CN-121657873-B - State identification method, device, electronic equipment and computer readable storage medium

CN121657873BCN 121657873 BCN121657873 BCN 121657873BCN-121657873-B

Abstract

The application discloses a state identification method, a state identification device, electronic equipment and a computer readable storage medium, and relates to the technical field of data processing, wherein the method comprises the following steps: when the wrist of the target user rotates, calculating accumulated rotation angle information according to the angular velocity information of the wrist rotation, calculating maximum continuous increase span information corresponding to the accumulated rotation angle information in the acquisition window according to a dual-stack aggregation strategy, and determining rotation state information of the wrist of the target user according to the maximum continuous increase span information and an angle threshold value. The calculation complexity of the maximum continuous increase span information is reduced, and therefore when the wrist rotation state information is identified and determined based on the maximum continuous increase span information, the identification efficiency of the wrist rotation state information can be improved, the resource occupation is reduced, and the instantaneity is improved.

Inventors

WANG YE
LI HAI
HE YIJIA

Assignees

雷鸟创新技术(深圳)有限公司

Dates

Publication Date: 20260512
Application Date: 20260203

Claims (9)

1. A method of state identification, the method comprising: When the wrist of the target user rotates, calculating accumulated rotation angle information according to the angular velocity information of the wrist rotation; calculating maximum continuous increase span information corresponding to accumulated rotation angle information in the acquisition window according to a dual-stack aggregation strategy; determining rotation state information of the wrist of the target user according to the maximum continuous growth span information and an angle threshold value; the calculating the maximum continuous growth span information corresponding to the accumulated rotation angle information in the acquisition window according to the dual-stack aggregation strategy comprises the following steps: Executing a push operation for an input stack and an output stack according to the accumulated rotation angle information by a collection window, wherein the accumulated rotation angle information of the input stack and the output stack has time sequence; Determining input aggregation information corresponding to the accumulated rotation angle information in the input stack, and determining output aggregation information corresponding to the accumulated rotation angle information in the output stack; Combining the input aggregation information and the output aggregation information to obtain the maximum continuous increase span information corresponding to the acquisition window; The input aggregation information comprises minimum input rotation angle information, maximum input rotation angle information and maximum input continuous span information; The determining the input aggregation information corresponding to the accumulated rotation angle information in the input stack includes: Setting the input stack top aggregation information corresponding to the stack top of the input stack as the input aggregation information; And when the accumulated rotation angle information is pushed onto the input stack, updating the minimum input rotation angle information, the maximum input rotation angle information and the maximum input continuous span information in the original input stack top aggregation information of the input stack according to the accumulated rotation angle information.
2. The state recognition method according to claim 1, wherein the updating of the maximum input continuous span information includes the steps of: calculating the difference between the accumulated rotation angle information and the minimum input rotation angle information in the original input stack top aggregation information to obtain input span information; and screening out the maximum input continuous span information of the original input stack top aggregation information, the self span information corresponding to the accumulated rotation angle information and the maximum value in the input span information, and obtaining updated maximum input continuous span information.
3. The state recognition method according to claim 2, wherein the output aggregation information includes minimum output rotation angle information, maximum output rotation angle information, and maximum output continuous span information; the determining output aggregation information corresponding to the accumulated rotation angle information in the output stack comprises the following steps: setting output stack top aggregation information corresponding to the stack top of the output stack as the output aggregation information; the output stack top aggregation information is obtained by updating minimum output rotation angle information, maximum output rotation angle information and maximum output continuous span information in the original output stack top aggregation information of the output stack according to the accumulated rotation angle information when the accumulated rotation angle information is inverted from the input stack to the output stack; the updating of the maximum output continuous span information comprises the following steps: Calculating the difference between the maximum output rotation angle information and the accumulated rotation angle information in the original output stack top aggregation information to obtain output span information; And screening out maximum output continuous span information of the original output stack top aggregation information, self span information corresponding to the accumulated rotation angle information and the maximum value in the output span information, and obtaining updated maximum output continuous span information.
4. The method of claim 3, wherein the merging the input aggregate information and the output aggregate information to obtain the maximum continuously-increasing span information corresponding to the collection window comprises: calculating the difference between the maximum input rotation angle information corresponding to the input aggregation information and the minimum output rotation angle information corresponding to the output aggregation information to obtain aggregation span information; and screening out the maximum value among the maximum input continuous span information, the maximum output continuous span information and the aggregate span information to obtain the maximum continuous increase span information.
5. The state recognition method according to claim 1, wherein the performing a push operation for an input stack and an output stack on the accumulated rotation angle information according to an acquisition window includes: adding the accumulated rotation angle information to the input stack in chronological order; If the data quantity of the accumulated rotation angle information in the input stack exceeds the length of the acquisition window, rewinding the accumulated rotation angle information in the input stack to the output stack; And controlling the stacking of the input stack and the stacking-out of the output stack according to the sliding of the acquisition window and the length.
6. A method of state identification according to claim 3, wherein the method further comprises: Screening out a minimum value from the minimum input rotation angle information and the minimum output rotation angle information to obtain target minimum rotation angle information; screening the maximum value from the maximum input rotation angle information and the maximum output rotation angle information to obtain target maximum rotation angle information; Calculating the difference between the target maximum rotation angle information and the target minimum rotation angle information to obtain maximum swing information; Carrying out abnormal span identification on the maximum continuous increase span information according to the maximum swing information to obtain an abnormal span identification result; and if the abnormal span identification result comprises span abnormality, controlling not to carry out the determining flow of the rotation state information or controlling not to respond to the rotation state information.
7. A state recognition device, the device comprising: The detection module is used for calculating accumulated rotation angle information according to the angular velocity information of wrist rotation when the wrist of the target user rotates; The calculation module is used for calculating the maximum continuous increase span information corresponding to the accumulated rotation angle information in the acquisition window according to the dual-stack aggregation strategy; the determining module is used for determining the rotation state information of the wrist of the target user according to the maximum continuous growth span information and the angle threshold value; the calculating the maximum continuous growth span information corresponding to the accumulated rotation angle information in the acquisition window according to the dual-stack aggregation strategy comprises the following steps: Executing a push operation for an input stack and an output stack according to the accumulated rotation angle information by a collection window, wherein the accumulated rotation angle information of the input stack and the output stack has time sequence; Determining input aggregation information corresponding to the accumulated rotation angle information in the input stack, and determining output aggregation information corresponding to the accumulated rotation angle information in the output stack; Combining the input aggregation information and the output aggregation information to obtain the maximum continuous increase span information corresponding to the acquisition window; The input aggregation information comprises minimum input rotation angle information, maximum input rotation angle information and maximum input continuous span information; The determining the input aggregation information corresponding to the accumulated rotation angle information in the input stack includes: Setting the input stack top aggregation information corresponding to the stack top of the input stack as the input aggregation information; And when the accumulated rotation angle information is pushed onto the input stack, updating the minimum input rotation angle information, the maximum input rotation angle information and the maximum input continuous span information in the original input stack top aggregation information of the input stack according to the accumulated rotation angle information.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the state identification method of any of claims 1-6 when the computer program is executed.
9. A computer readable storage medium, characterized in that the computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the state identification method according to any of claims 1-6.

Description

State identification method, device, electronic equipment and computer readable storage medium Technical Field The present application relates to the field of data processing technologies, and in particular, to a method and apparatus for identifying a state, an electronic device, and a computer readable storage medium. Background With the popularity of Augmented Reality (AR) and Virtual Reality (VR) technologies, smart wearable devices (e.g., smart watches, bracelets, etc.) are becoming an important human-computer interaction portal. In the interactive design of such devices, a state recognition for "wrist rotation reaching a specific gesture or angle interval" has a key role, which is often used as an activation switch or precondition for the interaction, e.g. the system responds to subsequent fine gesture instructions (e.g. single click, double click, pinch, etc.) only when it is detected that the user's wrist has rotated to a target state (e.g. wrist rotation satisfying "palm rest"), to reduce false triggering rate and improve reliability of the interaction. To meet the need for real-time interaction, wearable devices typically continuously collect gyroscope data in an Inertial Measurement Unit (IMU) at a higher sampling rate (e.g., 100 Hz or higher) and require real-time analysis and determination of wrist rotation over a window of time in the past (e.g., the last about 1 second, corresponding to 100-150 frames of data). The prior art mainly adopts two schemes for wrist rotation state identification, namely scheme A (global integration method) for carrying out accumulated integration on axial data of a gyroscope related to wrist rotation so as to estimate a rotation angle and judging whether the current angle is in a preset absolute angle range. Scheme B (traditional sliding window traversal) maintains a fixed length (e.g., n=150) fifo. All data within the window is traversed at each sampling instant to determine if enough rotation has occurred in the last period of time to reach the target state. However, the global integral of scheme a is susceptible to angular drift due to zero bias noise of the gyroscope, and the longer the run time, the larger the accumulated error, resulting in state decision misalignment. Every new data point in the scheme B needs to traverse a window with the length of N, and the time complexity is high. Resulting in significant CPU resource occupation and insufficient real-time. Therefore, a technical solution capable of efficiently recognizing the rotation state of the wrist is needed. Disclosure of Invention The embodiment of the application provides a state identification method, a state identification device, electronic equipment and a computer readable storage medium, which can improve the identification efficiency of a wrist rotation state, reduce the occupation of resources and improve the instantaneity. In a first aspect, an embodiment of the present application provides a method for identifying a state, where the method includes: when the wrist of the target user rotates, calculating accumulated rotation angle information according to the angular velocity information of the wrist rotation; calculating maximum continuous increase span information corresponding to accumulated rotation angle information in the acquisition window according to a dual-stack aggregation strategy; and determining the rotation state information of the wrist of the target user according to the maximum continuous growth span information and the angle threshold value. In a second aspect, an embodiment of the present application further provides a status identifying apparatus, where the apparatus includes: the detection module is used for calculating accumulated rotation angle information according to the angular velocity information of the wrist rotation when the wrist of the target user rotates; The calculation module is used for calculating the maximum continuous increase span information corresponding to the accumulated rotation angle information in the acquisition window according to the dual-stack aggregation strategy; and the determining module is used for determining the rotation state information of the wrist of the target user according to the maximum continuous growth span information and the angle threshold value. Optionally, in some embodiments of the present application, the calculating, according to a dual stack aggregation policy, maximum continuous growth span information corresponding to cumulative rotation angle information in an acquisition window includes: Executing a push operation for an input stack and an output stack according to the accumulated rotation angle information by a collection window, wherein the accumulated rotation angle information of the input stack and the output stack has time sequence; Determining input aggregation information corresponding to the accumulated rotation angle information in the input stack, and determining output aggregation information corr