A control mode switching method and wearable device based on wrist rotation direction recognition

Abstract

This invention discloses a control mode switching method and wearable device based on wrist rotation direction recognition. The system defaults to a free mode, allowing free hand movement. When wrist outward rotation is detected to reach an adaptive threshold, the system enters control mode, and the gesture is interpreted as a control command for the target device. When wrist inward rotation is detected, the system exits control mode, and the hand returns to free movement. Two consecutive rapid rotations detected in any mode trigger an emergency interrupt. The rotation threshold can be personalized through initial calibration, dynamic adjustment based on historical data, or manual setting. This invention achieves intuitive, safe, and widely applicable seamless switching of control modes through direction-discriminate state machine logic and an adaptive threshold scheme, and can be widely used in exoskeleton control, AR / VR interaction, drone control, and other scenarios.

Description

Technical Field

[0001] This invention belongs to the field of human-computer interaction technology, specifically relating to a method and wearable device for switching between control mode and free mode by recognizing the direction of the user's wrist rotation. Background Technology

[0002] Existing gesture control solutions generally suffer from the following pain points:

[0003] Continuous hand occupation issue: The user's hand must always remain in the "controller" role, and everyday actions can easily be misinterpreted as control commands. For example, when controlling a drone with gestures, the operator's hand needs to continuously make specific gestures, and once relaxed or performing other everyday actions, it may be accidentally triggered.

[0004] Lack of a clear mode switching mechanism: Existing solutions mostly rely on physical buttons, voice commands, or touch operations with the other hand to switch modes, which is extremely inconvenient when both hands are occupied and interrupts the natural flow of the current task.

[0005] Lack of personalized adaptation: The trigger thresholds of existing gesture recognition solutions are mostly fixed values, which do not take into account the individual differences in the range of wrist movements of different users (especially the elderly, hemiplegic patients or those with limited hand movement).

[0006] In the smartwatch industry, there are technical solutions that trigger specific functions (such as answering calls or silencing calls) with a single wrist flip. However, such solutions are essentially "single-action triggering single-function," meaning that wrist flips act as a virtual button, executing one action per press, and cannot achieve persistent mode switching or complex command encoding. Their interaction logic is unidirectional triggering, not state switching. Summary of the Invention

[0007] Purpose of the invention:

[0008] This provides a method for seamlessly switching between "control mode" and "free mode" using only wrist rotation, and can adapt to different users' wrist mobility.

[0009] Technical solution:

[0010] The system defines two working states, which are switched by a rapid flipping motion of the wrist wearing the wearable device:

[0011] Step S1 (Default Free): The system runs in Free Mode by default after startup. In this mode, the hand wearing the device can freely perform any daily activities. The system does not analyze the movement data for control commands, but only continuously monitors the wrist posture.

[0012] Step S2 (Reverse Wake-up): Real-time detection of hand posture. When the system detects that the wrist has completed a rapid flipping motion in the first direction (such as outward flipping) that meets preset conditions, the system switches from free mode to control mode. After a successful switch, confirmation is provided to the user via vibration feedback.

[0013] Step S3 (Control Execution): In control mode, the system parses the preset hand gestures and / or motion data into control commands for the target device. Target devices include, but are not limited to: exoskeleton robots, drones, AR / VR headsets, smart home devices, computers, etc.

[0014] Step S4 (Reverse Exit): In control mode, when the system detects again that the wrist has completed a rapid rotation movement in the second direction (such as inward rotation) that meets the preset conditions, the system switches back to free mode. Afterward, it stops parsing control commands, and the hand resumes free movement.

[0015] Step S5 (Emergency Interruption): Regardless of the current mode, when two consecutive rapid flips in any direction are detected within a preset time window (e.g., within 1.5 seconds), the system unconditionally executes an emergency interrupt, forcibly switches to free mode, and stops sending all control commands.

[0016] The preset conditions for the "rapid flipping action" are:

[0017] The change in angle around the wrist roll axis or pitch axis exceeds a preset threshold (e.g., greater than 90 degrees) within a preset time window (e.g., 0.3 seconds).

[0018] Or, the peak angular velocity around a specific axis exceeds a preset threshold (e.g., greater than 200 degrees / second).

[0019] Furthermore, after flipping, it remains in the target position for more than a preset duration (such as 0.3 seconds) to distinguish it from unconscious wrist rotations in daily life.

[0020] Adaptive threshold calibration scheme:

[0021] The aforementioned flipping threshold is not fixed, but can be adaptively adjusted according to individual user circumstances, specifically including at least one of the following three methods:

[0022] Automatic calibration upon initial use: After the user wears the device for the first time, the accompanying app or device voice prompts the user to "slowly rotate your wrist outward to your most comfortable maximum position, and then hold for 2 seconds." The system records the maximum rotation angle the user can achieve at this time using the inertial measurement unit (for example, an elderly person or a hemiplegic patient may only be able to rotate their healthy wrist outward to a maximum of 60 degrees), and automatically sets a preset percentage of this value (such as 80%, or 48 degrees) as the user's personalized trigger threshold. The same method can be used to calibrate the inversion threshold. The entire process takes only 1-2 minutes.

[0023] Continuous dynamic adjustment: During daily use, the system continuously records the angle data when the user actually triggers the rotation. If the user's wrist mobility gradually improves during rehabilitation and the actual rotation angle repeatedly exceeds the current threshold, the system can automatically adjust the threshold to a level closer to the user's current ability (e.g., gradually adjusting from 48 degrees to 60 degrees, 72 degrees), always maintaining the best match.

[0024] Manual settings: Users or their therapists and caregivers can directly input values ​​or drag sliders through the settings interface of the accompanying app to customize parameters such as trigger angle, angular velocity threshold and pause time, achieving fully personalized configuration.

[0025] Beneficial effects:

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] Truly freeing your hands: The "default freedom" design fundamentally eliminates the possibility of accidental triggering. Users only need to "awaken" the control function with a clear flip action when needed, and "return" freedom with another clear flip action after use, achieving a seamless interactive experience of "coming when needed and leaving when finished".

[0028] Intuitive operation: By distinguishing the flip direction (outward flip = enter control, inward flip = exit control), users do not need to remember the current system state, making operation intuitive and natural. This direction distinction also makes the system's recognition accuracy far exceed that of solutions based on number of attempts or time.

[0029] Absolutely safe: Setting "two consecutive rapid flips" as an independent, highest-priority emergency interrupt command provides users with a "safe word" that can be quickly and accurately triggered in any emergency, fundamentally solving the safety hazard of gesture control being difficult to reliably interrupt in critical moments.

[0030] Personalized adaptation: The adaptive threshold scheme enables this method to be adapted to a wide range of people with varying wrist mobility, including hemiplegic patients, the elderly, patients in the rehabilitation period, and healthy people, significantly expanding the applicable scenarios. Attached Figure Description

[0032] Figure 1 System state machine schematic diagram. This diagram illustrates the core state transition logic of this invention: the system defaults to free mode; when a first-direction flip (e.g., outward flip) is detected, it switches to control mode; when a second-direction flip (e.g., inward flip) is detected, it switches back to free mode; in any mode, detecting two consecutive rapid flips triggers an emergency interrupt and forces a return to free mode. The flip determination threshold can be determined through adaptive calibration.

[0033] Figure 2 : Schematic diagram defining wrist rotation direction. This diagram, with the back of the hand facing upwards as the natural position, defines two rotational directions around the forearm axis (roll axis): rotation towards the little finger side (palm facing upwards) is "outward rotation" (external rotation), and rotation towards the thumb side (palm facing downwards) is "inward rotation" (internal rotation). The actual trigger threshold is adaptively calibrated based on individual user circumstances.

[0034] Figure 3 System hardware block diagram. This diagram illustrates the overall architecture of the wearable device and the target device communicating via BLE. The wearable device integrates a 6-axis IMU, a low-power Bluetooth MCU (including an adaptive calibration module), and a vibration motor; the target device includes a command parsing module and a function execution module. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and specific examples.

[0037] Example 1: Rehabilitation Exoskeleton Control

[0038] In this embodiment, the target device is a rehabilitation exoskeleton glove worn by a hemiplegic patient on the affected side of their hand.

[0039] Hemiplegic patients wear a wristband with an integrated 6-axis IMU on their unaffected wrist. Upon first use, adaptive calibration is completed via the accompanying app: the system guides the patient to rotate their unaffected wrist outward to the most comfortable position and hold it for 2 seconds, automatically recording the angle and setting a personalized threshold.

[0040] In daily use, the system is in free mode by default, allowing the patient's healthy hand to freely perform activities such as turning pages, operating a mobile phone, and picking up items, without accidentally triggering the mobile phone lock on the affected side.

[0041] When a patient needs to use their affected hand to grasp an object (such as a water cup), they rotate their unaffected wrist outward to a pre-defined threshold. A single wristband vibration confirms entry into control mode. At this point, the patient makes a fist with their unaffected hand; this gesture is interpreted as a "grasp" command and sent via Bluetooth to the affected hand's device. This drives a motor to close the fingers and grasp the water cup. Once the grasp is complete, the patient rotates their unaffected wrist inward, and a single wristband vibration confirms exiting control mode. The unaffected hand is then free to perform other tasks (such as answering the phone or opening a door). The affected hand's device maintains the grasping state using its built-in mechanical self-locking mechanism, eliminating the need for continuous power supply to the motor.

[0042] When the patient needs to put down the water cup, the wrist is turned outward again to enter control mode, the healthy hand makes an opening gesture, and the affected hand unlocks and releases the phone.

[0043] During use, in case of an emergency (such as a sudden spasm in the affected hand), the patient only needs to quickly rotate the wrist of the healthy side twice in succession to trigger an emergency interruption. The affected hand glove will be forcibly unlocked and released from grip, and the system will automatically switch back to free mode.

[0044] Example 2: AR Glasses Menu Control

[0045] In this embodiment, the target device is augmented reality (AR) glasses.

[0046] The user wears AR glasses and a ring with an integrated IMU. The system defaults to free mode, where the user has normal vision and no virtual interface interference. When the user needs to operate the AR menu, they flip the finger wearing the ring outwards, the ring vibrates to confirm, and the AR glasses bring up the virtual menu. The user then operates using preset finger gestures (such as pinching the thumb and forefinger to confirm, or sliding the forefinger to select). After the operation is completed, flipping the finger inwards closes the virtual menu and returns to free vision.

[0047] Example 3: One-handed operation of a drone

[0048] In this embodiment, the target device is a consumer-grade drone.

[0049] The operator wears an IMU bracelet. In the default free mode, the drone flies autonomously along a preset route. When the operator needs to take over manually, they turn their wrist outward; the bracelet vibrates to confirm, and the drone switches to follow mode. The operator then controls the flight direction using hand gestures. After completing the task, turning the wrist inward exits follow mode, and the drone automatically hovers, awaiting the next command.

[0050] Example 4: Detailed Process of Adaptive Threshold Calibration

[0051] This embodiment describes in detail the specific implementation process of three methods for adaptive threshold calibration.

[0052] Method 1 (Initial Guided Calibration): The user clicks "Start Calibration" in the App. The App guides the user with animation and voice prompts, "Please slowly rotate your wrist outward to the maximum position and hold for 2 seconds." The MCU continuously reads the IMU roll axis angle data at a 100Hz sampling rate and records the stable maximum value within 2 seconds as the user's maximum outward rotation angle, MaxOutAngle. The system automatically sets the trigger threshold, TriggerOutAngle, to MaxOutAngle × 80%. Similarly, the user is guided to rotate inward, MaxInAngle is recorded, and TriggerInAngle is set to MaxInAngle × 80%. Upon completion of calibration, the App displays "Calibration Successful" and shows the currently set threshold for user confirmation.

[0053] Method 2 (Continuous Dynamic Adjustment): The system records the actual flip angle ActualAngle each time a mode switch is successfully triggered. When the average ActualAngle value of the most recent N times (e.g., 20 times) is consistently higher than the preset percentage (e.g., 110%) of the current threshold, the system automatically adjusts the threshold to 90% of the average value and notifies the user via App push notification "Your activity ability has improved, and the trigger threshold has been automatically adjusted".

[0054] Method 3 (Manual Setting): The App settings interface provides sliders and input boxes, allowing users or therapists to directly adjust four parameters: eversion threshold, inversion threshold, rotation speed threshold (angular velocity), and pause time. The adjustments take effect immediately.

Claims

1. A control mode switching method based on wrist flip direction recognition, applied to a wearable device comprising an inertial sensor, characterized in that, Includes the following steps: The system defaults to free mode, in which the motion data of the hand wearing the device is not parsed into control commands for the target device; The system detects the wrist posture of the hand in real time. When it recognizes that the wrist has completed a rapid flipping motion in the first direction that meets the preset conditions, the system switches from the free mode to the control mode. In the control mode, the preset gestures and / or motion data of the hand are parsed into control commands for the target device; When the system detects that the wrist has completed a rapid flipping motion in the second direction that meets the preset conditions, it switches from the control mode back to the free mode and stops parsing the hand's motion data into control commands. Wherein, the first direction and the second direction are opposite or different wrist rotation directions.

2. The method of claim 1, wherein, The first direction is outward flipping, and the second direction is inward flipping; or vice versa.

3. The method according to claim 1, characterized in that, Also includes: Regardless of the current mode, when two consecutive rapid flips in any direction are detected within a preset time window, an emergency interrupt is executed unconditionally, the system immediately switches to free mode and stops sending all control commands.

4. The method according to claim 1, characterized in that, The preset conditions for the "rapid flipping action" include: within a preset time window, the change in the angle of the wrist around a specific axis or the peak value of the angular velocity exceeds a preset threshold.

5. The method according to claim 4, characterized in that, The preset conditions also include: after flipping, the wrist remains in the target position for more than a preset time, so as to distinguish it from unconscious wrist rotation in daily life.

6. The method according to claim 4, characterized in that, The preset threshold can be determined in at least one of the following ways: Upon first use, the system automatically calibrates by guiding the user to complete a maximum rotation. During use, the threshold is dynamically adjusted by recording the user's historical flipping data; It is manually set by the user or caregiver through the accompanying software.

7. The method according to claim 1, characterized in that, When the mode switch is successful, the wearable device provides tactile feedback confirmation to the user.

8. A wearable control device for implementing the method of any one of claims 1 to 7, characterized in that, include: Inertial measurement unit: worn on the user's wrist, used to detect wrist rotation direction and posture changes; Processing unit: Executes the mode switching logic; Communication unit: used to send control commands to the target device and receive feedback; Feedback unit: Used to provide users with confirmation prompts when switching modes.

9. The apparatus according to claim 8, characterized in that, The wearable device may be in the form of a wristband, bracelet, glove, ring, or smartwatch.

10. An interactive system comprising the apparatus of claim 8 or 9, and a target device communicatively connected to the apparatus, the target device being at least one of the following: a smartphone, an augmented reality / virtual reality device, a drone, a smart home device, or an exoskeleton robot.

Images