CN110727353A - Method and device for controlling control components defined based on two-dimensional intent - Google Patents

Abstract

A control component control method and device defined based on a two-dimensional intent, characterized in that it includes a human body gesture action for expressing directional control intent and a head expression motion for expressing confirmation intent. The movement direction is controlled, and the purpose of confirming the intention is to confirm the current operation of the control component and its operation type; its beneficial technical effect is to use the two-dimensional combination of human body gestures and head expression movements or voice commands, not only to establish a body suitable for the body. The direction of the action controls the intention command mechanism, and an adaptive but independent "confirmation" intention command mechanism is also established, which overcomes the problem of uncontrolled voice technology and the limitations of EOG technology alone. , the effective cooperation between them effectively solves the problem of multi-dimensional control of human-computer interaction equipment, as well as the coordination problem between human actions and control intentions.

Description

Method and device for controlling control components defined based on two-dimensional intent

technical field

The present invention belongs to the technical field of human-computer interaction, and in particular relates to a control method and a control device for a control component defined based on two-dimensional intent, such as a control method and device based on EOG signals and head posture. EOG (electro-oculogram) membrane electro-oculogram is an electrical recording of eye movements, which is a detection of eye potential.

Background technique

In the prior art, the interaction between people and operating equipment (such as head-mounted display control components, computers, mobile phones and other living equipment is mainly through manual operation. For example, when a person interacts with a head-mounted display device, Physical buttons can be used to increase the volume, play or pause; when people interact with the computer, they need to manually operate the keyboard or mouse to play or open; It is very difficult to realize human-computer interaction by using traditional input devices (such as mouse, keyboard, manipulator, etc.)

In the field of human-computer interaction, the eyes are another important information interaction channel, and the line of sight reflects the direction of human attention. Therefore, the application of the line of sight to the field of human-computer interaction has the characteristics of naturalness, directness and interactivity, and has attracted much attention. people's attention. Patent Publication No. China CN104866100B discloses an eye control device, an eye control method and an eye control system, wherein the eye control device includes: a gaze point acquisition unit, a human eye motion detection unit and a control signal generation unit. The unit is used to obtain the position information of the gaze point of the human eye on the device to be operated; the human eye action detection unit is used to detect whether the human eye has made a preset action, and when it is detected that the human eye has made a preset action, it controls the gaze point to obtain The unit sends the current position information of the gaze point of the current human eye on the device to be operated to the control signal generation unit; the control signal generation unit is used to generate a corresponding table corresponding to the current position information according to the pre-stored position control correspondence table corresponding to the device to be operated. corresponding control signals, and send the control signals to the device to be operated so that the device to be operated performs corresponding operations. The technical solution can effectively realize the use of human eye movements to control the operating device, but this control method is too simple and has a single dimension, and is not suitable for use on human-computer interaction equipment with multi-dimensional control requirements.

SUMMARY OF THE INVENTION

In order to easily solve the problem of human-computer interaction in an unfavorable living environment, at least people first thought of language control technology, that is, using the instructions given by the human language itself, such as moving up, down, left, right, walking, etc., to let the controlled device follow the language command itself. However, the delay of the language itself and the limited degree of control, that is, the problem of uncontrolled control, determine that a single speech technology cannot be used completely and reliably in the field of human-computer interaction; secondly, someone also thought of it first. The eye control technology mentioned above in the prior art is also called EOG technology, and obviously this method is also a single control mode in control. None of these techniques can be done well when human-computer interaction devices require multi-dimensional control.

In order to solve multi-dimensional control problems such as human-computer interaction equipment, and to ensure the convenience and reliability of control, we have designed several control methods:

The first is a control component control method defined based on a two-dimensional intention, which is characterized in that it includes a human body gesture action for expressing a direction control intention and a head expression action for expressing a confirmation intention, and the direction control intention is to controlling the direction of movement of the control part, the confirmation is intended to confirm the current operation of the control part and its type of operation;

Implement the human body posture action, pick up the posture feature value of the action representation itself that characterizes the human body posture action, set the posture fusion calculation module, and identify the corresponding posture according to the picked up posture feature value and based on the posture fusion calculation module. The movement direction control command of the human body posture action, and the movement direction of the control component is controlled by the movement direction control command;

Implement the head expression action, pick up the bioelectric signal characteristic value of the head caused by the head expression action, set up a bioelectric signal calculation module, according to the picked up bioelectric signal characteristic value and based on the The bioelectric signal calculation module identifies a confirmation instruction corresponding to the head expression, and confirms the current operation of the control component and its operation type through the confirmation instruction.

The present invention also provides a two-dimensional intention-based control device for realizing the first control method, which is characterized by comprising:

An attitude feature value pickup device, the attitude feature value pickup device is used to pick up the attitude feature value of the action representation itself that characterizes the posture action of the human body and give the corresponding posture feature signal, and the human body posture action is used to express the movement of the control part the control intention to control the direction;

A head bioelectric signal characteristic value pickup device, the head bioelectric signal characteristic value pickup device is used to pick up the head bioelectric signal characteristic value caused by the movement of the head expression and give the corresponding bioelectric characteristic signal, The head expression action is used to express the intention to confirm the current operation of the control component and its operation type;

The control module includes a posture fusion calculation module and a bioelectric signal calculation module, the posture fusion calculation module is used to receive the posture feature signal and identify the motion direction control instruction corresponding to the human body posture action, and the control module uses In order to further control the movement direction of the control part according to the movement direction control instruction; the bioelectric signal calculation module is used to receive the bioelectricity signal and identify the confirmation instruction corresponding to the head expression action, so The control module is used for confirming the current operation of the control component and the operation type thereof according to the confirmation instruction.

The second is a control component control method defined based on a two-dimensional intention, which is characterized in that it includes a human gesture action for expressing a direction control intention and a voice command for expressing a confirmation intention, and the direction control intention is to control the control the direction of movement of the component is controlled, and the confirmation is intended to confirm the current operation of the control component and its type of operation;

Implement the human body posture action, pick up the posture feature value of the action representation itself that characterizes the human body posture action, set the posture fusion calculation module, and identify the corresponding posture according to the picked up posture feature value and based on the posture fusion calculation module. The movement direction control command of the human body posture action, and the movement direction of the control component is controlled by the movement direction control command;

Implement a voice command, pick up the voice signal feature value that characterizes the voice command, set a voice signal calculation module, identify a confirmation corresponding to the voice command according to the picked up voice signal feature value and based on the voice signal calculation module an instruction, and the current operation of the control part and its operation type are confirmed through the confirmation instruction.

The present invention also provides a two-dimensional intention-based control device for realizing the second control method, which is characterized by comprising:

An attitude feature value pickup device, the attitude feature value pickup device is used to pick up the attitude feature value of the action representation itself that characterizes the posture action of the human body and give the corresponding posture feature signal, and the human body posture action is used to express the movement of the control part the control intention to control the direction;

Voice command feature value pickup device for picking up voice signal feature values characterizing the voice command and giving a corresponding voice feature signal, the voice command being used to express the current operation of the control part and its intent to identify the type of operation;

A control module, including a posture fusion calculation module and a voice signal calculation module, the posture fusion calculation module is used to receive the posture feature signal and identify the motion direction control instruction corresponding to the human body posture action, and the control module is used for The movement direction of the control component is further controlled according to the movement direction control instruction; the voice signal calculation module is used for receiving the voice feature signal and identifying the confirmation instruction corresponding to the voice command, and the control module is used for The current operation of the control part and its operation type are confirmed according to the confirmation instruction.

In the above solution, the control component refers to a virtual physical unit such as a game doll that exists in a real physical unit in a living environment or exists in a virtual environment such as a display screen, AI, VR, etc. , the control part can perform not only the movement action but also the confirmation operation under the driving of the control software of the control module.

In the above solution, the posture feature value pickup device and the head expression action or voice information feature value pickup device can be combined into a wearable device, and secondly, the posture feature value pickup device and the head expression The action or voice information feature value pick-up device and the control module can either be wired or communicated through a wireless transceiver. In addition, the control module may be integrated in the same control circuit, or may be separately disposed in a device including the gesture feature value pickup device, the head expression action or voice information feature value pickup device, and the like.

Based on the above technical solution, compared with the prior art, its beneficial technical effect is that by using the two-dimensional combination of the human body gesture action and the head expression action or voice command, not only a directional control adapted to the body action is established Intention command mechanism, and also established an adapted but independent "confirmation" intention command mechanism, which overcomes the problem of uncontrolled voice technology alone, and also overcomes the limitation of EOG technology alone. The effective cooperation of the device effectively solves the problem of multi-dimensional control of human-computer interaction equipment, as well as the coordination problem between human actions and control intentions.

Since the present invention has the above-mentioned features and advantages, it can be applied to human-computer interaction control components, equipment and control systems thereof.

Description of drawings

FIG. 1 is a schematic flowchart of a first embodiment of applying the technical solution of the present invention.

FIG. 2 is a schematic flowchart of a second embodiment of applying the technical solution of the present invention.

Detailed ways

The following further describes the control method and control device of the control component defined based on the two-dimensional intention applying the technical solution of the present invention with reference to the accompanying drawings.

Embodiment 1, as shown in FIG. 1, specifically, a control component control method defined based on a two-dimensional intention is characterized in that:

Including the human body gesture action for expressing the intention of direction control and the head expression action for expressing the intention of confirming the direction, the intention of the direction control is to control the movement direction of the control part 4, and the intention of confirming is to confirm the control the current operation of component 4 and its operation type;

Implement the human body posture action, pick up the posture feature value of the action representation itself that characterizes the human body posture action, set the posture fusion calculation module, and identify the corresponding posture according to the picked up posture feature value and based on the posture fusion calculation module. The movement direction control command of the human body posture action, and the movement direction of the control component 4 is controlled by the movement direction control command;

Implement the head expression action, pick up the head bioelectric signal characteristic value associated with the head expression movement, set up the bioelectric signal calculation module 23, according to the picked up bioelectric signal characteristic value and based on the The bioelectric signal calculation module 23 identifies the confirmation instruction corresponding to the head expression action, and confirms the current operation and the operation type of the control component 4 through the confirmation instruction.

The two-dimensional intention-based control device for realizing the above method is characterized by comprising:

The posture feature value pickup device 21 is used to pick up the posture feature value of the action representation itself that represents the posture action of the human body and give the corresponding posture feature signal, and the posture feature value of the human body is used to express the control component. 4. The control intention to control the movement direction;

A head bioelectric signal characteristic value pickup device 11, the head bioelectric signal characteristic value pickup device 11 is used to pick up the head bioelectric signal characteristic value caused by the movement of the head expression and give the corresponding bioelectric characteristic Signal, the head expression is used to express the intention to confirm the current operation of the control part 4 and its operation type;

The control module 2 includes a posture fusion calculation module 22 and a bioelectric signal calculation module 23. The posture fusion calculation module 22 is used to receive the posture feature signal and identify the motion direction control instruction corresponding to the posture action of the human body. The control module 2 is used to further control the movement direction of the control part 4 according to the movement direction control instruction; the bioelectric signal calculation module 23 is used to receive the bioelectrical characteristic signal and identify the The confirmation command of the facial expression, the control module 2 is used to confirm the current operation of the control part 4 and its operation type according to the confirmation command.

The control component is the hospital bed 4, and the present invention is further described by taking the control of the hospital bed 4 as an example.

A control device based on EOG (eye electricity) information and attitude, including a signal acquisition module, a wireless communication module, a control module 2 and a display screen 3, the wireless communication module includes a wireless sending unit and a wireless receiving unit, and the wireless sending unit is set in the signal The signal acquisition end of the acquisition module and the wireless receiving unit are arranged on the control algorithm end of the control module 2, and use the Wifi wireless communication protocol. The control module 2 is installed on a desktop computer.

A method of utilizing eye-control EOG technology, including human gesture actions such as head shaking actions for expressing directional control intentions, and eye actions leading to EOG (Eoculography) for expressing confirmation intentions Control the movement direction of the control part 4, and the confirmation intention is to confirm the current operation of the control part 4 and its operation type; the movements of shaking the head include shaking to the left, shaking to the right, raising the head and lowering the head Move four gestures.

Implement the action of shaking the head, pick up the attitude feature value of the action image itself that characterizes the shaking head action, including the acceleration, angle and movement direction feature value of the action, set the attitude fusion calculation module 22, according to the picked up attitude The characteristic value is based on the gesture fusion calculation module 22 to identify the motion direction control instruction corresponding to the shaking head action, and the motion direction of the control component 4 is controlled by the motion direction control instruction;

Implement the blinking expression action of the head, pick up the EOG (Eoculography) signal characteristic value associated with the head expression action, and set the bioelectric signal calculation module 23, according to the picked up The bioelectric signal characteristic value and based on The bioelectric signal calculation module 23 identifies the confirmation instruction corresponding to the head expression, and confirms the current operation and the operation type of the control component 4 through the confirmation instruction.

The signal acquisition module is a wearable device, and it further comprises a head bioelectrical signal feature value pickup device 11 for picking up the EOG electrical signal associated with the blinking of the eyes of the user 1, a device 11 for picking up the gesture action feature value of the user 1. Attitude feature value pickup device 21, the signals collected by them are transmitted to the control module 2 through the wireless communication module; the signal acquisition module also includes a microprocessor unit, and the microprocessor unit includes an STM32F103 chip, which is responsible for the entire signal Synchronization and control of each component of the acquisition module.

In this embodiment, the attitude feature value pickup device 21 is a wearable attitude sensor unit, which is worn on the head of the user 1, and includes a PU9250 sensor chip connected to the microprocessor unit. The chip is nine The axis attitude sensor is composed of a three-axis acceleration sensor, a three-axis gyroscope and a three-axis magnetometer, which respectively collect the acceleration, angle and movement direction characteristic values of the shaking action of the user 1's head.

In this embodiment, the head bioelectric signal feature value pickup device 11 is also a wearable ophthalmic device based on EOG signals, which is worn on the head of the user 1 and includes an electrode unit and an EOG signal amplifying unit. The electrode unit and the EOG signal amplifying unit are sequentially connected with the microprocessor unit. The electrode unit contains three conductive electrodes, which are close to different parts of the skin of the head of the user 1, one of which is placed on the forehead of the user 1, and the other two are placed behind the ears; the EOG signal amplifying unit is provided with an AD8232 integrated instrumentation amplifier. The chip integrates an instrumentation amplifier, a high-pass filter, a low-pass filter and a right leg drive circuit.

The control module 2 includes the gesture fusion calculation module 22, which is used to analyze the gestures of the head of the user 1 (for example, shaking to the left, shaking to the right, moving up, and bowing the head) according to the collected gesture signal of shaking the head. Move four posture actions) to carry out algorithm identification and give corresponding posture feature signals, determine the recognition result according to the posture feature signals, and further determine and drive the control component 4 to move in the corresponding direction, for example, when the head-up action is implemented The control part 4 is also lifted up to facilitate control. The control module 2 also includes a bioelectric signal calculation module 23, which is used to perform algorithmic identification of the eye movements of the user 1 according to the collected EOG electrical signals corresponding to the blinking action and provide corresponding bioelectric characteristic signals, The identification result is determined according to the bioelectric characteristic signal and a confirmation control instruction is further expressed to confirm the current operation type of the control part 4 (stop rising or stop falling). That is to say, the hospital bed 4 moves in response to the movement direction control instruction contained in the gesture action, and completes the "confirmation" of the current operation and its operation type in response to the confirmation instruction contained in the head expression action .

In the above solution, in addition to a head swaying action, the human body posture action can also be a leftward swaying, a rightward swaying, a chest-lifting movement, and a bending-over movement of the entire body; It is easy to distinguish obvious features, and it can unify the direction control intention with the posture direction of the human body itself, so as to facilitate expression and control, and its action representation is easy to recognize and its feature value is easy to extract. A control method that is more convenient for people to get used to is to try to make the result of the movement direction control basically consistent with the direction of these gesture actions. For example, when you raise your head or lift your chest, the control module 2 preferably controls the movement of the moving cursor. The direction also moves up instead of left or right.

However, these human body gestures are only expressions of control intentions, and are neither the signals themselves to control the control component 4 nor the control signals themselves to be expressed. Therefore, in order to accurately express the control intention, it is better to set the standardized and standardized actions of the human body posture in advance, so as to facilitate the accurate display and expression of the control intention, and prevent the control system from being difficult to identify and extract feature values. A further technical solution for this may also be to establish a human body posture specification model for head shaking or body shaking, and the human posture model defines the type of human action used to express the direction control intention, such as head shaking or body shaking. can be effective, and the speed of shaking, or the postures such as the maximum and minimum angles, are standardized and defined by the human body posture model, and the movement of the control component 4 is expressed by implementing the human body posture actions defined by the human body posture model. Orientation control intent. After the human body posture model is established, the user 1 can start to operate after understanding and learning the requirements of the standardized normative action before use. For example, when starting the initial state of the controlled control component 4, let the posture of the user 1 be adjusted to the set initial posture, including straightening the body and facing the display screen 3, etc. For another example, it is required that the angle and speed of the user 1 shaking the body conform to the included specifications and the like.

In the above solution, the gesture fusion calculation module 22 is a software operation module established based on experience and data, and is used to identify and interpret the control intention of the current human body gesture action. In use, the posture fusion calculation module 22 can also collect the eigenvalues based on a standard posture action, that is, the human body posture model, as the benchmark eigenvalues corresponding to the posture actions, and compare the current posture eigenvalues picked up. , identify the current gesture and action intention of user 1 in real time.

Wherein, the picking up the gesture feature value representing the human body gesture action defines that the starting point of the control is to pick up the feature value, and pick up the feature parameter as a three-dimensional behavior action, and provide it to the control module 2 for use later.

Wherein, the action representation of the gesture and action of the human body refers to the external performance characteristics of the gesture action in the three-dimensional space. To this end, the gesture feature value of the action image itself representing the gesture action of the human body is picked up, which is actually the trajectory left by the gesture action in time and space.

Wherein, the current operation of the control component 4 refers to a confirmation operation performed on the control component 4. A blinking action is similar to performing a left-click or right-click confirmation operation on the operation component on the screen through the mouse cursor. The operation type refers to the type of the confirmation operation, for example, the control part 4 is started or stopped, intentional or negative, or the control part 4 is accelerated or decelerated. These operation types can be based on The control of the different control components 4 requires setting adjustments.

In the above solution, it is also possible to replace the current posture feature value with the reference posture feature value after a current posture action is completed and confirmed to be valid. That is to say, in order to prevent the problem of recognition error and recognition delay caused by the drift of the reference posture feature value, the current posture feature value is replaced by the reference posture feature value at the end of each action or periodically, and the reference feature value is refreshed appropriately. set.

In the above solution, it can also be that the posture feature value includes the angle, the action direction and the action acceleration of the human action that can be identified from the action representation. That is, the posture feature value is defined from several aspects of action angle, action direction and action acceleration. For the extraction of the attitude feature value, in the invention, a wearable rate gyroscope detector is used to measure the angle of the action and an acceleration sensor is used to measure the acceleration of the action. When the gesture action moves, the movement direction feature value in the gesture feature value is provided. In other embodiments, a camera sampling device may also be used to capture images of the human body gestures and actions, and image analysis techniques may be used to obtain the gesture feature values of the human body gestures and actions.

In the above solution, the head expression action is a micro-muscle action or thinking action based on the head. For this purpose, in addition to using the EOG signal caused by the eye blinking action, other micro-muscles can also be used. Actions such as eye movements, facial twitches, jaw movements of clenching and grinding teeth, or pure cerebral conscious thinking movements; when these facial expressions are performed, corresponding bioelectric signals with different characteristics can be generated in the head, commonly known as EEG signal, EMG, etc. For example, EOG and EMG electromyography, there are many methods and devices to pick up these bioelectrical signals. The eigenvalues of these bioelectrical signals mainly include the amplitude of the electrical signal, the frequency of the signal and its spectral composition, etc. Picking and distinguishing these eigenvalues can realize the confirmation instruction.

Comparing the characteristics of the human body gestures and the head expressions, it can be found that the difference between them is obvious. The former belongs to the body movements with a relatively large amplitude, which can be used to directly express the control intention related to the movement direction. , while the latter is only a micro-muscle action or thinking action of the head, and the head expression action can be implemented completely synchronously when the human body gesture action is implemented, or the head expression action can be implemented in a rapid response before or after the human body gesture action is implemented. The above-mentioned head expressions and actions can realize quick confirmation control without interfering with each other, which is easy for ordinary users to operate.

Therefore, in the present invention, these facial expressions are used to express or transmit a confirmation instruction. Based on the confirmation intention and instruction, in the present invention, the control program itself can either arrange an action to represent a confirmation meaning, such as a traditional single The confirmation action of clicking the mouse can also be two consecutive actions representing a confirmation meaning, such as the confirmation action of the traditional double-clicking of the mouse, that is, the repetition or combination of one action or several actions.

Secondly, the head expression action is not only an independent action and its signal, but also defines that the head expression action is mainly used to express the intention of confirming the control, that is, it is not expected that the control component 4 such as the mouse cursor ( Also known as a specific marker) to move on its own terms, but to express the meaning of confirmation like start or stop confirmation of mouse click confirmation.

In the above solution, it is also possible to replace the current bioelectric signal feature value with the reference bioelectric signal feature value after a current expressive action is completed and confirmed to be valid. That is, in order to prevent user 1's bioelectrical signal characteristic changes in different periods or the current bioelectrical signal characteristic value changes of different participants resulting in identification errors (not the baseline bioelectrical signal characteristic value drift), use the end of each action or periodically. The current posture feature value is replaced with the reference bioelectric signal feature value, and the reference bioelectric signal feature value is appropriately refreshed and reset.

Embodiment 2, as shown in Figure 2, specifically, the present invention also provides a control method defined based on two-dimensional intention, which is characterized in that it includes a human body gesture action for expressing direction control intention and a method for expressing confirmation. The voice command of the intention, the direction control intention is to control the movement direction of the control part 4, and the confirmation intention is to confirm the current operation of the control part 4 and its operation type;

Carry out the described human body posture action, pick up the posture feature value of the action representation itself that characterizes the described human body posture action, set the posture fusion calculation module 22, and identify the posture fusion calculation module 22 according to the posture feature value picked up and based on the posture fusion calculation module 22. Corresponding to the motion direction control instruction of the gesture action of the human body, the motion direction of the control component 4 is controlled by the motion direction control instruction;

Implement a voice command, pick up the voice signal characteristic value that characterizes the voice command, set the voice signal calculation module 24, and identify the voice signal corresponding to the voice command according to the voice signal characteristic value picked up and based on the voice signal calculation module 24 The confirmation command is used to confirm the current operation of the control component 4 and its operation type through the confirmation command.

Comparing the solution of the second embodiment above with that of the first embodiment, the main difference is that the head expression action for expressing the confirmation intention in the first embodiment is changed to a voice command issued by the user 1. Therefore, in the above solution, also It may be that the current voice command intention of the user 1 is recognized in real time by establishing the voice signal reference feature value corresponding to the voice command and comparing the picked-up feature value of the current voice signal through the voice signal calculation module.

In the above solution, it can also be that, after a current voice signal ends and is confirmed to be valid, the voice signal calculation module replaces the current voice signal feature value with a reference feature value. In this way, when the user 1 of the product is replaced or the accent characteristic of the user 1 is changed, the recognition ability can still be adjusted in time.

The present invention also provides a two-dimensional intention-based control device, characterized in that it includes:

The posture feature value pickup device 21 is used to pick up the posture feature value of the action representation itself that represents the posture action of the human body and give the corresponding posture feature signal, and the posture feature value of the human body is used to express the control component. 4. The control intention to control the movement direction;

Voice command feature value pickup device 5, described voice command feature value pickup device 5 is used to pick up the voice signal feature value characterizing the voice command and give the corresponding voice feature signal, and the voice command is used to express the control unit 4 the intent of the current operation and its type of operation to be confirmed;

The control module 2 includes a gesture fusion calculation module 22 and a speech signal calculation module 24, and the gesture fusion calculation module 22 is used to receive the gesture feature signal and identify the motion direction control instruction corresponding to the gesture action of the human body. The control module 2 is used to further control the movement direction of the control component 4 according to the movement direction control instruction; the voice signal calculation module 24 is used to receive the voice feature signal and identify the confirmation instruction corresponding to the voice command , the control module 2 is used to confirm the current operation of the control component 4 and its operation type according to the confirmation instruction.

Claims (27)

1. A control component control method defined based on a two-dimensional intention, characterized in that it includes a human body gesture action for expressing a direction control intention and a head expression action for expressing a confirmation intention, and the direction control intention is intended to control the control. the direction of movement of the component is controlled, and the confirmation is intended to confirm the current operation of the control component and its type of operation; Implement the human body posture action, pick up the posture feature value of the action representation itself that characterizes the human body posture action, set the posture fusion calculation module, and identify the corresponding posture according to the picked up posture feature value and based on the posture fusion calculation module. The movement direction control command of the human body posture action, and the movement direction of the control component is controlled by the movement direction control command; Implement the head expression action, pick up the bioelectric signal characteristic value of the head caused by the head expression action, set up a bioelectric signal calculation module, according to the picked up bioelectric signal characteristic value and based on the The bioelectric signal calculation module identifies a confirmation instruction corresponding to the head expression, and confirms the current operation of the control component and its operation type through the confirmation instruction. 2. control method according to claim 1, is characterized in that, establishes the reference characteristic value corresponding to the gesture action and compares the current attitude characteristic value picked up by the described attitude fusion calculation module, recognizes the user's current attitude in real time Action intent. 3 . The control method according to claim 2 , wherein after a current gesture action is completed and confirmed to be valid, the current gesture feature value is replaced with a reference feature value. 4 . 4 . The control method according to claim 2 , wherein the posture feature value includes the angle, direction and acceleration of the human action which can be identified from the action representation. 5 . 5. The control method according to claim 4, wherein a human body posture model is established, and the human body posture model defines a human body action type for expressing the direction control intention, and is defined in accordance with the human body posture model by implementing It expresses the control intention of the movement direction of the control part. 6. The control method according to claim 1, wherein the bioelectric signal calculation module is used to establish a reference characteristic value of the bioelectric signal corresponding to the facial expression and compare the characteristic value of the current bioelectric signal picked up, Real-time recognition of the user's current expression and action intention. 7 . The control method according to claim 6 , wherein after a current facial expression is completed and confirmed to be valid, the current bioelectric signal feature value is replaced with a reference posture feature value. 8 . 8 . The control method according to claim 6 , wherein the characteristic value of the bioelectric signal includes a signal amplitude and a signal frequency. 9 . 9 . The control method according to claim 1 , wherein the operation type includes whether to start or stop, or whether to accelerate or decelerate. 10 . 10. The control method according to any one of claims 1 to 9, wherein the human body gesture actions include four kinds of gesture actions capable of independently expressing control intentions of movement directions, and each of the gesture actions defines a movement direction Control intent, the four gesture actions are used to express the control intent of moving in four directions: front, back, left, and right, respectively. 11 . The control method according to claim 10 , wherein the gesture action of the human body is a shaking action of the body. 12 . 12 . The control method according to claim 11 , wherein the body shaking action includes four posture actions: leftward shaking of the body, rightward shaking, chest up movement and bending over movement. 13 . 13 . The control method according to claim 10 , wherein the gesture action of the human body is a shaking action of the head. 14 . 14 . The control method according to claim 13 , wherein the shaking action of the head includes four gesture actions of the head tilting to the left, tilting to the right, moving the head up and moving the head down. 15 . 15. The control method according to any one of claims 1 to 9, wherein the head expression action is a head muscle action or a thinking action that can cause the head to generate bioelectric signals that can be picked up. 16 . The control method according to claim 15 , wherein the head expression action comprises one action or a repetition or combination of several actions among eye action, facial action, jaw action or brain awareness action. 17 . 17. The control method according to any one of claims 1 to 9, wherein the control component is a hospital bed or a wheelchair, and the control component moves in response to a movement direction control command contained in the gesture action, and The "confirmation" of the current operation and its operation type is completed in response to the confirmation instruction implied by the head expression action. 18. A control part control method defined based on a two-dimensional intention, characterized in that it comprises a human body gesture action for expressing a direction control intention and a voice command for expressing a confirmation intention, the direction control intention being aimed at controlling the control part. The direction of movement is controlled, and said confirmation is intended to confirm the current operation of said control part and its type of operation; Implement the human body posture action, pick up the posture feature value of the action representation itself that characterizes the human body posture action, set the posture fusion calculation module, and identify the corresponding posture according to the picked up posture feature value and based on the posture fusion calculation module. The movement direction control command of the human body posture action, and the movement direction of the control component is controlled by the movement direction control command; Implement a voice command, pick up the voice signal feature value that characterizes the voice command, set a voice signal calculation module, identify a confirmation corresponding to the voice command according to the picked up voice signal feature value and based on the voice signal calculation module an instruction, and the current operation of the control part and its operation type are confirmed through the confirmation instruction. 19. The control method according to claim 18, characterized in that, establishing the reference feature value of the voice signal corresponding to the voice command by the voice signal calculation module and comparing the feature value of the current voice signal picked up, identifying in real time The user's current voice command intent. 20. The control method according to claim 19, characterized in that, after a current voice signal ends and is confirmed to be valid, the voice signal calculation module replaces the current voice signal feature value with a reference voice signal feature value. 21. A two-dimensional intent-based control device, characterized in that it comprises: An attitude feature value pickup device, the attitude feature value pickup device is used to pick up the attitude feature value of the action representation itself that characterizes the posture action of the human body and give the corresponding posture feature signal, and the human body posture action is used to express the movement of the control part the control intention to control the direction; A head bioelectric signal characteristic value pickup device, the head bioelectric signal characteristic value pickup device is used to pick up the head bioelectric signal characteristic value caused by the movement of the head expression and give the corresponding bioelectric characteristic signal, The head expression action is used to express the intention to confirm the current operation of the control component and its operation type; The control module includes a posture fusion calculation module and a bioelectric signal calculation module, the posture fusion calculation module is used to receive the posture feature signal and identify the motion direction control instruction corresponding to the human body posture action, and the control module uses In order to further control the movement direction of the control part according to the movement direction control instruction; the bioelectric signal calculation module is used to receive the bioelectricity signal and identify the confirmation instruction corresponding to the head expression action, so The control module is used for confirming the current operation of the control component and the operation type thereof according to the confirmation instruction. 22. The control device according to claim 21, wherein the control part is a hospital bed or a wheelchair, and the control part moves in response to a movement direction control command contained in the gesture action, and in response to the The confirmation command contained in the head expression action completes the "confirmation" of the current operation and its operation type. 23 . The control device according to claim 21 , wherein the attitude feature value pickup device comprises a rate gyroscope detector and an acceleration sensor, and the rate gyroscope detector is used to provide a rate gyroscope detector when moving with the human body attitude action. 24 . The angular velocity change characteristic value in the attitude characteristic value; the acceleration sensor is used to provide the acceleration change characteristic value in the attitude characteristic value when moving with the gesture action of the human body. 24 . The control device according to claim 23 , wherein the device for picking up the posture characteristic value further comprises a magnetometer, which is used to provide the posture characteristic value when moving with the posture action of the human body. 25 . Motion direction eigenvalues. 25 . The control device according to claim 21 , wherein the attitude feature value pickup device is a camera sampling device, and the camera sampling device is used to capture images of the human body gestures and use image analysis technology to obtain the obtained images. 26 . The posture feature value of the human body posture action. 26. The control device according to any one of claims 21 to 25, wherein the head bioelectric signal feature value pickup device is a wearable device based on an EOG signal, comprising an electrode unit and an EOG signal amplifying unit; The electrode unit contains three conductive electrodes, which are close to the skin of the user's head, one of which is placed on the user's forehead, and the other two are placed behind the ear; the EOG signal amplifying unit is provided with an integrated instrumentation amplifier, which is integrated on the integrated instrumentation amplifier. There are instrumentation amplifier, high pass filter, low pass filter and right leg drive circuit. 27. A two-dimensional intent-based control device, comprising: An attitude feature value pickup device, the attitude feature value pickup device is used to pick up the attitude feature value of the action representation itself that characterizes the posture action of the human body and give the corresponding posture feature signal, and the human body posture action is used to express the movement of the control part the control intention to control the direction; Voice command feature value pickup device for picking up voice signal feature values characterizing the voice command and giving a corresponding voice feature signal, the voice command being used to express the current operation of the control part and its intent to identify the type of operation; A control module, including a posture fusion calculation module and a voice signal calculation module, the posture fusion calculation module is used to receive the posture feature signal and identify the motion direction control instruction corresponding to the human body posture action, and the control module is used for The movement direction of the control component is further controlled according to the movement direction control instruction; the voice signal calculation module is used for receiving the voice feature signal and identifying the confirmation instruction corresponding to the voice command, and the control module is used for The current operation of the control part and its operation type are confirmed according to the confirmation instruction.

Images