Posture recognition and correction method, apparatus, system, and electronic device

Abstract

The application discloses a posture recognition and correction method applied to an electronic device. The posture recognition and correction method comprises the following steps: acquiring first environment distance information, wherein the first environment distance information is distance information between the electronic device and surrounding environment objects; determining first human body posture feature information and first human body distance feature information according to the first environment distance information; and if it is determined that the user posture is incorrect according to the first human body posture feature information and the first human body distance feature information and preset correct human body posture feature information and correct human body distance feature information, generating a correction reminding information. The posture recognition and correction method can more accurately and conveniently recognize the posture according to the human body posture feature information and the human body distance feature information, improves the robustness and accuracy of the posture recognition, and effectively improves the user experience. The application further discloses a posture recognition and correction device, a posture recognition and correction system and an electronic device.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a posture recognition and correction method, apparatus, system, and electronic device. Background Technology

[0002] In daily work, study, exercise, and life, it is necessary to maintain correct postures (also known as stances) in sitting, standing, walking, sleeping, and exercising. Correct posture is beneficial to maintaining physical health, while incorrect posture can negatively impact it. Especially now that people are spending increasingly more time using electronic devices such as mobile phones and computers, the importance of correct posture is even greater.

[0003] Currently, there are many types of devices used for posture correction. Common examples include physical correction devices such as "BackJoy" and posture correction desks, as well as wearable devices such as glasses, watches, and wristbands that correct posture by detecting distance parameters between the wearable device and the user or the user's head angle parameters. These devices suffer from inaccurate posture recognition and correction, as well as inconvenience, resulting in a poor user experience. Summary of the Invention

[0004] This application provides a posture recognition and correction method, apparatus, system, and electronic device to solve the problems of inaccurate and inconvenient posture recognition and correction. It enables more accurate and convenient posture recognition and correction, improving the user experience.

[0005] To address the aforementioned technical problems, in a first aspect, embodiments of this application provide a posture recognition and correction method applied to an electronic device. The method includes: acquiring first environmental distance information, where the first environmental distance information is the distance information between the electronic device and surrounding environmental objects; determining first human posture feature information and first human distance feature information based on the first environmental distance information, where the first human posture feature information represents the user's current posture, and the first human distance feature information represents the current distance between the user and the electronic device; if it is determined that the user's posture is incorrect based on the first human posture feature information, the first human distance feature information, and preset correct human posture feature information and correct human distance feature information, a correction reminder message is generated, where the correct human posture feature information represents the user's correct posture, and the correct human distance feature information represents the correct distance between the user and the electronic device.

[0006] By comparing the real-time determined first human posture feature information and first human distance feature information with the preset correct human posture feature information and correct human distance feature information, the user's posture can be judged to be correct. This allows for more accurate and convenient posture recognition, improving the robustness and accuracy of posture recognition and effectively enhancing the user experience.

[0007] In one possible implementation of the first aspect described above, the first environmental distance information can be an environmental distance matrix. For example, the environmental distance matrix can be KM = [K1, K2, K3...Kn], where K1, K2, K3...Kn represent multiple distance information between the watch 100 and surrounding environmental objects.

[0008] In one possible implementation of the first aspect above, the method further includes: if it is determined from the first environmental distance information that the surrounding environmental objects include the user, determining the user's first human posture feature information and first human distance feature information from the first environmental distance information.

[0009] Determining whether surrounding environmental objects include users based on the first environmental distance information can be achieved by using machine learning algorithms to determine the outline information of surrounding environmental objects based on the first environmental distance information to determine the type or category of surrounding environmental objects, and further determining whether surrounding environmental objects include human body outlines to determine whether users exist in the surrounding environment.

[0010] In one possible implementation of the first aspect above, if the working mode is automatic mode, the method further includes determining whether the user's posture is correct by: determining a first posture mode based on first human posture feature information and first human distance feature information; determining whether the user's posture is correct based on the first human posture feature information and first human distance feature information, as well as the pre-stored correct human posture feature information and correct human distance feature information corresponding to the first posture mode.

[0011] In one possible implementation of the first aspect above, the method further includes determining the operating mode by displaying a first menu, the first menu including an automatic mode control; and determining the operating mode as automatic mode if a user selection operation on the automatic mode control is detected.

[0012] Users can manually set the working mode. In automatic working mode, the electronic device determines the user's current first posture mode based on the first human posture feature information and the first human distance feature information, and determines whether the user's current posture is correct, which can improve the user experience.

[0013] In one possible implementation of the first aspect described above, if the working mode is a non-automatic mode, the method further includes determining whether the user's posture is correct by: determining a first posture mode based on first human posture feature information and first human distance feature information; if the first posture mode is inconsistent with a preset second posture mode, generating a mode switching reminder message for the user to perform a posture mode selection operation; if the user's selection operation for the first posture mode is received, determining whether the user's posture is correct based on the first human posture feature information, the first human distance feature information, and the pre-stored correct human posture feature information and correct human distance feature information corresponding to the first posture mode; if the user's selection operation for the second posture mode is received, determining whether the user's posture is correct based on the first human posture feature information, the first human distance feature information, and the pre-stored correct human posture feature information and correct human distance feature information corresponding to the second posture mode.

[0014] In one possible implementation of the first aspect above, the method further includes determining the working mode by: displaying a first menu, the first menu including at least one posture mode control; if a user's selection operation on the posture mode control is detected, determining that the working mode is a non-automatic mode, and determining that the posture mode corresponding to the posture mode control is a preset second posture mode.

[0015] Users can manually set the operating mode. For example, a user can select a certain posture mode as the second posture mode, in which case the electronic device is in a non-automatic operating mode. In the non-automatic operating mode, the electronic device determines the user's current first posture mode based on the first human posture feature information and the first human distance feature information, and determines whether the first posture mode is consistent with the second posture mode selected by the user. If they are inconsistent, the electronic device can remind the user to switch posture modes, and determine whether the posture is correct based on the user's final selected posture mode, which can improve the user experience.

[0016] In one possible implementation of the first aspect described above, the posture mode control can be any one of a reading mode control, a writing mode control, and a computer viewing mode control. The posture mode control can also be other controls, such as those for viewing a tablet computer.

[0017] In one possible implementation of the first aspect described above, the method further includes recording the correct posture to obtain correct human posture feature information and correct human distance feature information by: displaying a second menu, the second menu including at least one posture mode control; if a user's selection operation on the posture mode control is received, acquiring second environmental distance information, the second environmental distance information being the distance information between the electronic device and surrounding environmental objects; determining second human posture feature information and second human distance feature information based on the second environmental distance information, the second human posture feature information representing the user's current posture, and the second human distance feature information representing the current distance between the user and the electronic device; using the second human posture feature information as the correct human posture feature information, and using the second human distance feature information as the correct human distance feature information.

[0018] The electronic device extracts human posture feature information and human distance feature information based on a second environmental distance matrix, which are then used as the second human posture feature information and the second human distance feature information, respectively. These are then stored as the correct human posture feature information and the correct human distance feature information, respectively. During posture recognition and correction, the electronic device can compare the obtained first human posture feature information and first human distance feature information corresponding to the user's current posture with the stored correct human posture feature information and correct human distance feature information to identify posture anomalies and determine whether the user's posture is correct. This effectively improves the dynamic range and adaptability of the electronic device and enhances the accuracy of posture recognition.

[0019] In one possible implementation of the first aspect above, the method further includes: displaying a reminder message to continue recording the correct posture for the user to select the correct posture for recording; if the user selects to continue recording the correct posture, continuing to record the correct posture information; if the user cancels the selection to continue recording the correct posture, ending the recording of the correct posture information.

[0020] For a given posture mode, users can continue to input reminder information by maintaining the correct posture. They can input the correct posture once or multiple times as needed, which can effectively improve the user experience.

[0021] In one possible implementation of the first aspect described above, the method further includes: performing a corresponding reminder operation based on the correction reminder information; and / or sending the correction reminder information to other electronic devices that have established a communication connection with the electronic device, so that the other electronic devices perform the corresponding reminder operation based on the correction reminder information. The electronic device can provide correction reminders on its own, or it can link with other devices to provide correction reminders, allowing for convenient correction reminders and achieving the goal of more timely reminders to the user to adjust their posture.

[0022] In one possible implementation of the first aspect described above, the method further includes: transmitting an ultrasonic signal to the surrounding environment; receiving an ultrasonic signal reflected back from an object in the surrounding environment; and obtaining distance information between the electronic device and the object in the surrounding environment based on the time of transmitting the ultrasonic signal, the time of receiving the reflected ultrasonic signal, and the transmission rate of the ultrasonic signal.

[0023] Electronic devices emit ultrasonic signals in different directions towards their surroundings. Ultrasonic detection technology can then be used to obtain multiple distance parameters between the electronic device and objects in the environment, resulting in the aforementioned environmental distance matrix. This environmental distance matrix can be used to determine the types of objects in the surrounding environment and the user's posture, among other things.

[0024] Secondly, embodiments of this application provide a posture recognition and correction device, comprising: an ultrasonic module for transmitting and receiving ultrasonic signals to obtain first environmental distance information, wherein the first environmental distance information is the distance information between the posture recognition and correction device and surrounding environmental objects; an information storage module for storing preset correct human posture feature information and correct human distance feature information, wherein the correct human posture feature information represents the correct posture of the user, and the correct human distance feature information represents the correct distance between the user and the electronic device; and an information processing module for determining the first human posture feature information and the first human distance feature information based on the first environmental distance information, and generating correction reminder information if it is determined that the user's posture is incorrect based on the first human posture feature information and the first human distance feature information, as well as the correct human posture feature information and the correct human distance feature information stored in the information storage module, wherein the first human posture feature information represents the user's current posture, and the first human distance feature information represents the current distance between the user and the electronic device.

[0025] By comparing the real-time determined first human posture feature information and first human distance feature information with the preset correct human posture feature information and correct human distance feature information, the user's posture can be judged to be correct. This allows for more accurate and convenient posture recognition, improving the robustness and accuracy of posture recognition and effectively enhancing the user experience.

[0026] In one possible implementation of the second aspect above, there is one ultrasonic module, which is used to send ultrasonic signals in different directions according to a preset period; or there are multiple ultrasonic modules, which are used to send ultrasonic signals in different directions simultaneously.

[0027] By emitting ultrasonic signals in different directions towards the surrounding environment, ultrasonic detection technology can be used to obtain multiple distance information between the posture recognition and correction device and objects in the surrounding environment, thus generating the aforementioned environmental distance matrix. This environmental distance matrix can be used to determine the types of objects in the surrounding environment and the user's posture, among other things.

[0028] The posture recognition and correction apparatus provided in this application includes a module for performing the posture recognition and correction method provided by the first aspect and / or any possible implementation of the first aspect, and therefore can also achieve the beneficial effects (or advantages) of the posture recognition and correction method provided by the first aspect.

[0029] Thirdly, embodiments of this application provide a posture recognition and correction system, including a first electronic device and a second electronic device that establish a communication connection. The first electronic device includes an ultrasonic module for transmitting and receiving ultrasonic signals to obtain first environmental distance information, which is the distance information between the posture recognition and correction device and surrounding environmental objects. The second electronic device includes an information storage module and an information processing module. The information storage module is used to store preset correct human posture feature information and correct human distance feature information. The information processing module is used to determine the first human posture feature information and the first human distance feature information based on the first environmental distance information. If it is determined that the user's posture is incorrect based on the first human posture feature information, the first human distance feature information, and the correct human posture feature information and the correct human distance feature information stored in the information storage module, a correction reminder message is generated. The first human posture feature information represents the user's posture, the first human distance feature information represents the distance between the user and the electronic device, the correct human posture feature information represents the user's correct posture, and the correct human distance feature information represents the correct distance between the user and the electronic device.

[0030] The posture recognition and correction system provided in this application includes a first electronic device and a second electronic device for performing the posture recognition and correction method provided by the first aspect and / or any possible implementation of the first aspect, and thus can also achieve the beneficial effects (or advantages) of the posture recognition and correction method provided by the first aspect.

[0031] Fourthly, embodiments of this application provide an electronic device, including: a memory for storing a computer program, the computer program including program instructions; and a control unit for executing the program instructions to cause the electronic device to perform the posture recognition and correction method provided by the first aspect and / or any possible implementation of the first aspect.

[0032] Fifthly, embodiments of this application provide a computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, the computer program including program instructions, the program instructions being executed by a computer to cause the computer to perform the posture recognition and correction method provided as described in the first aspect and / or any possible implementation of the first aspect. Attached Figure Description

[0033] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0034] Figure 1A and Figure 1B Based on some embodiments of this application, some application scenarios of the posture recognition and correction method provided in this application are shown;

[0035] Figure 2 A schematic diagram of the structure of a watch 100 is shown according to some embodiments of this application;

[0036] Figure 3A-3G This is a schematic diagram of the display interface of a watch 100 when recording the correct posture, according to some embodiments of this application;

[0037] Figure 4A-4I These are schematic diagrams of the display interface of a watch 100 when performing posture recognition and correction, according to some embodiments of this application.

[0038] Figure 5 This is a flowchart illustrating a correct posture input process according to some embodiments of this application;

[0039] Figure 6 According to some embodiments of this application, a flowchart illustrating a process for performing posture recognition and correction is shown;

[0040] Figure 7 This is a flowchart illustrating a process for anomaly identification according to some embodiments of this application;

[0041] Figure 8 The present application provides some embodiments of a posture recognition and correction device, which is shown in the following schematic diagram.

[0042] Figure 9 The present application provides some embodiments of a posture recognition and correction system, and provides a structural schematic diagram of such a system.

[0043] Figure 10 The present application provides some embodiments of an electronic device, which are shown in the following schematic diagram.

[0044] Figure 11 The present application illustrates a schematic diagram of a system-on-a-chip (SoC) structure according to some embodiments thereof. Detailed Implementation

[0045] The technical solution of this application will now be described more clearly and completely with reference to the accompanying drawings.

[0046] This application describes the technical solution of this application in detail using the scenario of posture recognition and correction.

[0047] Myopia is currently the most common eye disease, affecting an estimated 1.5 billion people worldwide. It can occur in all regions and among all ethnic groups. Research indicates that most cases of myopia are non-hereditary, and their occurrence is closely related to close-range visual activity. This is especially true for teenagers, whose eyes are still developing and whose eye walls are more elastic. Coupled with their relatively poor self-control, teenagers are prone to incorrect posture, making them highly susceptible to myopia if they spend long periods reading or writing at close range. Preventing non-hereditary myopia has become one of the top priorities for teenagers besides their studies. Furthermore, for working professionals, prolonged computer use, especially with incorrect posture, can easily lead to eye strain and other health problems, such as affecting the cervical spine.

[0048] Currently, there are three methods for correcting sitting posture: using physical devices, using head-mounted wearable devices, and using hand-mounted wearable devices.

[0049] Physical devices can be used to correct posture, such as tables with railings or supports (e.g., corrective desks) to ensure a certain distance between the user's head and the desktop. Alternatively, a posture corrector or chair with a posture corrector can be used, using the posture corrector's straps to restrain the user's body and force them to maintain a certain distance from the desktop.

[0050] Posture correction devices, such as "BackJoy" (a posture corrector), have limited adjustability and a restrictive feel due to individual differences in body shape and habits, often causing discomfort during correction. Correction desks confine users to a limited space, making movement inconvenient; they are also less portable and require desk modifications, resulting in higher costs.

[0051] This method utilizes a wearable head-mounted device to correct posture. The user wears the device on their head, which detects the distance between the device and objects in front of them, as well as the tilt angle of the user's head. When the distance is less than a distance threshold or the tilt angle is greater than a tilt angle threshold, the device triggers an alarm, prompting the user to adjust their posture. The wearable head-mounted device can be, for example, an earring, a headband, or glasses.

[0052] Head-worn devices for correcting posture rely on a single detection parameter—distance and angle—making them inaccurate in identifying sitting posture and prone to missed or false detections. Furthermore, the limited dynamic range and adaptability of head-worn devices limit their application to users with varying body types. Additionally, they can be inconvenient for some users, such as those who don't need glasses but require them to wear glasses.

[0053] This solution uses a wearable hand device to correct posture. The user wears the device on their hand, and it detects the distance between the device and objects in front of them. When the distance is less than a threshold, an alarm is triggered, prompting the user to adjust their posture. The wearable hand device can be, for example, a wristband or a watch.

[0054] The solution of correcting sitting posture using a wearable hand device has limitations. Because the wearable hand device detects only a single parameter—distance information—it cannot accurately identify sitting posture, easily leading to missed or false detections. Furthermore, due to differences in body shape, the wearable hand device has a limited dynamic range and weak adaptability.

[0055] This application provides a posture recognition and correction method that can be applied to a watch used as a wearable hand device. See also... Figure 1A and Figure 1B , Figure 1A and Figure 1B A schematic diagram illustrating a usage scenario for watch 100 is shown. In this scenario, the user (person) wears watch 100 and is viewing computer 200. Watch 100 is positioned between the user's body and computer 200. Figure 1A The image shows the correct posture for viewing a computer screen (computer 200). Figure 1B An incorrect posture is a posture that is not in the correct position.

[0056] Watch 100 includes an ultrasonic module. Watch 100 emits ultrasonic signals to different directions in its surrounding environment via this module. Objects in the surrounding environment, such as the user's body, computer 200, and a table, reflect these ultrasonic signals. After receiving the ultrasonic signals reflected back from these objects, watch 100 uses ultrasonic detection technology to determine multiple distance parameters representing the distance between the watch 100 and the objects. For example, during use of watch 100, the user... Figure 1B If the computer 200 is viewed in the posture shown, then the watch 100 can obtain the following: Figure 1BThe watch 100 is shown to have multiple distance information K1, K2, K3, and K4 between itself and the user, multiple distance information K5, K6, and K7 between itself and the computer 200, and multiple distance information between itself and other objects in its surroundings. Then, the watch 100 obtains an environmental distance matrix including these multiple distance information as first environmental distance information. The environmental distance matrix can be, for example, KM = [K1, K2, K3…Kn], where K1, K2, K3…Kn are multiple distance information between the watch 100 and surrounding environmental objects. Furthermore, the watch 100 extracts the user's posture information (e.g., posture contour) as first human posture feature information based on this environmental distance matrix, and extracts distance information between itself and its user (e.g., the distance between the watch 100 and the user's head) as first human distance feature information. Next, the watch 100 analyzes and compares the first human posture feature information and the first human distance feature information, as well as the preset correct human posture feature information and the correct human distance feature information, to identify posture anomalies and determine whether the user's posture is correct. That is, the posture recognition and correction method provided in this application allows the watch 100 to determine whether the user's posture is correct through human posture feature information and human distance feature information, enabling more accurate and convenient posture recognition, improving the robustness and accuracy of posture recognition, and effectively enhancing the user experience.

[0057] The posture recognition and correction method provided in this application allows the watch 100 to receive the user's correct posture input (or correct posture learning) to obtain correct human posture feature information and correct human distance feature information. For example, the user can pose in a certain way. Figure 1AAs shown in the correct posture, watch 100 obtains correct human posture feature information and correct human distance feature information based on the user's correct posture. Specifically, watch 100 uses ultrasonic detection technology to obtain distance information between watch 100 and surrounding objects, and forms an environmental distance matrix as the second environmental distance information. Watch 100 extracts the user's posture information as the second human posture feature information and the distance information between watch 100 and the user as the second human distance feature information based on this environmental distance matrix. These two information are then stored as the correct human posture feature information and the correct human distance feature information, respectively. During posture recognition and correction, watch 100 analyzes and compares the obtained first human posture feature information and first human distance feature information corresponding to the user's current posture with the stored correct human posture feature information and correct human distance feature information to identify posture anomalies and determine whether the user's posture is correct. This effectively improves the dynamic range and adaptability of watch 100 and enhances the accuracy of posture recognition.

[0058] The posture recognition and correction method provided in this application allows the watch 100 to generate correction reminders when it determines that the user's posture is incorrect. Firstly, the watch 100 can trigger an alarm, such as by vibrating or emitting a sound, based on the correction reminder. Secondly, the watch 100 can use other electronic devices (such as a computer 200 or a mobile phone) that have established a communication connection with it (e.g., Bluetooth or WiFi) or other electronic devices bound to an account as linked devices. Correction reminders can be completed through linkage with these linked devices. For example, the watch 100 sends correction reminder information to the linked device, causing the linked device to perform corresponding reminder operations, such as displaying a reminder on the linked device, turning off the linked device, or locking the linked device's screen. This facilitates convenient correction reminders, enabling more timely reminders for users to adjust their posture.

[0059] Please see Figure 2 , Figure 2 A structural schematic diagram of watch 100 is shown.

[0060] The watch 100 may include a processor 110, an external memory interface 120, an internal memory 121, a universal serial bus (USB) connector 130, a charging management module 140, a power management module 141, a battery 142, an antenna 1, an antenna 2, a mobile communication module 150, a wireless communication module 160, an audio module 170, a speaker 170A, a receiver 170B, a microphone 170C, a headphone jack 170D, a sensor module 180, buttons 190, a motor 191, an indicator 192, a camera 193, a display screen 194, and a subscriber identification module (SIM) card interface 195, etc. The sensor module 180 may include a pressure sensor 180A, a gyroscope sensor 180B, a barometric pressure sensor 180C, a magnetic sensor 180D, an accelerometer sensor 180E, a distance sensor 180F, a proximity sensor 180G, a fingerprint sensor 180H, a temperature sensor 180J, a touch sensor 180K, an ambient light sensor 180L, a bone conduction sensor 180M, and an ultrasonic sensor 180N, etc.

[0061] It is understood that the structures illustrated in the embodiments of the present invention do not constitute a specific limitation on the watch 100. In other embodiments of this application, the watch 100 may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0062] Processor 110 may include one or more processing units, such as application processors (APs), modem processors, graphics processing units (GPUs), image signal processors (ISPs), controllers, video codecs, digital signal processors (DSPs), baseband processors, and / or neural network processing units (NPUs). These different processing units may be independent devices or integrated into one or more processors.

[0063] The processor can generate operation control signals based on the instruction opcode and timing signals to control the instruction fetching and execution.

[0064] The processor 110 may also include a memory for storing instructions and data. In some embodiments, the memory in the processor 110 is a cache memory. This memory can store instructions or data that the processor 110 has just used or that are used repeatedly. If the processor 110 needs to use the instruction or data again, it can retrieve it directly from the memory. This avoids repeated accesses, reduces the waiting time of the processor 110, and thus improves the efficiency of the system.

[0065] The wireless communication function of watch 100 can be achieved through antenna 1, antenna 2, mobile communication module 150, wireless communication module 160, modem processor and baseband processor.

[0066] Antennas 1 and 2 are used to transmit and receive electromagnetic wave signals. Each antenna in watch 100 can be used to cover one or more communication frequency bands. Different antennas can also be multiplexed to improve antenna utilization. For example, antenna 1 can be multiplexed as a diversity antenna for a wireless local area network. In some other embodiments, the antennas can be used in conjunction with a tuning switch.

[0067] The mobile communication module 150 can provide wireless communication solutions, including 2G / 3G / 4G / 5G, for use on the watch 100. The mobile communication module 150 may include at least one filter, switch, power amplifier, low-noise amplifier (LNA), etc. The mobile communication module 150 can receive electromagnetic waves via the antenna 1, and perform filtering, amplification, and other processing on the received electromagnetic waves before transmitting them to the modem processor for demodulation. The mobile communication module 150 can also amplify the signal modulated by the modem processor and convert it into electromagnetic waves for radiation via the antenna 1. In some embodiments, at least some functional modules of the mobile communication module 150 may be housed in the processor 110. In some embodiments, at least some functional modules of the mobile communication module 150 and at least some modules of the processor 110 may be housed in the same device.

[0068] The wireless communication module 160 can provide solutions for wireless communication applications on the watch 100, including wireless local area networks (WLAN) (such as wireless fidelity (Wi-Fi) networks), Bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), and infrared (IR) technologies. The wireless communication module 160 can be one or more devices integrating at least one communication processing module. The wireless communication module 160 receives electromagnetic waves via antenna 2, performs frequency modulation and filtering of the electromagnetic wave signals, and sends the processed signal to processor 110. The wireless communication module 160 can also receive signals to be transmitted from processor 110, perform frequency modulation and amplification, and convert them into electromagnetic waves for radiation via antenna 2.

[0069] In some embodiments, antenna 1 of watch 100 is coupled to mobile communication module 150, and antenna 2 is coupled to wireless communication module 160, enabling watch 100 to communicate with networks and other devices via wireless communication technology. The wireless communication technology may include Global System for Mobile Communications (GSM), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Time-Division Code Division Multiple Access (TD-SCDMA), Long Term Evolution (LTE), BT, GNSS, WLAN, NFC, FM, and / or IR technologies, etc. The GNSS may include the Global Positioning System (GPS), the Global Navigation Satellite System (GLONASS), the BeiDou Navigation Satellite System (BDS), the Quasi-Zenith Satellite System (QZSS), and / or satellite-based augmentation systems (SBAS).

[0070] The watch 100 implements its display function through a GPU, a display screen 194, and an application processor. The GPU is a microprocessor for image processing, connected to the display screen 194 and the application processor. The GPU is used to perform mathematical and geometric calculations and for graphics rendering. The processor 110 may include one or more GPUs, which execute program instructions to generate or modify display information.

[0071] Display screen 194 is used to display images, videos, etc. Display screen 194 includes a display panel. The display panel may be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode (AMOLED), a flexible light-emitting diode (FLED), a miniature LED, a microLED, a quantum dot light-emitting diode (QLED), etc. In some embodiments, watch 100 may include one or N displays 194, where N is a positive integer greater than 1.

[0072] The watch 100 can achieve shooting functions through the ISP, camera 193, video codec, GPU, display 194 and application processor.

[0073] NPU stands for Neural Network (NN) Computing Processor. By borrowing the structure of biological neural networks, such as the transmission patterns between neurons in the human brain, it can quickly process input information and continuously learn on its own. Through the NPU, applications such as intelligent cognition in the Watch100 can be realized, including image recognition, facial recognition, speech recognition, and text understanding.

[0074] The internal memory 121 can be used to store computer executable program code, which includes instructions. The internal memory 121 may include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback, image playback, etc.), etc. The data storage area may store data created during the use of the watch 100 (such as audio data, phonebook, etc.). Furthermore, the internal memory 121 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, universal flash storage (UFS), etc. The processor 110 executes various functional applications and data processing of the watch 100 by running instructions stored in the internal memory 121 and / or instructions stored in memory located within the processor.

[0075] The watch 100 can achieve audio functions, such as music playback and recording, through an audio module 170, speaker 170A, receiver 170B, microphone 170C, headphone jack 170D, and application processor.

[0076] The pressure sensor 180A is used to sense pressure signals and can convert the pressure signals into electrical signals. In some embodiments, the pressure sensor 180A may be disposed on the display screen 194.

[0077] A distance sensor 180F is used to measure distance. The watch 100 can measure distance via infrared or laser. In some embodiments, during a shooting scenario, the watch 100 can utilize the distance sensor 180F to measure distance for rapid focusing.

[0078] The touch sensor 180K is also known as a "touch device". The touch sensor 180K can be set on the display screen 194. The touch sensor 180K and the display screen 194 together form a touch screen, also known as a "touch screen".

[0079] The ultrasonic sensor 180N, also known as the "ultrasonic module", is used to transmit ultrasonic signals to different directions in the surrounding environment of the watch 100, and to receive ultrasonic signals reflected back by objects in the surrounding environment. Then, it uses ultrasonic detection technology to determine multiple distance information between the watch 100 and the surrounding objects, which are used to characterize the distance between the watch 100 and the objects in the environment.

[0080] Buttons 190 include a power button, volume buttons, etc. Buttons 190 can be mechanical buttons or touch-sensitive buttons. The watch 100 can receive button input and generate key signal inputs related to user settings and function control of the watch 100.

[0081] Motor 191 can generate vibration alerts. Motor 191 can be used for incoming call vibration alerts or for touch vibration feedback. For example, different vibration feedback effects can correspond to touch operations performed on different applications (such as taking photos, playing audio, etc.). Motor 191 can also correspond to different vibration feedback effects for touch operations performed on different areas of the display screen 194. Different application scenarios (such as time reminders, receiving messages, alarm clocks, games, etc.) can also correspond to different vibration feedback effects. The touch vibration feedback effect can also be customized.

[0082] Indicator 192 can be an indicator light, used to indicate charging status, power changes, or to indicate messages, missed calls, notifications, etc.

[0083] In one implementation of this application, the watch 100 can provide the function of recording the correct posture for the user.

[0084] Please see Figure 3A The watch 100 displays the main screen, which includes information such as time and date. If the watch 100 detects a tap on the screen, it displays... Figure 3B The display interface shown includes information such as icons and names of applications in watch 100, such as the posture correction application G.

[0085] If watch 100 detects a user clicking the icon of the posture correction application G, watch 100 will display as follows: Figure 3C The display interface shown includes a "Correct Posture Input" control 300 and a "Posture Recognition and Correction" control 400.

[0086] If the watch 100 detects a user's click on the "Correct Posture Input" control 300, the watch 100 will display the following: Figure 3D The display interface shown is a second menu 310, which provides posture mode controls such as "computer viewing mode", "reading mode", "writing mode" and other modes (such as "tablet viewing mode") for the user to select.

[0087] If the watch 100 detects a user clicking the "Computer Viewing Mode" control, the watch 100 will display... Figure 3E The display interface shown above displays a reminder message 301: "Correct posture recording is in progress. Please assume the correct posture and remain still for at least 3 seconds." The user can then assume the posture described above. Figure 1A The correct posture is recorded as shown. Then, watch 100 acquires multiple distance information between watch 100 and surrounding objects corresponding to the user's current correct posture, forming an environmental distance matrix. This environmental distance matrix can be, for example, KM = [K1, K2, K3…Kn], where K1, K2, K3…Kn represent multiple distance information between watch 100 and surrounding objects. Then, watch 100 obtains and stores human posture feature information and human distance feature information based on the environmental distance matrix KM to complete the correct posture recording. The human posture feature information is the user's posture information, and the human distance feature information is the distance information between the user and watch 100.

[0088] After the correct posture is recorded on the watch 100, it will display as follows: Figure 3F The display interface shown includes a reminder message 302, "Please confirm whether to continue recording the correct posture," as a reminder to continue recording the correct posture. If the watch 100 detects the user's triggering of the "OK" control, the watch 100 continues to display... Figure 3EThe display interface shown is for the user to continue recording the correct posture. If the watch 100 detects the user clicking the "Cancel" control, the watch 100 completes the correct posture recording in the current mode and displays as shown. Figure 3D The display interface shown is shown below. At this time, the user can also select other posture modes to record the correct posture in other posture modes. The recording method is the same as or similar to the correct posture recording method corresponding to the "computer viewing mode" mentioned above, and will not be repeated here.

[0089] If watch 100 obtains an environment matrix based on a correct posture adopted by the user in a posture mode, and obtains a set of human posture feature information and distance feature information based on the environment matrix, then watch 100 stores the human posture feature information and human distance feature information as correct human posture feature information and correct human distance feature information. If watch 100 obtains multiple environment matrix information based on multiple correct postures adopted by the user in a posture mode, and obtains multiple sets of human posture feature information and human distance feature information, then watch 100 stores the multiple sets of human posture feature information and human distance feature information as correct human posture feature information and correct human distance feature information, respectively.

[0090] Watch 100 assigns a number to the correct posture input by the user and stores the corresponding correct posture features and correct distance features for each number. These features are used as comparison objects for posture anomaly identification when performing user posture recognition and correction.

[0091] In one implementation of this application, if the watch 100 detects a user's click on the "Correct Posture Input" control 300, the watch 100 can also display something like this. Figure 3G The display interface shown is a user menu 320 that provides user name controls such as "User A" and "User B" for the user to select. If the watch 100 detects a click operation by the user on a user name control (e.g., User A), the watch 100 displays the above-mentioned... Figure 3D The display interface shown allows the watch 100 to record the correct posture and distance information of different users, thus meeting the usage needs of different users.

[0092] In one implementation of this application, the watch 100 can be used for posture recognition and correction.

[0093] Watch 100 displays as follows Figure 4A As shown in the display interface, if the watch 100 detects a click operation by the user on the "Posture Recognition and Correction" control 400, the watch 100 will display as follows: Figure 4BThe display interface shown includes a posture mode menu 410 as the first menu, which provides posture mode option controls such as "Auto Mode", "Computer Viewing Mode", "Reading Mode", and "Writing Mode". The display interface may also include a "Settings" control 420.

[0094] If the watch 100 detects a user clicking the "Computer Viewing Mode" control, the watch 100 will display... Figure 4C The display interface shown is labeled 401, which displays the message "Posture recognition and correction enabled." The watch 100 can then recognize the user's posture to determine if the current posture is the correct posture corresponding to the current "Computer Viewing Mode."

[0095] If the watch 100 detects that the user's posture is incorrect, the watch 100 can alert the user to the incorrect posture through vibration, ringing, or displaying a reminder message. For example... Figure 4D As shown, watch 100 displays reminder message 402, "Your current posture is incorrect, please adjust it," as a correction reminder. Additionally, reminder message 402 may also include posture adjustment suggestions such as "Keep your back straight and keep a safe distance from the computer 200."

[0096] If the watch 100 detects that the user's posture is correct, the watch 100 may not need to issue any reminder message, or it may remind the user that the current posture is correct through voice or other means. For example, it may issue a voice message saying "Correct posture, please continue to maintain it".

[0097] In one implementation of this application, the watch 100 may also default to using the computer 200, with which it has established a communication connection, as a linked device. The watch 100 sends reminder information to the computer 200, which then performs a reminder operation based on the reminder information to alert the user to incorrect posture. For example, the computer 200 may issue the reminder through methods such as screen locking, screen flashing, or screen blackout.

[0098] In another implementation of this application, the watch 100 can also set up linked devices based on the user's operation.

[0099] If watch 100 detects that the user has... Figure 4B When the "Settings" control 420 is clicked, the watch 100 displays the following: Figure 4E The display interface shown includes a "Linked Devices" setting control 421.

[0100] If the watch 100 detects a user clicking the "Linked Device" setting control 421, the watch 100 displays... Figure 4FThe display interface shown is shown below. This display interface shows the "Established Communication Connection Device" control 4211 and the "Contacts in Address Book" control 4212.

[0101] If the watch 100 detects a user's click on the "Established Communication Connection Device" control 4211, the watch 100 will display as follows: Figure 4G The display interface shown includes the names of electronic devices that have established a communication connection with the watch 100, such as computer 200, mobile phone, tablet, speaker, etc. If the watch 100 detects a user's click operation on the name of an electronic device, the watch 100 will designate the electronic device corresponding to the name selected by the user as the linked device.

[0102] Additionally, if the watch 100 detects a user clicking on the name of an electronic device, it can display a "Reminder Method" menu 422. The "Reminder Method" menu 422 includes various reminder methods such as screen lock, vibration, ringtone, and screen blur. If the watch 100 detects a user clicking on the selection box following a reminder method, it determines that the reminder method corresponding to the currently selected electronic device is the user's chosen reminder method. For example, the reminder methods for the computer 200 include screen lock and ringtone. Users can set linked reminder methods for different electronic devices as needed.

[0103] If watch 100 detects that the user has... Figure 4F When the "Contacts" control 4212 on the displayed interface is clicked, the watch 100 displays the following: Figure 4H The display interface shown includes contacts in the address book. If the watch 100 detects the user's selection of a contact, the watch 100 will treat the corresponding electronic device as a linked device and can send reminders via SMS or other means by default.

[0104] In another implementation of this application, if the application on watch 100 is account-linked with the application on the electronic device corresponding to the contact, watch 100 can also send correction reminder information to the application on the electronic device corresponding to the contact through the application on watch 100. This application could be, for example, various chat applications, social applications, etc.

[0105] In another implementation of this application, the watch 100 uses a default linkage method, such as the watch 100 itself reminding users by vibration, the computer 200 and tablet computer reminding users by screen flashing or blacking out, and the electronic devices corresponding to contacts in the address book reminding users via SMS, etc.

[0106] In another implementation of this application, if the watch 100 determines the user's posture mode as "writing mode" based on the environmental distance matrix during posture recognition, which is inconsistent with the "computer viewing mode" manually selected by the user, the watch 100 can also display... Figure 4I The reminder message 403, "Please confirm whether to switch to 'Writing Mode'", serves as a mode switching reminder. If the watch 100 detects a user's click on the "Confirm" control, the watch 100 switches the posture mode to "Writing Mode" and determines whether the user's current posture is the correct posture corresponding to "Writing Mode". If the watch 100 detects a user's click on the "Cancel" control, the watch 100 determines whether the user's current posture is the correct posture corresponding to "Computer Viewing Mode".

[0107] In another implementation of this application, if the watch 100 detects that the user... Figure 4B The watch 100 can also display the "Auto Mode" control provided by the posture mode menu 410 shown in the figure. Figure 4C The displayed interface performs posture recognition and correction. Furthermore, the watch 100 determines the working mode to be automatic, and during the posture recognition process, the watch 100 automatically determines the user's current posture mode (e.g., "computer viewing mode" or "reading mode").

[0108] Additionally, if watch 100 detects that the user... Figure 4E Clicking the "Recognition and Correction Time" setting control on the interface shown allows the Watch 100 to also provide a time setting display interface for users to set the time for posture recognition and correction. Alternatively, the Watch 100 can default to setting the posture recognition and correction time to one hour after the user activates the posture recognition and correction function, which can be set as needed.

[0109] In one implementation of this application, if the watch 100 is as described above... Figure 3C As shown, upon detecting a user's click on the "Computer Viewing Mode" control, the watch 100 interprets this as a command from the user to correctly input the posture for "Computer Viewing Mode," and the watch 100 can then input the correct posture corresponding to "Computer Viewing Mode." The following is based on... Figure 5 This document explains and illustrates the process of recording the correct posture for the "computer viewing mode" of the watch 100.

[0110] Please see Figure 5 The process of recording the correct posture for "computer viewing mode" on Watch 100 includes the following steps.

[0111] S101, Watch 100 emits ultrasonic signals at different angles to its surroundings. The ultrasonic signals are reflected when they encounter objects in the surrounding environment.

[0112] S102, Watch 100 receives the reflected ultrasonic signal.

[0113] The distance between objects in the surrounding environment and watch 100 varies, and the time it takes for the ultrasonic signals to be reflected also varies. That is, when the ultrasonic signal is reflected by objects at different distances during transmission, the time it takes for watch 100 to receive the ultrasonic signals reflected back from different objects is different.

[0114] S103, based on the time of transmitting the ultrasonic signal, the time of receiving the ultrasonic signal, and the transmission rate of the ultrasonic wave, the watch 100 can calculate the distance information between the watch 100 and surrounding objects. Furthermore, the watch 100 can combine multiple distance information sets to obtain an environmental distance matrix KM = [K1, K2, K3...Kn].

[0115] Specifically, if the speed of ultrasonic wave transmission is V, and the time it takes for watch 100 to receive the corresponding reflected ultrasonic wave signal after emitting an ultrasonic signal is T, then the distance between the watch and surrounding objects is K = (V × T) / 2.

[0116] The watch 100 can obtain multiple distance information based on the ultrasonic signals reflected back from multiple different angles. Then, based on the distance information corresponding to multiple different angles, the environmental distance matrix KM = [K1, K2, K3...Kn] can be obtained.

[0117] S104, watch 100 determines whether there is a human body in the surrounding environment by using the environmental distance matrix. If a human body is present, watch 100 executes S105. If no human body is present, watch 100 executes S101.

[0118] Watch 100 determines the presence of a human body, i.e., the presence of a user, in its surrounding environment using an environmental distance matrix. This can be achieved by combining the environmental distance matrix with a human detection algorithm, which can be a deep neural network classifier. For example, Watch 100 inputs the environmental distance matrix KM into a deep neural network classifier composed of several convolutional layers, pooling layers, or fully connected layers. The classifier obtains the contour information of surrounding objects to determine their types or categories, and further determines whether these objects include human contours to ascertain the presence of a human body. The contour information of surrounding objects can be, for example, a point cloud model of object contours obtained based on distance information. The deep neural network classifier could be, for example, ShuffleNet, MobileNet, EfficientNet, or SqueezeNet.

[0119] Human detection algorithms can also be other machine learning algorithms. Watch 100 determines the presence of a human in its surroundings using an environmental distance matrix. Alternatively, Watch 100 can first extract points within a specific distance range from the environmental distance matrix KM using a machine learning algorithm, for example, points within the 0m to 1m distance range. Then, Watch 100 normalizes these extracted points using a machine learning algorithm and calculates the correlation parameters between these extracted points and a preset human distance template.

[0120] The human distance template can be obtained by collecting distance matrix data (e.g., 10,000 sets of distance matrix data) when a human body is in front of the watch 100, and then normalizing the distance matrix data to obtain the average distance matrix. The human distance template is this average distance matrix. The correlation parameter characterizes the similarity between the distance matrix formed by the extracted points and the human distance template, and can be represented by the average absolute error of all parameters between these two matrices. Specifically, if the distance matrix formed by the extracted points is completely different from the human distance template, the correlation parameter is 0; if it is completely identical, the correlation parameter is 1. Furthermore, if the correlation parameter is greater than a preset correlation parameter threshold, a human body is determined to be present; otherwise, it is determined not to be present. The preset correlation parameter threshold can range from 0.5 to 0.9, for example, 0.5, 0.65, 0.8, 0.9, etc. Of course, the correlation parameter threshold can also be other values.

[0121] S105, Watch 100 determines human posture feature information and human distance feature information through the environmental distance matrix.

[0122] The watch 100 determines the human body posture feature information and the human body distance feature information through the environmental distance matrix. It can be that the watch 100 extracts the human body posture feature information and the human body distance feature information from the environmental distance matrix by means of a deep neural network. For example, for the human body posture feature information, the watch 100 inputs the environmental distance matrix KM into an autoencoder network, and obtains the matrix k = [k1, k2, k3... km] (m < n) after dimensionality reduction as the human body posture feature information. The matrix k after dimensionality reduction can be used to describe the body contour of the user. For example, it can be a body contour point cloud model, so that it can be used to determine the posture of the user. For the human body distance feature information, the watch 100 can also input the environmental distance matrix KM into a deep neural network composed of several convolutional layers, pooling layers or fully connected layers to predict the distance value M between the user's head and the watch 100 (or it can also be called M

[0126] , ,

[0124] , ,

[0125] , ,

[0127] ,

[0123] , , 头 , ), as the human body distance feature information. M 头 can be the distance from a determined position on the user's head (such as the middle position between two eyes) to the watch 100, or it can be the average value of the distances from multiple positions on the user's head to the watch 100, etc.

[0123] Of course, the watch 100 can also extract the human body posture feature information and the human body distance feature information through other machine learning algorithms.

[0124] For example, for the human body posture feature information, the watch 100 can extract the points in a specific distance interval from the environmental distance matrix KM through a machine learning algorithm. For example, it extracts the points in the range of 0m to 1m, and performs normalization processing on them through a machine learning algorithm. Then, through a machine learning algorithm, it calculates the mean l1, variance l2, gradient l3, peak-to-peak value l4, peak factor l5 of these points, and calculates the correlation parameter l6 between these points and the human body typical correct posture distance template, etc., to obtain the human body posture feature matrix k = [k1, k2, k3... km] (m can be greater than n or less than n) as the human body posture feature information.

[0125] The human body typical correct posture distance template can be obtained by collecting a certain number of distance matrix data of human bodies with typical correct postures (such as 10,000 groups of distance matrix data), performing normalization processing, and then obtaining the average distance matrix. That is, the human body typical correct posture distance template is this average distance matrix. [[ID=​​​​For human body distance feature information, Watch 100 can also use machine learning algorithms to calculate the average value of distance points in the environmental distance matrix KM whose correlation parameter with the user's head distance template is greater than the correlation parameter threshold, based on the user's head distance template, and predict the distance value M from the user's head to Watch 100 as human body distance feature information.

[0128] The head distance template can be obtained by collecting a certain number of distance matrix data of human heads in front (e.g., 10,000 sets of distance matrix data), extracting the distance matrix information of the head position, and then normalizing it to obtain the average head distance matrix. That is, the head distance template is the average head distance matrix.

[0129] The specific process for obtaining human body distance feature information involves calculating the average absolute error of all parameters between the environmental distance matrix KM and the head distance template. If the environmental distance matrix KM and the head distance template are completely different, the correlation parameter is 0; if they are completely identical, the correlation parameter is 1. Furthermore, if the correlation parameter is greater than a preset correlation parameter threshold, the presence of a head is determined, and the average distance to the head is calculated as the human body distance feature information. The preset correlation parameter threshold can range from 0.5 to 0.9, for example, 0.5, 0.65, 0.8, 0.9, etc. Of course, the correlation parameter threshold can also be other values.

[0130] S106, watch 100 determines the posture number and stores the posture number, as well as the human posture feature information and human distance feature information obtained above, as the correct human posture feature information and correct human distance feature information.

[0131] The correspondence between the posture number, correct human posture feature information, and correct human distance feature information stored by Watch 100 can be, for example, as: posture mode number - posture number - correct human posture feature information - correct human distance feature information.

[0132] S107, watch 100 determines whether to continue recording the correct posture. If it continues, watch 100 executes S101; otherwise, watch 100 ends the recording of the correct posture.

[0133] The watch 100 determines whether correct posture recording has been performed by checking if the preset threshold for the number of correct posture recordings corresponding to the current "computer viewing mode" has been reached. If the threshold has been reached, the watch 100 ends the correct posture recording. If the threshold has not been reached, the watch 100 continues to execute S101 to perform correct posture recording.

[0134] Watch 100 determines whether correct posture recording has been performed, or it can determine this based on a received instruction from the user to continue recording the correct posture. For example, if watch 100 detects that the user has performed the above-mentioned posture recording... Figure 3F If the confirmation control shown is clicked, the watch 100 interprets this as a command from the user to continue recording the correct posture, and the watch 100 continues to execute S101 to record the correct posture. If the watch 100 detects that the user has clicked the confirmation control shown above... Figure 3F If the user clicks the cancel control shown, the watch 100 will determine that it has received a command from the user not to continue recording the correct posture, and the watch 100 will then end the correct posture recording.

[0135] In one implementation of this application, if the watch 100 receives a user's input of the correct posture corresponding to the "computer viewing mode", the watch 100 can obtain a set of posture number corresponding to posture mode 1 and the corresponding relationship between the correct human posture feature information and the correct human distance feature information, as shown in Table 1 below.

[0136]

[0137] Table 1

[0138] In another implementation of this application, if the watch 100 receives multiple inputs from the user of the correct posture corresponding to the "computer viewing mode", the watch 100 can obtain the correspondence between multiple sets of posture numbers corresponding to posture mode 2 and the correct human posture feature information and the correct human distance feature information, as shown in Table 2 below.

[0139]

[0140] Table 2

[0141] Furthermore, if the watch 100 receives a user's instruction to input the correct posture for other posture modes such as "reading mode" or "writing mode," the watch 100 will also perform operations similar to S101 to S107 described above to input the correct posture for the corresponding posture mode. The watch 100 can then obtain the corresponding relationship shown in Table 3 below.

[0142]

[0143] Table 3

[0144] In another implementation of this application, the watch 100 provides a correct posture recording function for multiple different users, for the purpose of recognizing and correcting the correct posture of multiple different users. After receiving the correct posture recording of different modes corresponding to different users, the watch 100 can obtain the correspondence shown in Table 4 below.

[0145]

[0146] Table 4

[0147] In another implementation of this application, the watch 100 includes S posture modes, and the watch 100 can also extract feature information P times for each posture mode by default. If the watch 100 completes S×P correct posture feature recordings, then the watch 100 determines that the correct posture recording is complete.

[0148] In this application, the watch 100 stores the obtained correct human posture feature information and correct human distance feature information by number, so that when performing posture recognition and correction, it can compare the obtained correct human posture feature information and correct human distance feature information corresponding to the current user's posture with the stored correct human posture feature information and correct human distance feature information to determine whether the user's posture is correct.

[0149] Furthermore, by utilizing the correct posture recording process, it is possible to learn the correct human posture characteristics and distance characteristics of people with different body types and habits in a personalized manner, demonstrating strong adaptability and robustness. Simultaneously, it can also record and store the correct posture under different posture modes (or eye use states), such as when looking at a computer, doing homework, or reading, improving the accuracy of subsequent posture recognition and correction stages.

[0150] In this implementation, machine learning algorithms are used to extract human posture features and human distance features, which can effectively improve the robustness and accuracy of recognition.

[0151] This application also provides a method for human posture recognition and correction using a watch 100. For example, if the watch 100 detects that the user is sitting... Figure 4B Clicking the "Computer Viewing Mode" control as shown will cause the watch 100 to recognize and correct human posture. The following is based on... Figure 6 The process of human posture recognition and correction for watch 100 is explained and illustrated.

[0152] Please see Figure 6 In one implementation of this application, the process of human posture recognition and correction by the watch 100 includes the following steps.

[0153] S201, Watch 100 determines the current working mode, which includes automatic mode and non-automatic mode.

[0154] If the watch detects the above... Figure 4B The user's click on the "Auto Mode" control indicates that the current operating mode is Auto Mode. If the watch 100 detects the user's click on the "Auto Mode" control, the watch will confirm that the current operating mode is Auto Mode. Figure 4B For the click operation of the "View Computer Mode" control shown, it is determined that the current working mode is the non-automatic mode.

[0155] In the automatic mode, the watch 100 directly and automatically determines the current posture mode according to the user's posture. In the non-automatic mode, the watch 100 determines the posture mode according to the user's selection operation of the posture mode.

[0156] S202, the watch 100 emits an ultrasonic signal.

[0157] S203, the watch 100 receives the reflected ultrasonic signal.

[0158] S204, the watch 100 calculates the environmental distance matrix using the time information of the obtained reflected ultrasonic signal.

[0159] S205, based on the machine learning algorithm, the watch 100 determines whether there is a human body in the surrounding environment of the watch 100 through the environmental distance matrix. If there is a human body, the watch 100 executes S206; if there is no human body, the watch 100 executes S202.

[0160] S206, the watch 100 calculates the human body posture feature information of the user and the human body distance feature information between the watch 100 and the user through the environmental distance matrix. Among them, the human body posture feature information of the user can be the matrix k = [k1, k2, k3... km] (m < n) after dimensionality reduction of the environmental distance matrix KM as described above. The dimensionality-reduced matrix k can be used to describe the body contour of the user. The human body distance feature information between the watch 100 and the user is the distance value M between the user's head and the watch 100.

[0161] It should be noted that the above S202 - S206 correspond to the aforementioned 101 - 105 respectively and are the same in sequence, so they will not be elaborated here.

[0162] S207, the watch 100 performs posture anomaly detection.

[0163] For S207, the watch 100 performs posture anomaly detection, which can be as Figure 7 shown, including the following steps:

[0164] S2071, the watch 100 determines the working mode and determines the current posture mode corresponding to the human body posture feature information according to the human body posture feature information.

[0165] The watch 100 determines the current posture mode corresponding to the human posture feature information based on the human posture feature information. This can be achieved by the watch 100 calculating the average Euclidean distance between the human posture feature information and the correct human posture feature information stored in various posture modes, and taking the posture mode represented by the human posture feature information with the largest average Euclidean distance as the current posture mode.

[0166] Furthermore, if watch 100 determines that the operating mode is automatic, then watch 100 executes S2072. If watch 100 determines that the operating mode is non-automatic, then watch 100 executes S2073.

[0167] S2072, the watch 100 compares the human posture feature information and human distance feature information with the correct human posture feature information and correct human distance feature information corresponding to the current posture mode to determine whether the current posture is correct, i.e., whether it is abnormal.

[0168] Specifically, the Watch 100 is based on an anomaly detection algorithm, which can be a deep neural network. For example, the Watch 100 inputs the current human posture feature information, human distance feature information, and the correct human posture information and correct human distance information stored in the Watch 100 into a deep neural network discriminator composed of several convolutional layers, pooling layers, or fully connected layers to determine whether the posture is abnormal, that is, whether the posture is correct.

[0169] The anomaly detection algorithm can also be other machine learning algorithms. For example, Watch100 uses a machine learning algorithm to calculate the Euclidean distance between the current human posture feature information and the stored correct human posture feature information, as well as the Euclidean distance between the current human distance feature information and the stored correct human distance feature information. Then, the average of the two Euclidean distances is used as the final anomaly Euclidean distance result, and it is determined whether the final anomaly Euclidean distance result is greater than a preset final anomaly Euclidean distance result threshold. If the final anomaly Euclidean distance result is greater than the final anomaly Euclidean distance result threshold, then the posture is determined to be abnormal.

[0170] The final abnormal Euclidean distance threshold is a statistical value. It can be calculated by statistically analyzing a certain number of users (e.g., 100 people), calculating the abnormal Euclidean distance results for each user under several groups (e.g., 50 groups) of abnormal postures, and taking the minimum value of the abnormal Euclidean distance results as the abnormal Euclidean distance threshold for that user. The average value of the abnormal Euclidean distance thresholds of this certain number of users (e.g., 100 people) is then used as the final abnormal Euclidean distance threshold.

[0171] S2073, Watch 100 determines whether the current posture mode is consistent with the posture mode manually set by the user. If they are inconsistent, Watch 100 executes S2074. If they are consistent, Watch 100 executes S2077.

[0172] S2074, Watch 100 displays a posture mode switching reminder message. Watch 100 then executes S2075.

[0173] S2075, watch 100 determines whether to switch posture modes. If it determines to switch posture modes, it switches to the current posture mode determined above based on human posture feature information and human distance feature information, and watch 100 executes S2076. If it determines not to switch posture modes, watch 100 executes S2077.

[0174] The posture mode switching reminder displayed on Watch 100 could be, for example, as follows: Figure 4I The reminder message 403 is shown. If the watch 100 detects a user clicking the "Confirm" control, the watch 100 determines to switch the posture mode. If the watch 100 detects a user clicking the "Cancel" control, the watch 100 determines not to switch the posture mode.

[0175] In addition, posture mode switching reminders can be one or more of the following: vibration, sound, text, and images. Alternatively, they can be generated on other devices connected to the same Bluetooth or WiFi network as the wristband, through device linkage. This includes generating one or more of the following: vibration, sound, text, and images. It also includes changing the display parameters, display characteristics, touch characteristics, and control characteristics of other devices, such as turning off the screen, blurring the image, forcibly exiting, or forcibly locking the screen.

[0176] S2076, the watch 100 compares the human posture feature information and human distance feature information with the correct human posture feature information and correct human distance feature information corresponding to the switched posture mode, for example, by calculating the aforementioned Euclidean distance and the final abnormal Euclidean distance result, to determine whether the current posture is correct.

[0177] S2077, the watch 100 compares the human posture feature information and human distance feature information with the correct human posture feature information and correct human distance feature information corresponding to the posture mode manually selected by the user, such as calculating the aforementioned Euclidean distance and the final abnormal Euclidean distance result, to determine whether the current posture is correct.

[0178] Watch 100 executes S2076 or S2077 to determine whether the current posture is accurate, similar to the aforementioned S2072, which will not be repeated here.

[0179] S208, the watch 100 determines whether to issue a correction reminder based on the judgment result of step S207. Specifically, if the current posture is normal, the watch 100 executes S210. If the current posture is abnormal, the watch 100 executes S209.

[0180] S209, Watch 100 sends a correction signal to remind the user to make corrections.

[0181] When Watch 100 provides posture correction reminders to the user, the reminders specifically include: Watch 100 generating its own correction signals (or reminder information) such as vibration, sound, text, and images to remind the user. Alternatively, Watch 100 can utilize device linkage to generate reminders on other linked devices within the same Bluetooth or Wi-Fi network. This includes the linked devices generating correction signals (or reminder information) such as vibration, sound, text, and images to remind the user, or the devices changing display parameters (e.g., dimming the screen), display characteristics (e.g., enlarging text), touch characteristics (e.g., freezing the screen and not responding to user actions), and control characteristics (e.g., powering off) to remind the user. Using device linkage for correction reminders can increase the "penalty" of posture correction, forcing the user to better correct their posture to some extent.

[0182] S210, watch 100 does not emit a correction signal and does not remind the user to correct the error.

[0183] S211, watch 100 determines whether the posture recognition and correction process is complete. If complete, watch 100 exits the posture recognition and correction process. If not complete, watch 100 continues to execute S202.

[0184] The watch 100 determines whether the posture recognition and correction work is finished. This can be done by the watch 100 determining whether the posture recognition and correction time set by the user has been reached, or by detecting whether the user has manually exited the posture recognition and correction task.

[0185] Watch 100 continues to execute S202, which means that watch 100 periodically emits ultrasonic signals according to the user's preset posture recognition and correction cycle in order to recognize and correct the user's posture.

[0186] The posture recognition and correction method provided in this implementation allows the watch 100 to identify posture anomalies by comparing the real-time posture features and distance features of the user with stored correct posture features and distance features, thereby determining whether the user's posture is correct. This effectively improves the dynamic range and adaptability of electronic devices, enhancing the robustness and accuracy of posture recognition. Furthermore, the watch 100 is easy for users to carry and use, and is aesthetically pleasing and comfortable to wear. Additionally, the watch 100 boasts a wide dynamic range and good adaptability.

[0187] Please see Figure 8 , Figure 8 The diagram shown is a schematic diagram of the module structure of the posture recognition and correction device provided in this application.

[0188] The posture recognition and correction device includes an ultrasonic module 101, an information processing module 102, and an information storage module 103. The ultrasonic module 101 is used to obtain environmental matrix information; the information processing module 102 is used to handle various computational tasks during device operation, such as determining posture anomalies; and the information storage module 103 is used to store various types of data that need to be saved during device operation, such as storing correct human posture feature information and correct human distance feature information corresponding to multiple different posture modes, as well as other data such as program data, calculation process data, and input information data.

[0189] In one implementation of this application, the watch 100 described above includes the posture recognition and correction device.

[0190] The ultrasonic module 101 includes an ultrasonic transmitting module 1011 and an ultrasonic receiving module 1012.

[0191] The ultrasonic transmitting module 1011 is used to perform the aforementioned S101, or to perform the aforementioned S202, that is, the ultrasonic transmitting module 1011 is used to transmit ultrasonic signals and record the ultrasonic transmission time.

[0192] The ultrasonic receiving module 1012 is used to execute S102 and S103 as described above, or S203 and S204 as described above. That is, the ultrasonic receiving module 1012 is used to receive the ultrasonic signals returned by surrounding objects after the ultrasonic waves emitted by the ultrasonic transmitting module 1011 pass through the ultrasonic objects, and records the time of receiving the returned ultrasonic signals. Furthermore, the ultrasonic transmitting module 1011 sends the ultrasonic emission time to the ultrasonic receiving module 1012. The ultrasonic receiving module 1012 calculates the distance information between the attitude recognition and correction device and various objects in the surrounding environment based on the ultrasonic signal emission time, the time of receiving the reflected ultrasonic signals, and the ultrasonic transmission speed, and further calculates the environmental distance matrix KM.

[0193] The information processing module 102 includes a human body detection module 1021, a working mode detection module 1022, a human body distance feature extraction module 1023, a human body posture feature extraction module 1024, and an anomaly detection module 1025.

[0194] The human body detection module 1021 is used to execute S104 or S205 as described above. That is, the human body detection module 1021 receives the environmental distance matrix KM information sent by the ultrasonic module 101, and determines whether there is a human body in the surrounding environment of the watch 100 through the environmental distance matrix KM. If there is a human body, the human body detection module 1021 sends (informs) the information that there is a human body in the surrounding environment of the watch 100 to the working mode detection module 1022.

[0195] The working mode detection module 1022 is used to execute S2071 in S207 above, that is, the working mode detection module 1022 is used to determine whether the current working mode of the watch 100 for posture recognition and correction is automatic mode or non-automatic mode. Then, the working mode detection module 1022 sends the determined working mode type information to the human body distance feature extraction module 1023.

[0196] The human body distance feature extraction module 1023 is used to perform S105 as described above, that is, the human body distance feature extraction module 1023 is used to extract the distance feature information between the watch 100 and the user through the environmental distance matrix KM. The human body distance feature extraction module 1023 is also used to send the distance feature information to the information storage module 103 for storage as correct distance feature information.

[0197] The human posture feature extraction module 1024 is used to perform S105 as described above, that is, the human posture feature extraction module 1024 is used to extract the human posture feature information of the user of the watch 100 through the environmental distance matrix. The human posture feature extraction module 1024 is also used to send the posture feature information to the information storage module 103 for storage as correct posture feature information.

[0198] The information storage module 103 is used to execute the aforementioned S106, that is, the information storage module 103 is used to receive distance feature information sent by the human distance feature extraction module 1023 and store it as correct distance feature information, and to receive human posture feature information sent by the human posture feature extraction module 1024 and store it as correct human posture feature information.

[0199] The human body distance feature extraction module 1023 is also used to perform S206 as described above, that is, the human body distance feature extraction module 1023 is used to extract human body distance feature information between the watch 100 and the user through the environmental distance matrix KM. The human body distance feature extraction module 1023 is also used to send the distance feature information to the anomaly detection module 1025.

[0200] The human posture feature extraction module 1024 is used to perform S206 as described above, that is, the human posture feature extraction module 1024 is used to extract the human posture feature information of the user of the watch 100 through the environmental distance matrix. The human posture feature extraction module 1024 is also used to send the human posture feature information to the anomaly detection module 1025.

[0201] The anomaly detection module 1025 is used to execute S2072 to S2075 in the above-mentioned S207, that is, to perform posture anomaly detection based on the human distance feature information sent by the human distance feature extraction module 1023, the human posture feature information sent by the human posture feature extraction module 1024, and the correct human posture feature information and correct human distance feature information stored in the information storage module 103.

[0202] In one implementation of this application, the attitude recognition and correction device may consist of only one ultrasonic module 101. The ultrasonic module 101 is used to send and receive ultrasonic signals in different directions according to a preset period to obtain an environmental distance matrix. The preset period can be, for example, 0.01s. Of course, it can also be other values.

[0203] In another implementation of this application, the attitude recognition and correction device may also include an array of multiple ultrasonic modules 101, which are used to simultaneously send and receive ultrasonic signals in different directions to obtain an environmental distance matrix.

[0204] In another implementation of this application, the ultrasonic module 101 may further include a calculation module. The ultrasonic transmitting module 1011 is used to transmit ultrasonic waves and record the ultrasonic wave transmission time. The ultrasonic receiving module 1012 is used to receive the ultrasonic wave signals returned by objects in the surrounding environment after the ultrasonic waves transmitted by the ultrasonic transmitting module 1011 have passed through it, and records the time of receiving the returned ultrasonic wave signals. The ultrasonic transmitting module 1011 sends the ultrasonic wave transmission time to the calculation module. The calculation module calculates the distance information between the attitude recognition and correction device and various objects in the surrounding environment based on the ultrasonic wave signal transmission time, the time of receiving the reflected ultrasonic wave signals, and the ultrasonic wave transmission speed, and further calculates the environmental distance matrix.

[0205] In some other implementations of this application, the watch 100 can also select a posture mode to record the correct posture after receiving a user's voice command, as well as perform posture recognition and correction.

[0206] In other implementations of this application, the watch 100 can also be other wearable hand devices such as bracelets, bangles, necklaces, rings, gloves, etc., which are convenient for users to carry and use, and are aesthetically pleasing and comfortable to wear. Furthermore, it has the advantages of a wide dynamic range and good adaptability.

[0207] In other implementations of this application, the watch 100 may also be a head-worn wearable device such as glasses or earrings, or other electronic devices such as mobile phones, tablets, speakers, electronic pens, televisions, laptops, ultra-mobile personal computers (UMPCs), personal digital assistants (PDAs), and virtual reality devices.

[0208] In other implementations of this application, the watch 100 can also obtain standard human posture feature information and standard human distance feature information from the application server corresponding to the posture recognition and correction application as correct human posture feature information and correct human distance feature information. The standard human posture feature information and standard human distance feature information are obtained by analyzing big data of the correct postures corresponding to different posture modes.

[0209] Please see Figure 9 , Figure 9 The diagram shown is a schematic of the posture recognition and correction system provided in this application.

[0210] The posture recognition and correction system provided in this application includes a wristband 500 and a mobile phone 600, wherein the user wears the wristband 500. The wristband 500 is used to transmit and receive ultrasonic signals to obtain first environmental distance information and to send the first environmental distance information to the mobile phone 600. The mobile phone 600 is used to store preset correct human posture feature information and correct human distance feature information, and to determine the first human posture feature information and the first human distance feature information based on the first environmental distance information. If it is determined that the user's posture is incorrect based on the first human posture feature information, the first human distance feature information, the correct human posture feature information, and the correct human distance feature information, it generates a correction reminder message to remind the user to adjust their sitting posture.

[0211] The wristband 500 includes the aforementioned ultrasonic module 101, which includes an ultrasonic transmitting module 1011 and an ultrasonic receiving module 1012. The ultrasonic module 101 is used to obtain environmental matrix information.

[0212] The mobile phone 600 includes the aforementioned information processing module 102 and information storage module 103. The information processing module 102 includes a human detection module 1021, a working mode detection module 1022, a human distance feature extraction module 1023, a human posture feature extraction module 1024, and an anomaly detection module 1025, used for various computational tasks such as determining posture anomalies during device operation. The information storage module 103 is used to store various types of data that need to be saved during device operation, such as storing correct human posture feature information and correct human distance feature information, such as storing correct human posture feature information and correct human distance feature information corresponding to multiple different posture modes, as well as other data such as program data, calculation process data, and input information data.

[0213] The wristband 500 acquires environmental matrix information via the ultrasonic module 101 and sends this information to the mobile phone 600. Upon receiving the environmental distance matrix information from the wristband 500, the mobile phone 600 performs operations similar to those described in S205-S211 to recognize and correct posture. Furthermore, if the mobile phone 600 detects an incorrect user posture, it can send a correction reminder to the wristband 500, causing the wristband 500 to vibrate or otherwise indicate that the user's posture is correct. Additionally, the mobile phone 600 can also perform operations similar to those described in S104-S107 to record the correct posture.

[0214] In other implementations of this application, the posture recognition and correction method provided can also be used in other posture recognition and correction scenarios. For example, it can be used to recognize sleeping postures and generate reminder information when a user's sleeping posture is incorrect. Additionally, it can be used to recognize and correct movement postures, enabling users to achieve better exercise results and reducing injuries caused by incorrect movement postures. Of course, it can also be used for other purposes such as recognizing standing postures and walking postures, and providing reminders for incorrect postures.

[0215] Please see Figure 10 , Figure 10The diagram illustrates the structure of an electronic device 900 according to one embodiment of this application. The electronic device 900 may include one or more processors 901 coupled to a controller hub 904. In at least one embodiment, the controller hub 904 communicates with the processor 901 via a multi-branch bus such as a Front Side Bus (FSB), a point-to-point interface such as a QuickPath Interconnect (QPI), or a similar connection. The processor 901 executes instructions controlling general types of data processing operations. In one embodiment, the controller hub 904 includes, but is not limited to, a Graphics Memory controller hub (GMCH) (not shown) and an Input / Output Hub (IOH) (which may be on a separate chip) (not shown), wherein the GMCH includes memory and a graphics controller and is coupled to the IOH.

[0216] Electronic device 900 may also include a coprocessor 906 and a memory 902 coupled to a controller hub 904. Alternatively, one or both of the memory 902 and the GMCH may be integrated within the processor 901 (as described in this application), with the memory 902 and the coprocessor 906 directly coupled to the processor 901 and the controller hub 904, which is located on a single chip with the IOH.

[0217] The memory 902 may be, for example, dynamic random access memory (DRAM), phase change memory (PCM), or a combination of the two.

[0218] In one embodiment, the coprocessor 906 is a dedicated processor, such as, for example, a high-throughput many-integrated-core (MIC) processor, a network or communication processor, a compression engine, a graphics processor, a general-purpose graphics processing unit (GPGPU), or an embedded processor, etc. Optional properties of the coprocessor 906 are indicated by dashed lines. Figure 9 middle.

[0219] In one embodiment, the electronic device 900 may further include a Network Interface Card (NIC) 903. The network interface 903 may include a transceiver for providing a radio interface for the electronic device 900 to communicate with any other suitable device (such as a front-end module, antenna, etc.). In various embodiments, the network interface 903 may be integrated with other components of the electronic device 900. The network interface 903 can implement the functions of the communication unit in the above embodiments.

[0220] Electronic device 900 may further include input / output (I / O) device 905. Input / output (I / O) device 905 may include: a user interface designed to enable a user to interact with electronic device 900; a peripheral component interface designed to enable peripheral components to also interact with electronic device 900; and / or sensors designed to determine environmental conditions and / or location information related to electronic device 900.

[0221] It is worth noting that, Figure 10 This is merely an example. That is, although... Figure 10 The electronic device 900 shown includes multiple devices such as a processor 901, a controller hub 904, and a memory 902. However, in practical applications, devices using the methods of this application may include only a portion of the devices in the electronic device 900. For example, it may include only the processor 901 and the NIC 903. Figure 10 The properties of the optional devices are shown by dashed lines.

[0222] The memory of the electronic device 900 may include one or more tangible, non-transitory computer-readable media for storing data and / or instructions. The computer-readable storage medium stores instructions, specifically, temporary and permanent copies of those instructions.

[0223] In this application, the electronic device 900 may specifically be a terminal device such as a mobile phone, tablet computer, personal digital assistant (PDA), or desktop computer. The instructions stored in the memory of the electronic device may include instructions that, when executed by at least one unit in the processor, cause the electronic device to perform the posture recognition and correction method as mentioned above.

[0224] Please see Figure 11 , Figure 11 The diagram shown is a schematic representation of the structure of a SoC (System on Chip) 1000 according to an embodiment of this application. Figure 11In the figures, similar components have the same reference numerals. Additionally, the dashed box is an optional feature of the more advanced SoC 1000. This SoC 1000 can be used in any electronic device according to this application, and can implement corresponding functions depending on the device it is in and the instructions stored within it.

[0225] exist Figure 11 In this SoC 1000, the following are included: an interconnect unit 1002 coupled to a processor 1001; a system proxy unit 1006; a bus controller unit 1005; an integrated memory controller unit 1003; a group or one or more coprocessors 1007, which may include integrated graphics logic, an image processor, an audio processor, and a video processor; a static random-access memory (SRAM) unit 1008; and a direct memory access (DMA) unit 1004. In one embodiment, the coprocessor 1007 includes a dedicated processor, such as, for example, a network or communication processor, a compression engine, a GPGPU, a high-throughput MIC processor, or an embedded processor, etc.

[0226] SRAM cell 1008 may include one or more computer-readable media for storing data and / or instructions. The computer-readable storage medium may store instructions, specifically, temporary and permanent copies of those instructions. These instructions may include, when executed by at least one unit of processor 1001, instructions that cause the electronic device to perform the attitude recognition and correction methods as described above.

[0227] It should be noted that the terms "first" and "second" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0228] It should be noted that some structural or methodological features may be shown in the accompanying drawings in a specific arrangement and / or order. However, it should be understood that such a specific arrangement and / or order may not be necessary. Rather, in some embodiments, these features may be arranged in a manner and / or order different from that shown in the illustrative drawings. Furthermore, including structural or methodological features in a particular figure does not imply that such features are required in all embodiments, and in some embodiments, these features may be omitted or may be combined with other features.

[0229] Although this application has been illustrated and described with reference to certain preferred embodiments, those skilled in the art should understand that the above description is a further detailed explanation of the application in conjunction with specific embodiments, and should not be construed as limiting the specific implementation of the application to these descriptions. Those skilled in the art can make various changes in form and detail, including some simple deductions or substitutions, without departing from the spirit and scope of this application.

Claims

1. A posture recognition and correction method, applied to electronic devices, characterized in that, The method comprises: obtaining first environment distance information, the first environment distance information being distance information between the electronic device and surrounding environment objects; determining first human body posture feature information and first human body distance feature information of a user of the electronic device according to the first environment distance information, the first human body posture feature information representing a current posture of the user, and the first human body distance feature information representing a current distance between the user and the electronic device; if it is determined that the posture of the user is incorrect according to the first human body posture feature information and the first human body distance feature information, and preset correct human body posture feature information and correct human body distance feature information, generating a correction reminder information, wherein the correct human body posture feature information and the correct human body distance feature information are input and stored according to a correct posture of the user, the correct human body posture feature information representing the correct posture of the user, and the correct human body distance feature information representing a correct distance between the user and the electronic device; wherein the first environment distance information is an environment distance matrix, the environment distance matrix comprising a plurality of distance information between the electronic device and surrounding environment objects; the first human body posture feature information and the first human body distance feature information of the user of the electronic device are determined according to the first environment distance information, comprising: using the first environment distance information after dimension reduction as the first human body posture feature information, the first human body posture feature information being used to indicate a body contour of the user, or extracting points in a preset distance interval in the first environment distance information, performing normalization processing on the points in the preset distance interval, and determining a mean value of the normalized points in the preset distance interval, a variance of the normalized points in the preset distance interval, a gradient of the normalized points in the preset distance interval, a peak-to-peak value of the normalized points in the preset distance interval, a peak value factor of the normalized points in the preset distance interval, and a correlation parameter of the normalized points in the preset distance interval and a typical correct posture distance template of a human body, to obtain the first human body posture feature information; inputting the first environment distance information into a deep neural network to obtain a distance between the head of the user and the electronic device as the first human body distance feature information, or obtaining a head distance template, determining an average value of distance points in the first environment distance information having a correlation parameter greater than a correlation parameter threshold with the head distance template based on the head distance template, to obtain a distance value from the head of the user to the electronic device as the first human body distance feature information.

2. The method of claim 1, wherein, The method further comprises: if it is determined that the surrounding environment objects include the user according to the first environment distance information, determining the first human body posture feature information and the first human body distance feature information of the user according to the first environment distance information.

3. The method of claim 1, wherein, if the working mode is an automatic mode, the method further comprises determining whether the posture of the user is correct by: determine a first posture mode according to the first human body posture feature information and the first human body distance feature information; determine whether the posture of the user is correct according to the first human body posture feature information and the first human body distance feature information, and correct human body posture feature information and correct human body distance feature information corresponding to the first posture mode pre-stored.

4. The method of claim 3, wherein, The method further comprises determining the working mode by: displaying a first menu, the first menu comprising an automatic mode control; if detecting a selection operation of the automatic mode control by the user, determining the working mode as an automatic mode.

5. The method of claim 1, wherein, If the working mode is a non-automatic mode, the method further comprises determining whether the posture of the user is correct by: determining a first posture mode according to the first human body posture feature information and the first human body distance feature information; if the first posture mode is inconsistent with a pre-set second posture mode, generating mode switching reminding information for the user to perform a posture mode selection operation; if receiving a selection operation of the first posture mode by the user, determining whether the posture of the user is correct according to the first human body posture feature information and the first human body distance feature information, and correct human body posture feature information and correct human body distance feature information corresponding to the first posture mode pre-stored; if receiving a selection operation of the second posture mode by the user, determining whether the posture of the user is correct according to the first human body posture feature information and the first human body distance feature information, and correct human body posture feature information and correct human body distance feature information corresponding to the second posture mode pre-stored.

6. The method of claim 5, wherein, The method further comprises determining the working mode by: displaying a first menu, the first menu comprising at least one posture mode control; if detecting a selection operation of the posture mode control by the user, determining the working mode as a non-automatic mode, and determining the posture mode corresponding to the posture mode control as the second posture mode.

7. The method of claim 6, wherein, The posture mode control is any one of a reading mode control, a writing mode control and a computer viewing mode control.

8. The method of claim 1, wherein, The method further comprises performing correct posture input to obtain the correct human body posture feature information and the correct human body distance feature information by: displaying a second menu, the second menu comprising at least one posture mode control; if receiving a selection operation of the posture mode control by the user, obtaining second environment distance information, the second environment distance information being distance information between the electronic device and surrounding environment objects; determining second human body posture feature information and second human body distance feature information according to the second environment distance information, the second human body posture feature information representing a current posture of the user, and the second human body distance feature information representing a current distance between the user and the electronic device; taking the second human body posture feature information as the correct human body posture feature information, and taking the second human body distance feature information as the correct human body distance feature information.

9. The method of claim 8, wherein, The method further comprises: displaying a correct posture continuing entry reminding information for the user to select an operation of correct posture entry; if the user continues to select the operation of correct posture entry, continuing correct posture information entry; if the user cancels the operation of correct posture entry, ending correct posture information entry.

10. The method of claim 1-9, wherein, The method further comprises: performing a corresponding reminding operation according to the correction reminding information; and / or sending the correction reminding information to other electronic devices in communication connection with the electronic device, so that the other electronic devices perform a corresponding reminding operation according to the correction reminding information.

11. The method of claim 1-9, wherein, The method further comprises: emitting an ultrasonic signal to the surrounding environment; receiving the ultrasonic signal reflected by the object in the surrounding environment; obtaining distance information between the electronic device and the object in the surrounding environment according to the time of emitting the ultrasonic signal, the time of receiving the reflected ultrasonic signal, and the transmission rate of the ultrasonic signal.

12. A posture recognition and correction device, characterized by, comprises: an ultrasonic module for emitting and receiving an ultrasonic signal to obtain first environment distance information, wherein the first environment distance information is distance information between the posture recognition and correction device and an object in the surrounding environment, and the first environment distance information is an environment distance matrix comprising a plurality of distance information between the posture recognition and correction device and the object in the surrounding environment; an information storage module for storing preset correct human posture feature information and correct human distance feature information, wherein the correct human posture feature information represents a correct posture of a user, and the correct human distance feature information represents a correct distance between the user and an electronic device; an information processing module for determining first human posture feature information and first human distance feature information according to the first environment distance information, and generating correction reminding information if it is determined that the posture of the user is incorrect according to the first human posture feature information and the first human distance feature information, and the correct human posture feature information and the correct human distance feature information stored by the information storage module according to the correct posture entry of the user, wherein the first human posture feature information represents the posture of the user, and the first human distance feature information represents the distance between the user and the electronic device. The information processing module is configured to use the first environment distance information after dimension reduction as the first human posture feature information, and the first human posture feature information is used to indicate the body contour of the user, or extracting points in a preset distance interval from the first environment distance information, performing normalization processing on the points in the preset distance interval, and determining the mean value of the normalized points in the preset distance interval, the variance of the normalized points in the preset distance interval, the gradient of the normalized points in the preset distance interval, the peak-to-peak value of the normalized points in the preset distance interval, the peak value factor of the normalized points in the preset distance interval, and the correlation parameter of the normalized points in the preset distance interval and a typical correct human posture distance template, to obtain the first human posture feature information. The information processing module is configured to input the first environment distance information into a deep neural network to obtain a distance between the user's head and the electronic device as the first human distance feature information, or The information processing module is configured to obtain a head distance template, determine an average value of distance points in the first environment distance information that have a correlation parameter greater than a correlation parameter threshold in the head distance template, and obtain a distance value from the user's head to the electronic device as the first human distance feature information.

13. The posture recognition and correction device according to claim 12, characterized in that, The ultrasonic module is provided with one, and the ultrasonic module is used for emitting and receiving ultrasonic signals in different directions according to a preset period; or The ultrasonic module is provided with multiple, and multiple ultrasonic modules are used for simultaneously emitting and receiving ultrasonic signals in different directions.

14. A posture recognition and correction system, characterized by, The first electronic device and the second electronic device are connected through a communication connection, The first electronic device is configured to emit and receive ultrasonic signals to obtain first environment distance information, and send the first environment distance information to the second electronic device, the first environment distance information being distance information between the first electronic device and surrounding environment objects, the first environment distance information being an environment distance matrix, and the environment distance matrix including multiple distance information between the electronic device and surrounding environment objects; The second electronic device is configured to store preset correct human posture feature information and correct human distance feature information, determine first human posture feature information and first human distance feature information according to the first environment distance information, and generate a correction reminder information if it is determined that the user's posture is incorrect according to the received first human posture feature information and first human distance feature information sent by the first electronic device, and the correct human posture feature information and the correct human distance feature information, the first human posture feature information representing the user's posture, the first human distance feature information representing the distance between the user and the first electronic device, the correct human posture feature information and the correct human distance feature information being recorded and stored according to the correct posture of the user, the correct human posture feature information representing the correct posture of the user, and the correct human distance feature information representing the correct distance between the user and the first electronic device; wherein The second electronic device is configured to use the first environment distance information after dimension reduction as the first human posture feature information, and the first human posture feature information is used to indicate the user's body contour, or extracting a point in a preset distance interval in the first environment distance information, performing normalization processing on the point in the preset distance interval, and determining a mean value of the normalized point in the preset distance interval, a variance of the normalized point in the preset distance interval, a gradient of the normalized point in the preset distance interval, a peak-to-peak value of the normalized point in the preset distance interval, a peak factor of the normalized point in the preset distance interval, and a correlation parameter of the normalized point in the preset distance interval and a typical correct posture distance template of a human body, to obtain the first human body posture feature information; the second electronic device is configured to input the first environment distance information into a deep neural network to obtain a distance between a head of a user and the electronic device as the first human body distance feature information, or obtaining a head distance template, determining an average value of distance points in the first environment distance information that have a correlation parameter greater than a correlation parameter threshold with respect to the head distance template based on the head distance template, and obtaining a distance value of the head of the user to the electronic device as the first human body distance feature information.

15. An electronic device, comprising: comprise: a memory configured to store a computer program, the computer program comprising program instructions; a control component configured to execute the program instructions to cause the electronic device to perform the posture recognition and correction method according to any one of claims 1-11.

16. A computer readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program comprises program instructions. The program instructions are run by the computer to cause the computer to execute the posture recognition and correction method according to any one of claims 1-11.

Images