh) * 2018-10-15 2019-04-05 西安理工大学 一种基于Notch传感器的不良坐姿监测方法
CN113892920B (zh) * 2020-07-06 2023-06-27 华为技术有限公司 可穿戴设备的佩戴检测方法、装置及电子设备

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003070596A (ja) * 2001-08-31 2003-03-11 Itoki Crebio Corp 着座姿勢検出機能付き椅子
CN103400476A (zh) * 2013-07-30 2013-11-20 步步高教育电子有限公司 一种坐姿提醒方法及装置
CN104916093A (zh) * 2015-06-05 2015-09-16 重庆智韬信息技术中心 实现学生坐姿智能矫正提醒的方法
CN105389960A (zh) * 2015-11-16 2016-03-09 上海斐讯数据通信技术有限公司 一种坐姿检测装置、系统及方法
CN105844867A (zh) * 2016-03-18 2016-08-10 广东小天才科技有限公司 一种矫正用户坐姿的方法及装置
CN105894752A (zh) * 2014-11-10 2016-08-24 王恒厂 一种多功能坐姿矫正装置
CN106643708A (zh) * 2016-09-21 2017-05-10 苏州坦特拉自动化科技有限公司 一种基于imu的交互式坐姿矫正装置、坐姿矫正器及监测软件

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008264188A (ja) * 2007-04-19 2008-11-06 Sanko:Kk ゲーム使用時間カウント機能付き学習机
CN104484984A (zh) * 2014-12-17 2015-04-01 天津工业大学 可调节的人体姿势检测提醒装置
CN104900019A (zh) * 2015-04-23 2015-09-09 上海维锐智能科技有限公司 一种人体坐姿探测装置及探测方法
CN205810084U (zh) * 2016-07-06 2016-12-14 欧普照明股份有限公司 一种坐姿矫正设备及照明设备

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003070596A (ja) * 2001-08-31 2003-03-11 Itoki Crebio Corp 着座姿勢検出機能付き椅子
CN103400476A (zh) * 2013-07-30 2013-11-20 步步高教育电子有限公司 一种坐姿提醒方法及装置
CN105894752A (zh) * 2014-11-10 2016-08-24 王恒厂 一种多功能坐姿矫正装置
CN104916093A (zh) * 2015-06-05 2015-09-16 重庆智韬信息技术中心 实现学生坐姿智能矫正提醒的方法
CN105389960A (zh) * 2015-11-16 2016-03-09 上海斐讯数据通信技术有限公司 一种坐姿检测装置、系统及方法
CN105844867A (zh) * 2016-03-18 2016-08-10 广东小天才科技有限公司 一种矫正用户坐姿的方法及装置
CN106643708A (zh) * 2016-09-21 2017-05-10 苏州坦特拉自动化科技有限公司 一种基于imu的交互式坐姿矫正装置、坐姿矫正器及监测软件

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111062244A (zh) * 2019-10-31 2020-04-24 河南科技大学 一种基于机器视觉的坐姿提醒方法
CN111062244B (zh) * 2019-10-31 2023-04-18 河南科技大学 一种基于机器视觉的坐姿提醒方法
CN113312938A (zh) * 2020-02-26 2021-08-27 北京君正集成电路股份有限公司 一种坐姿检测防检测器前无目标时产生误报的方法及系统
CN113313917A (zh) * 2020-02-26 2021-08-27 北京君正集成电路股份有限公司 一种解决坐姿检测中检测器前无目标时产生误报的方法
CN113313917B (zh) * 2020-02-26 2022-12-16 北京君正集成电路股份有限公司 一种解决坐姿检测中检测器前无目标时产生误报的方法
CN113591522A (zh) * 2020-04-30 2021-11-02 百度在线网络技术（北京）有限公司 图像处理方法、装置及存储介质
CN112733626A (zh) * 2020-12-28 2021-04-30 歌尔光学科技有限公司 基于手环的坐姿提醒方法、手环及计算机可读存储介质
CN112733626B (zh) * 2020-12-28 2023-02-28 歌尔科技有限公司 基于手环的坐姿提醒方法、手环及计算机可读存储介质
CN113378762A (zh) * 2021-06-25 2021-09-10 宁波星巡智能科技有限公司 坐姿智能监测方法、装置、设备及存储介质
CN113378762B (zh) * 2021-06-25 2023-08-08 宁波星巡智能科技有限公司 坐姿智能监测方法、装置、设备及存储介质

Also Published As

Publication number Publication date
CN108209929A (zh) 2018-06-29
CN108209929B (zh) 2021-08-13

Similar Documents

Publication Publication Date Title
WO2018113582A1 (zh) 一种坐姿识别系统和坐姿识别方法
JP6281290B2 (ja) リストバンド型腕動作判定装置
WO2021093896A1 (zh) 一种基于智能手表的读写距离识别方法
US10422814B2 (en) Fall detection using machine learning
CN104424771B (zh) 一种告警方法及装置
CN109116586B (zh) 多功能智能眼镜
US20120274554A1 (en) Body movement detection device and display control method of body movement detection device
CN101536053A (zh) 预防跌倒的系统以及采用这样的系统预防跌倒的方法
CN108665687A (zh) 一种坐姿监测方法和装置
US10653353B2 (en) Monitoring a person for indications of a brain injury
WO2020228307A1 (zh) 一种跌倒检测方法、装置及可穿戴式设备
CN107949320A (zh) 一种血压监测前进行提醒的方法及相应的动态血压监测仪
US9901291B2 (en) Activity meter and sleep/awake state recording system
KR20140100076A (ko) 신체에 걸리는 방식으로 착용되는 낙상 감지 장치 및 그 방법 그리고 이를 이용한 낙상 사고 통합 관리 시스템 및 낙상 사고 관리 방법
CN109799624A (zh) 一种智能儿童护眼镜及终端设备
CN109528209A (zh) 穿戴式设备的运动状态提醒方法、装置、设备及存储介质
WO2017032161A1 (zh) 一种监控形体姿态的方法和装置
KR20210031221A (ko) 사용자의 머리에 착용되는 자세 센서의 전후 좌우 방향 결정 방법
CN109141346A (zh) 穿戴式设备的坐姿更正引导方法、装置、设备及存储介质
US11369316B2 (en) Heart rate detection architecture
US20090063088A1 (en) Wristwatch type acceleration detection module
CN113040753A (zh) 坐姿监测系统
CN113053074A (zh) 坐姿监测方法
CN109697835A (zh) 一种监控坐姿和预防近视的方法与装置
KR20150004524A (ko) 목 디스크 예방을 위한 모니터 사용자의 자세 측정 시스템 및 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17884488

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17884488

Country of ref document: EP

Kind code of ref document: A1