Search

CN-121980167-A - Action recognition method based on electromyographic signals, equipment control method and electronic equipment

CN121980167ACN 121980167 ACN121980167 ACN 121980167ACN-121980167-A

Abstract

The embodiment of the invention discloses an electromyographic signal-based action recognition method, an electromyographic signal-based equipment control method, electronic equipment and a storage medium. The electromyographic signal-based action recognition method is used for recognizing specific actions executed by a target part of a user, wherein the target part is provided with at least two muscle groups, the method comprises the steps of obtaining electromyographic signals collected by each muscle group, and determining the specific actions executed by the target part according to the energy of each electromyographic signal. The scheme can accurately distinguish different specific actions of the same part by utilizing the electromyographic signals generated by each of a plurality of muscle groups.

Inventors

  • ZHANG RUIZHONG

Assignees

  • 北京镁伽机器人科技有限公司

Dates

Publication Date
20260505
Application Date
20251224

Claims (10)

  1. 1. An electromyographic signal-based action recognition method is characterized by being used for recognizing a specific action executed by a target part of a user, wherein the target part is provided with at least two muscle groups, and the method comprises the following steps: Acquiring myoelectric signals acquired by each muscle group; and determining the specific action executed by the target part according to the respective energy of each electromyographic signal.
  2. 2. The method as recited in claim 1, further comprising: Pre-configuring a preset mapping relation between electromyographic signal energy and a plurality of action types, wherein the preset mapping relation comprises a one-to-one correspondence relation between a plurality of possible comparison results between the electromyographic signal energy and the plurality of action types; And determining a specific action executed by the target part according to the respective energy of each electromyographic signal, wherein the specific action comprises the following steps: and determining the action type of the specific action executed by the target part according to the comparison result between the energy of at least two electromyographic signals acquired by the at least two muscle groups in one-to-one correspondence and the preset mapping relation.
  3. 3. The method of claim 2, wherein the target site is a wrist joint, the at least two muscle groups include a longus palmaris and a extensor carpi radialis, and the predetermined mapping relationship includes: when the comparison result is that the energy of the first electromyographic signal is larger than the product of the energy of the second electromyographic signal and the preset multiple, the corresponding action type of the specific action is wrist bending action; when the comparison result shows that the energy of the first electromyographic signal is larger than the energy of the second electromyographic signal and smaller than the product of the energy of the second electromyographic signal and the preset multiple, the corresponding action type of the specific action is cubit bias action; When the comparison result shows that the energy of the second electromyographic signal is larger than the energy of the first electromyographic signal and smaller than the product of the energy of the first electromyographic signal and the preset multiple, the corresponding action type of the specific action is wrist radial deviation action; When the comparison result shows that the energy of the second electromyographic signal is larger than the product of the energy of the first electromyographic signal and the preset multiple, the corresponding action type of the specific action is wrist stretching action; Wherein the first electromyographic signal is an electromyographic signal collected for the palmaris longus; the second electromyographic signal is an electromyographic signal collected for extensor carpi radialis.
  4. 4. A method according to any one of claims 1-3, wherein the method further comprises: And determining the action times of the specific action within the preset time according to the target electromyographic signals generated by the target muscle group for executing the specific action within the preset time.
  5. 5. The method of claim 4, further comprising, prior to determining the number of actions of the particular action within the preset time: When the target site is a wrist joint, the at least two muscle groups include a longus palmaris and a extensor carpi radialis, the longus palmaris is determined as the target muscle group when the specific motion type is a wrist flexion motion or a cubit deviation motion, and the extensor carpi radialis is determined as the target muscle group when the specific motion type is a wrist extension motion or a wrist radiodeviation motion.
  6. 6. The method of claim 4, wherein the preset time comprises a plurality of time windows, and wherein the determining the number of actions of the particular action within the preset time comprises: Counting energy values of the target electromyographic signals acquired in each of the plurality of time windows to determine the number of target windows with energy values greater than a preset energy threshold; wherein the number of actions of the specific action is equal to the number of target windows.
  7. 7. A device control method, characterized by being used for controlling a peripheral auxiliary device which a user desires to operate, the peripheral auxiliary device being in communication connection with an electroencephalogram signal acquisition device and an electromyogram signal acquisition device, the electroencephalogram signal acquisition device being used for acquiring an electroencephalogram signal of the user, the electromyogram signal acquisition device being used for acquiring an electromyogram signal of the user, the method comprising: Acquiring an electroencephalogram signal acquired by the electroencephalogram signal acquisition equipment, and determining a first target action corresponding to the electroencephalogram signal, wherein the electroencephalogram signal is generated due to intention associated with the first target action in the consciousness of the user; determining a specific action by adopting the action recognition method based on the electromyographic signals according to any one of claims 1-6 based on the electromyographic signals acquired by the electromyographic signal acquisition device when the user executes the specific action, and determining a second target action corresponding to the recognized specific action according to the correspondence between a plurality of preset specific actions and a plurality of target actions; and generating control information according to the first target action and the second target action, wherein the control information is used for controlling the movement of the peripheral auxiliary equipment to execute the combined action of the first target action and the second target action.
  8. 8. The method of claim 7, wherein the plurality of target actions comprises a plurality of target action sets, each target action set comprising one or more target actions, the plurality of target action sets each comprising a different target action; for each target action set of the plurality of target action sets, the target action set can be associated to other target action sets except the target action set through the plurality of action types, and the plurality of action types and the other target action sets have a one-to-one correspondence; Determining a second target action corresponding to the identified specific action according to the corresponding relation between the preset specific actions and the target actions, wherein the second target action comprises the following steps: Determining a first target action set in which the first target action is located according to the corresponding relation between a plurality of preset action types and a plurality of target action sets, and determining a second target action set associated with the first target action set through the identified action types according to the one-to-one corresponding relation between the plurality of action types and the target action sets; and selecting the action in the second target action set according to the identified action times as the second target action.
  9. 9. An electronic device comprising a processor and a memory, characterized in that the memory has stored therein computer program instructions which, when executed by the processor, are adapted to carry out the electromyographic signal based action recognition method according to any of claims 1-6 and/or the device control method according to any of claims 7-8.
  10. 10. A storage medium having stored thereon program instructions for performing the electromyographic signal based action recognition method according to any of claims 1-6 and/or the device control method according to any of claims 7-8 when run.

Description

Action recognition method based on electromyographic signals, equipment control method and electronic equipment Technical Field The present invention relates to the field of biometric identification, and in particular, to an electromyographic signal-based motion identification method, an apparatus control method, an electronic apparatus, and a storage medium. Background The human body surface electromyographic signals are closely related to the generation of motion, and thus, the human body surface electromyographic signals can be used for motion recognition. Electromyographic signal recognition techniques have been applied in a number of fields, such as prosthetic control, rehabilitation engineering, human-computer interaction, etc. However, in the conventional electromyographic signal recognition technology, it is difficult to distinguish complex actions generated by the cooperative work of a plurality of muscle groups by the electromyographic signals. Disclosure of Invention The present invention has been made in view of the above-described problems. The invention provides an electromyographic signal-based action recognition method, an electromyographic signal-based action recognition device, an electromyographic signal-based action control method, an electromyographic device and a storage medium. The scheme can accurately distinguish different specific actions of the same part by utilizing the electromyographic signals generated by each of a plurality of muscle groups. According to one aspect of the invention, an electromyographic signal-based action recognition method is provided, and is used for recognizing a specific action executed by a target part of a user, wherein the target part is provided with at least two muscle groups. Optionally, the method further comprises the step of pre-configuring a preset mapping relation between electromyographic signal energy and a plurality of action types, wherein the preset mapping relation comprises a one-to-one correspondence relation between a plurality of possible comparison results between the energy of at least two electromyographic signals and the plurality of action types, and the step of determining the specific action executed by the target part according to the respective energy of each electromyographic signal comprises the step of determining the action type of the specific action executed by the target part according to the comparison result between the respective energy of at least two electromyographic signals acquired by at least two muscle groups in one-to-one correspondence and the preset mapping relation. Optionally, the target site is a wrist joint, the at least two muscle groups comprise a longus palmaris muscle and a extensor carpi radialis muscle, and the preset mapping relation comprises a wrist deflection action corresponding to a specific action type when the comparison result is that the energy of a first electromyographic signal is larger than the product of the energy of a second electromyographic signal and a preset multiple, a extensor carpi radialis action corresponding to a specific action type when the comparison result is that the energy of the first electromyographic signal is larger than the energy of the second electromyographic signal and smaller than the product of the energy of the first electromyographic signal and a preset multiple, a extensor carpi radialis action corresponding to a specific action type when the comparison result is that the energy of the second electromyographic signal is larger than the product of the first electromyographic signal and a preset multiple, wherein the first electromyographic signal is the electrical signal, and the extensor carpi radialis action corresponding to a specific action corresponding to the specific action type when the comparison result is that the energy of the second electromyographic signal is larger than the product of the first electromyographic signal and the preset multiple. Optionally, the method further comprises determining the number of actions of the specific action within the preset time according to target electromyographic signals generated by target muscle groups for executing the specific action within the preset time. Optionally, before determining the number of actions of the specific action within the preset time, the method further comprises determining the longus palmaris as a target muscle group when the action type of the specific action is a wrist bending action or a cubit deviation action and determining the extensor carpi radialis as a target muscle group when the action type of the specific action is a wrist stretching action or a wrist radial deviation action in the case that the target site is a wrist joint and at least two muscle groups comprise the longus palmaris and extensor radialis. Optionally, the preset time comprises a plurality of time windows, and determining the action times of the specific actions in the preset time comprises counti