Physiotherapy robot control method and physiotherapy robot with replaceable physiotherapy head

Abstract

The application relates to a physiotherapy robot control method and a physiotherapy robot with replaceable physiotherapy heads. The physiotherapy robot comprises a main machine and a plurality of different physiotherapy heads. The main machine and the physiotherapy heads are detachably mechanically and electrically connected. The method comprises the following steps: identifying the type of the physiotherapy head; confirming the mechanical structure and the center of gravity of the physiotherapy head based on the type of the physiotherapy head; and adjusting the force control parameters of the mechanical arm based on the mechanical structure and the center of gravity of the physiotherapy head. The application improves the safety of the physiotherapy robot with replaceable physiotherapy heads.

Description

Technical Field

[0001] This invention relates to the field of physiotherapy robot control, specifically to a physiotherapy robot control method with a replaceable physiotherapy head and a physiotherapy robot. Background Technology

[0002] When a physiotherapy robot provides physiotherapy to a customer, the physiotherapy function depends on the physiotherapy principle of the physiotherapy head itself, such as RF radio frequency physiotherapy, EMS microcurrent physiotherapy, Bian stone physiotherapy, shock wave physiotherapy, fascia gun physiotherapy, negative pressure physiotherapy, red light physiotherapy, microwave physiotherapy, moxibustion physiotherapy, etc.

[0003] In traditional physiotherapy devices, each device is fixedly connected to a specific type of treatment head. Because different treatment heads have different centers of gravity, different torques are generated when changing treatment heads, resulting in an unreasonable distance between the treatment head and the skin, which poses a significant safety risk during physiotherapy. Summary of the Invention

[0004] This invention addresses the control problem after the replacement of the physiotherapy head by providing a control method for a physiotherapy robot with a replaceable physiotherapy head and a physiotherapy robot in general.

[0005] Firstly, the replaceable therapy head control method for a physiotherapy robot provided in this application is applied to a physiotherapy robot, which includes a main unit and multiple different therapy heads. The main unit and the therapy heads are detachably mechanically and electrically connected. The replaceable therapy head control method for the physiotherapy robot includes:

[0006] Identify the type of therapy head;

[0007] Based on the type of the therapy head, confirm the mechanical structure and center of gravity of the therapy head;

[0008] Based on the mechanical structure and center of gravity of the therapy head, the force control parameters of the robotic arm are adjusted.

[0009] In some embodiments, the physiotherapy robot control method with a replaceable physiotherapy head further includes:

[0010] Based on the type of treatment head, adjust the retraction parameters of the robotic arm to adjust the minimum distance between the treatment head and the skin during operation.

[0011] In some embodiments, the backoff parameter is used to perform backoff processing on point cloud data generated based on facial images during physiotherapy to obtain backoff point cloud data, so that the distance between the backoff point cloud data and the target facial area is greater than a preset safe distance.

[0012] In some embodiments, the physiotherapy robot control method with a replaceable physiotherapy head further includes:

[0013] Acquire facial images in real time from the 3D camera of the physiotherapy robot;

[0014] Point cloud data is generated based on the facial image;

[0015] The point cloud data is superimposed with a unit vector of a preset multiple to obtain backdated point cloud data, wherein the distance between the backdated point cloud data and the target facial region is greater than a preset safe distance;

[0016] Generate target trajectory data based on the reverted point cloud data;

[0017] Based on the target trajectory data, the physiotherapy head of the physiotherapy robot is controlled to move in at least three degrees of freedom to perform physiotherapy on the target facial area.

[0018] In some embodiments, the type of the therapy head includes one or more combinations of red light therapy, microwave therapy, and moxibustion therapy; wherein...

[0019] Different treatment heads have different retraction parameters.

[0020] In some embodiments, generating target trajectory data based on backtracked point cloud data includes:

[0021] Calculate the normal vector of the local plane corresponding to the backed-up point cloud data;

[0022] The target trajectory data is calculated based on the normal vector.

[0023] In some embodiments, the target trajectory data includes a first deflection angle, a second deflection angle, and a third deflection angle; wherein the axes corresponding to the first deflection angle, the second deflection angle, and the third deflection angle are mutually perpendicular; the method includes:

[0024] Based on satisfying the first deflection angle constraint, the second deflection angle constraint, and the third deflection angle constraint, the corresponding target trajectory data is calculated and generated.

[0025] The first deflection angle constraint conditions include:

[0026] First formula:

[0027] And, the second formula:

[0028]

[0029] Wherein, the normal vector of the local plane RX represents the first deflection angle;

[0030] The second deflection angle constraint includes:

[0031] Third formula:

[0032] And, the fourth formula:

[0033]

[0034] Where RY represents the second deflection angle;

[0035] The third deflection angle constraint condition is: the third deflection angle is equal to 180°.

[0036] In some embodiments, the physiotherapy robot control method with replaceable physiotherapy heads further includes:

[0037] Based on the treatment plan selected by the user, the regression parameters are confirmed, wherein the regression for facial treatment is greater than the regression parameter for back treatment.

[0038] In some embodiments, the host computer is provided with a main control board, and the therapy head is provided with an auxiliary control board. Identifying the type of therapy head includes:

[0039] The auxiliary control board encrypts the random code based on a preset algorithm and sends it to the main control board;

[0040] The main control board decrypts the random code based on the preset algorithm and then confirms the type of the physiotherapy head.

[0041] Secondly, this application provides a physiotherapy robot, comprising:

[0042] At least two different treatment heads;

[0043] The host and the physiotherapy head are detachably mechanically and electrically connected. The host is used to execute the physiotherapy robot control program to realize the above-mentioned physiotherapy robot control method with replaceable physiotherapy head.

[0044] This application embodiment identifies the type of physiotherapy head, and based on that type, confirms the mechanical structure and center of gravity to determine the force control parameters of the physiotherapy head. This setting ensures that when changing the physiotherapy head, the force control parameters will make the distance between the physiotherapy head and the skin reasonable, thereby improving the safety of the physiotherapy robot. Attached Figure Description

[0045] Figure 1 This is a flowchart of an embodiment of the control method for a physiotherapy robot with a replaceable physiotherapy head according to this application;

[0046] Figure 2 This is a flowchart of another embodiment of the control method for a physiotherapy robot with a replaceable physiotherapy head according to this application;

[0047] Figure 3 This is a schematic diagram of the local plane normal vector of an embodiment of the control method for a replaceable physiotherapy head physiotherapy robot according to this application. Detailed Implementation

[0048] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0050] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0051] Example 1

[0052] Reference Figure 1 Firstly, the replaceable therapy head control method for a physiotherapy robot provided in this application is applied to a physiotherapy robot, which includes a main unit and multiple different therapy heads. The main unit and the therapy heads are detachably mechanically and electrically connected. The replaceable therapy head control method for the physiotherapy robot includes:

[0053] S100, Identify the type of therapy head;

[0054] In this embodiment, the type of therapy head can be identified using sensors. For example, different numbers and polarities of magnets can be placed on the therapy head, such as one magnet with its N pole facing the main unit; one magnet with its S pole facing the main unit; two magnets, both with their N poles facing the main unit; two magnets, both with their S poles facing the main unit; or two magnets, one with its N pole facing the main unit and the other with its S pole facing the main unit. Then, a Hall sensor can be placed on the main unit to detect the type of therapy head.

[0055] In some embodiments, the type of physiotherapy head can be identified by the host communicating with the physiotherapy head after the physiotherapy head is installed, thereby confirming the type of the physiotherapy head. At this time, it is only necessary to assign different random sequence codes to different physiotherapy heads. Based on the correspondence between the sequence codes and the physiotherapy head types, the type of physiotherapy head can be confirmed.

[0056] S200. Based on the type of the therapy head, confirm the mechanical structure and center of gravity of the therapy head;

[0057] It should be noted that considering only the center of gravity will cause the center of gravity to change during six-axis motion, ultimately leading to control logic errors.

[0058] S300. Based on the mechanical structure and center of gravity of the therapy head, adjust the force control parameters of the robotic arm.

[0059] In this embodiment, the mechanical structure and center of gravity of the physiotherapy head can be calibrated first to obtain the corresponding force control parameters. The specific calibration method is the same as that of the traditional single physiotherapy head, and will not be traced here. The calibration is completed at the factory, and each type of physiotherapy head corresponds to a set of force control parameters.

[0060] Furthermore, due to errors in the manufacturing process of the therapy head, there may be a deviation between the current center of gravity of the therapy head and the center of gravity of the therapy head used during calibration. In this embodiment, after determining the force control parameters based on the type of therapy head, the following steps are also included:

[0061] A preset command is output to the robotic arm, and the difference between the robotic arm's motion and the expected motion corresponding to the preset command is detected. Based on this difference, the force control parameters are adjusted. Furthermore, the difference between the adjusted force control parameters and the original force control parameters is less than a preset threshold; in other words, the adjustable range of each force control parameter is constrained to avoid over-adjustment. The specific constraint range can be set according to the actual application.

[0062] This application embodiment identifies the type of physiotherapy head, and based on that type, confirms the mechanical structure and center of gravity to determine the force control parameters of the physiotherapy head. This setting ensures that when changing the physiotherapy head, the force control parameters will make the distance between the physiotherapy head and the skin reasonable, thereby improving the safety of the physiotherapy robot.

[0063] Reference Figure 1 In some embodiments, the physiotherapy robot control method with replaceable physiotherapy heads further includes:

[0064] S400: Based on the type of treatment head, adjust the retraction parameters of the robotic arm to adjust the minimum distance between the treatment head and the skin during operation.

[0065] Reference Figure 2In some embodiments, the backoff parameter is used to perform backoff processing on the point cloud data generated based on the facial image during the physiotherapy process to obtain backoff point cloud data, so that the distance between the backoff point cloud data and the target facial area is greater than a preset safe distance.

[0066] Specifically, the physiotherapy robot control method with replaceable physiotherapy head of this application also includes:

[0067] S500: Acquires facial images in real time from the 3D camera of the physiotherapy robot;

[0068] In this embodiment, the 3D camera can use speckle laser measurement to acquire 3D images of the target face in real time through a speckle laser and an infrared sensor.

[0069] S600. Generate point cloud data based on the facial image; wherein, the conversion between the real-time facial image and the point cloud data is a conventional technical means, which will not be elaborated here.

[0070] S700. Perform distance constraint processing on the point cloud data to obtain backdated point cloud data, wherein the distance between the backdated point cloud data and the target facial region is greater than a preset safe distance.

[0071] For example, the generated point cloud image is theoretically aligned with the human face, but the therapeutic head of the physiotherapy robot generally needs to maintain a distance from the human face during operation; at the same time, the position of the human face inevitably changes during the physiotherapy process, causing the therapeutic head to come into contact with the human body.

[0072] This embodiment performs distance constraint processing on the point cloud data, ensuring that the distance between the retrieved point cloud data and the target facial region is greater than a preset safe distance. Specifically, the distance between the retrieved point cloud data and the target facial region can refer to the distance between the point cloud data of the facial region and the corresponding data points in the retrieved point cloud data.

[0073] S800: Generate target trajectory data based on the reverted point cloud data;

[0074] The specific calculation method for generating target trajectory data based on the backtracked point cloud data is not limited here.

[0075] S900: Based on the target trajectory data, control the physiotherapy robot's physiotherapy head to move in at least three degrees of freedom to perform physiotherapy on the target facial area.

[0076] The physiotherapy robot can achieve 6-axis motion, such as three-axis translation and three-axis rotation. In this embodiment, after hand-eye calibration, tool calibration, binocular calibration, and optimized calibration, the physiotherapy robot can acquire matching data from a 3D camera. Pixels acquired from an RGB camera... and matching 3D camera 3D data After dual-target positioning This allows you to obtain the 3D data of the matched target points. The target's 3D data, based on the 3D camera coordinate system, is calibrated by hand and eye. This allows us to obtain the representation of the target's 3D data in the tool coordinates of the robotic arm. After tool calibration After conversion, the target's 3D data can be output based on the robotic arm's base coordinate system. After subsequent optimization and calibration This can make the output value closer to the true value. Its expression is as follows:

[0077]

[0078] This application generates point cloud data by acquiring real-time facial images from a 3D camera of a physiotherapy robot. Then, distance constraint processing is applied to the point cloud data to obtain regression point cloud data, where the distance between the regression point cloud data and the target facial region is greater than a preset safety distance. Finally, target trajectory data is generated based on the regression point cloud data, and the physiotherapy head of the robot is controlled to move in at least three degrees of freedom to perform physiotherapy on the target facial region. By generating point cloud data from real-time acquired facial images, the point cloud data is matched to the human face. Distance constraint processing of the point cloud data ensures that the generated target trajectory data allows the physiotherapy head to maintain a safe distance from the face during operation.

[0079] Reference Figure 3 In one embodiment, step S700, which performs distance constraint processing on the point cloud data to make the distance between the point cloud data and the target facial region greater than a preset safe distance, includes: superimposing a unit vector of a preset multiple on the point cloud data to obtain backtracking point cloud data.

[0080] In this embodiment, based on the unit vector i of the point cloud The point is then vector-backed.

[0081]

[0082] Assuming a safe distance of H, the point cloud points After vector regression, maintaining a safe distance from the facial area, the point cloud data is regressed. It can be represented as:

[0083]

[0084] It's important to note that maintaining one posture for extended periods can cause muscle soreness. Therefore, during physical therapy, the position of the face will inevitably change. However, after this change, the face will quickly return to roughly the same position, such as when the neck is turned. Because the face returns to its original position quickly, the physical therapy robot does not need to adjust itself in real time to follow the user's facial changes. It only needs to maintain its original function within a safe distance.

[0085] In one embodiment, to address this scenario, the method further includes:

[0086] The target trajectory data for the duration of caching includes target trajectory data at a first moment and target trajectory data at a second moment after the first moment;

[0087] Based on the target trajectory data at the second time moment, the target trajectory data at the first time moment is smoothed to make the target trajectory data at the first time moment smoother and to satisfy the distance constraint.

[0088] In this embodiment, each time target trajectory data is generated, it can be cached in the cache of the physiotherapy robot, so that the target trajectory data in the cache is a continuous trajectory data for a period of time, and the "period of time" can be 0.5 seconds to 3 seconds.

[0089] Based on the target trajectory data at the second time moment, the target trajectory data at the first time moment is smoothed, which can be understood as:

[0090] (1) The first moment refers to a point in time, and the second moment is a cached multiple points in time after the first moment; in other embodiments, the first moment may also refer to multiple consecutive points in time.

[0091] (2) Based on the target trajectory data of multiple time points cached after the first moment, that is, the target trajectory data of the second moment, confirm whether the next target trajectory data corresponds to the human face changing position and the human face quickly returning to the same position as before.

[0092] (3) If so, keep the target trajectory data at the first moment roughly as it is; otherwise, let the target trajectory data at the first moment smoothly follow the target trajectory data at the second moment.

[0093] Then, the trajectory data at the first moment is smoothed. (3) The whole process can be called smoothing.

[0094] (4) Use the target trajectory data at the first moment to control the operation of the physiotherapy robot;

[0095] This process continues, ensuring that each initial moment fully considers whether the human face will quickly return to a roughly the same position after a change in position in the second moment. Therefore, the therapy robot, operating based on the trajectory data from the first moment, will not follow sudden head movements of the user.

[0096] It should be noted that while "the human face will quickly return to a position roughly the same as before," it is not exactly the same. Therefore, when controlling the trajectory at the first moment, it is still necessary to ensure that the trajectory data at the first moment maintains a safe distance from the human face. Thus, after confirming the target trajectory data at the first moment through the second moment, it is necessary to verify whether the target trajectory at the first moment meets the distance constraint. If not, the target trajectory data composed of the first and second moments needs to be smoothed multiple times until the target trajectory data at the first moment meets the distance constraint.

[0097] In one embodiment, the target trajectory data at the first time moment is smoothed based on the target trajectory data at the second time moment to make the target trajectory data at the first time moment smoother and satisfy the distance constraint, including:

[0098] Based on the preset physiotherapy needs, generate the objective function corresponding to each time period;

[0099] Using the minimization of the objective function as the convergence condition, the target trajectory data at the first moment is iteratively smoothed until convergence, so that the target trajectory data at the first moment is smoother and satisfies the distance constraint.

[0100] It should be noted that during the iterative smoothing process, the trajectory path may change, and different physiotherapy needs require different changes in the path. In this embodiment, minimizing the objective function makes the motion path of the physiotherapy head more suitable for the actual physiotherapy needs. The objective function can be a similarity function between the path corresponding to the target trajectory data and the path preset based on the physiotherapy needs.

[0101] In this embodiment, iteratively smoothing the target trajectory data at the first moment until convergence may include:

[0102] (1) An update step that updates the current estimate in a way that reduces the value of the objective function without considering any constraints;

[0103] (2) Based on the trajectory data obtained in (1) at the first moment, calculate whether the safe distance constraint is met. If not, adjust the trajectory data at the first moment until the distance constraint is just met. Then repeat step (1) until the trajectory data at the first moment output in step (1) simultaneously meets the requirements of the safe distance constraint and the objective function minimization.

[0104] In one embodiment, the objective function is a similarity function between the path corresponding to the target trajectory data and the path preset based on the physiotherapy needs.

[0105] In one embodiment, the above control method further includes:

[0106] Calculate the normal vector of the local plane corresponding to the backed-up point cloud data;

[0107] It should be noted that when the sampling surface of the point cloud is smooth, the local neighborhood of any point can be fitted with a plane. The embodiment uses a local plane fitting method for normal vector estimation. For each scanned point p in the point cloud data, its K nearest neighbors are searched. The scanned point p and its K nearest neighbors constitute a local plane P. The normal vector n of the fitted local plane P can be the desired normal vector n. In this case, the normal vector n of the local plane P in the least squares sense can be expressed as:

[0108]

[0109] Where argminf(x) represents the value of x when f(x) reaches its minimum. (Principal components analysis)

[0110] The normal vector n of the local plane P can be obtained using the principal component analysis (PCA) algorithm. The centroid of point K passes through the domain point. normal vector satisfy For the covariance matrix Feature analysis shows that the eigenvector corresponding to the smallest eigenvalue is the normal vector of the local plane P.

[0111]

[0112] The target trajectory data is calculated based on the normal vector.

[0113] The target trajectory data includes a first deflection angle RX, a second deflection angle RY, and a third deflection angle RZ, wherein the axes corresponding to the first deflection angle RX, the second deflection angle RY, and the third deflection angle RZ are perpendicular to each other.

[0114] In this embodiment, the step of calculating the target trajectory data based on the normal vector includes:

[0115] Based on the constraints of the first deflection angle RX, the second deflection angle RY, and the third deflection angle RZ, the corresponding trajectory data to be processed is calculated and generated. It should be noted that the constraints can also be understood as the output model, that is, after the normal vector input value is output as the model (constraints), the corresponding deflection angle can be calculated.

[0116] For example, the first deflection angle RX constraint includes:

[0117] First formula:

[0118] And, the second formula:

[0119]

[0120] Wherein, the normal vector of the local plane RX represents the first deflection angle RX; This represents the unit normal vector value along the X-axis; α represents the normal vector's position relative to the x-axis. The Z-axis angle; W represents the normal vector value in the Z-axis direction. It should be noted that after system calibration, the backtracking point cloud data acquired by the 3D camera can be transformed into the robotic arm's output coordinate system, allowing the robotic arm to move to a specified target position. Controlling the robotic arm to the target pose also requires related attitude control; the robotic arm's attitude output is determined by the deflection angle. Control can be achieved by its unit normal vector. Figure it out.

[0121] Attitude output deflection angle To surround The axis rotates at an angle, and the end effector of the robotic arm rotates around... Axis rotation angle Space. When the robotic arm is working, the tool tip should be facing the target object. Based on the robot's structural layout, with the tool tip pointing downwards, the unit normal vector of the object... middle . Controlling the robotic arm to swing left and right, based on the spatial relationship between the target being measured and the tool's end effector, then... Its angle of motion in the direction should be within Within the interval. When the end effector of the robotic arm is deflected to the left (e.g., rotated to a 135° position), then its normal vector at this time... middle , The value should be in First deflection angle The output model is shown in the first formula.

[0122] When the end effector of the robotic arm is deflected to the right (e.g., rotated to -135°), its normal vector is... middle , The value should be in , The output model exists as shown in the second formula.

[0123] For example, the second deflection angle RY constraint includes:

[0124] Third formula:

[0125] And, the fourth formula:

[0126]

[0127] Where RY represents the second deflection angle RY; Represents the unit normal vector value along the Y-axis; Representing the normal vector and Angle between axes;

[0128] Robotic arm posture output deflection angle middle Indicated as surrounding The rotation angle of the axis; the rotation angle of the robotic arm end about the X-axis. Space. When the robotic arm is working, the tool end of the robotic arm faces downwards, and the unit normal vector of the same object... middle . Controlling the robotic arm to swing back and forth, based on the spatial relationship between the target being measured and the tool's end effector, then... Its angle of motion in the direction should be within Within the interval. When the end effector of the robotic arm deflects forward (e.g., rotates to a 45° position), then its normal vector at this time... , The value should be in The following relationship exists:

[0129]

[0130] When the end effector of the robotic arm is tilted backward (e.g., rotated to -45°), then its normal vector is... , The value should be in There exists a fourth formula.

[0131] Rotation of the tool end of the robotic arm It can achieve the target posture, and From the angle, The angle can be set to a fixed value according to requirements.

[0132] For example, the constraint condition for the third deflection angle RZ is: the third deflection angle RZ is equal to 180°. This is because when the robotic arm is working, its tool tip always points downwards, and after analysis, its parameter value can be set as follows: .

[0133] In summary, the output from the normal vector to the robot's pose has been completed.

[0134] In one embodiment, the time span of the target trajectory data for the duration of the cache can be used to identify unexpected movements. For example, 0.5 seconds to 3 seconds.

[0135] In one embodiment, the number of 3D cameras is multiple, and they are arranged to capture facial images from multiple angles; S500 includes: acquiring multi-angle facial images captured by the multiple 3D cameras of the physiotherapy robot.

[0136] In this embodiment, since the areas of the human face requiring treatment are mainly divided into three parts—the forehead, left cheek, and right cheek—it is difficult to obtain full profile data using a normal frontal view photography method. Therefore, it is necessary to add a corresponding shooting angle to obtain three-dimensional information of the profile while ensuring the identification of key facial points.

[0137] Therefore, this embodiment uses multiple 3D cameras, which are arranged to capture facial images from multiple angles.

[0138] In some embodiments, the point cloud data can also be smoothed and denoised to obtain backdated point cloud data;

[0139] It should be noted that the real-time facial images captured by the 3D camera contain non-target areas and noise interference. In order to improve the quality of the point cloud data,

[0140] In this embodiment, point cloud segmentation uses a pass-through filtering method, which sets a value range based on the extreme value and structural range of a certain dimension to generate new point cloud data;

[0141] In this embodiment, Gaussian filtering is used for smoothing and denoising the point cloud data. Statistical analysis is applied to nearby point clouds to determine if they are noise. A KD_Tree neighborhood search is performed to find N neighborhood points, and the average Euclidean distance from the sampling point to the neighborhood points is calculated.

[0142] Initial point cloud coordinates Sampling domain search point set The average Euclidean distance from the sampling point to the neighborhood point Distance can be expressed as:

[0143]

[0144] Initial point cloud The average Euclidean distance between it and its K neighborhood points mean and standard deviation They can be represented as:

[0145]

[0146]

[0147] The average Euclidean distance from the sampling points to the neighborhood points in the 3D point cloud acquired by the 3D camera should form a Gaussian distribution. By setting its mean and variance and replacing the original data with the calculated Gaussian smoothed data, the effects of noise reduction and data smoothing of the point cloud can be achieved.

[0148] In some embodiments, the physiotherapy robot control method with replaceable physiotherapy heads further includes:

[0149] Based on the treatment plan selected by the user, the regression parameters are confirmed, wherein the regression for facial treatment is greater than the regression parameter for back treatment.

[0150] The backoff parameter can refer to a unit vector of the aforementioned preset multiple. In practical applications, one backoff parameter is sufficient for each treatment head. However, in practical applications, taking moxibustion as an example, the safe distance between the treatment head and the skin varies for different temperatures. Therefore, the safe distance can be determined based on the treatment plan, and the backoff parameter can be calculated to ensure compatibility with different treatment plans using the same treatment head.

[0151] In some embodiments, the host computer is provided with a main control board, and the therapy head is provided with an auxiliary control board. Step S100, identifying the type of therapy head, includes:

[0152] The auxiliary control board encrypts the random code based on a preset algorithm and sends it to the main control board;

[0153] The main control board decrypts the random code based on the preset algorithm and then confirms the type of the physiotherapy head.

[0154] In practical applications, since the physiotherapy head may be copied by third parties, but the quality of third-party physiotherapy heads is not guaranteed, in this embodiment, only physiotherapy heads produced by manufacturers who know the preset algorithm can be recognized by the host, thus improving security.

[0155] In one embodiment, the time span of the target trajectory data for the duration of the cache can be used to identify unexpected movements. For example, 0.5 seconds to 3 seconds.

[0156] In one embodiment, the number of 3D cameras is multiple, and they are arranged to capture facial images from multiple angles; S500 includes: acquiring multi-angle facial images captured by the multiple 3D cameras of the physiotherapy robot.

[0157] In this embodiment, since the areas of the human face requiring treatment are mainly divided into three parts—the forehead, left cheek, and right cheek—it is difficult to obtain full profile data using a normal frontal view photography method. Therefore, it is necessary to add a corresponding shooting angle to obtain three-dimensional information of the profile while ensuring the identification of key facial points.

[0158] Therefore, this embodiment uses multiple 3D cameras, which are arranged to capture facial images from multiple angles.

[0159] In some embodiments, the point cloud data can also be smoothed and denoised to obtain backdated point cloud data;

[0160] It should be noted that the real-time facial images captured by the 3D camera contain non-target areas and noise interference. In order to improve the quality of the point cloud data,

[0161] In this embodiment, point cloud segmentation uses a pass-through filtering method, which sets a value range based on the extreme value and structural range of a certain dimension to generate new point cloud data;

[0162] In this embodiment, Gaussian filtering is used for smoothing and denoising the point cloud data. Statistical analysis is applied to nearby point clouds to determine if they are noise. A KD_Tree neighborhood search is performed to find N neighborhood points, and the average Euclidean distance from the sampling point to the neighborhood points is calculated.

[0163] Initial point cloud coordinates Sampling domain search point set The average Euclidean distance from the sampling point to the neighborhood point Distance can be expressed as:

[0164]

[0165] Initial point cloud The average Euclidean distance between it and its K neighborhood points mean and standard deviation They can be represented as:

[0166]

[0167]

[0168] The average Euclidean distance from the sampling points to the neighborhood points in the 3D point cloud acquired by the 3D camera should form a Gaussian distribution. By setting its mean and variance and replacing the original data with the calculated Gaussian smoothed data, the effects of noise reduction and data smoothing of the point cloud can be achieved.

[0169] Secondly, this application provides a physiotherapy robot, comprising:

[0170] At least two different treatment heads;

[0171] The host and the physiotherapy head are detachably mechanically and electrically connected. The host is used to execute the physiotherapy robot control program to realize the above-mentioned physiotherapy robot control method with replaceable physiotherapy head.

[0172] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0173] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0174] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0175] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0176] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.

[0177] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0178] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A physiotherapy robot, characterized in that, include: At least two different treatment heads; The main unit and the therapy head are detachably mechanically and electrically connected, the main unit being used for: Identify the type of the therapy head when changing the therapy head; Based on the type of the therapy head, confirm the mechanical structure and center of gravity of the therapy head; Based on the mechanical structure and center of gravity of the treatment head, the force control parameters of the robotic arm are adjusted to ensure a reasonable distance between the treatment head and the skin.

2. The physiotherapy robot as described in claim 1, characterized in that, The physiotherapy robot is also used for: Based on the type of treatment head, adjust the retraction parameters of the robotic arm to adjust the minimum distance between the treatment head and the skin during operation.

3. The physiotherapy robot as described in claim 2, characterized in that, The backoff parameter is used to perform backoff processing on the point cloud data generated based on the facial image during the physiotherapy process to obtain backoff point cloud data, so that the distance between the backoff point cloud data and the target facial area is greater than a preset safe distance.

4. The physiotherapy robot as described in claim 3, characterized in that, The physiotherapy robot is also used for: Acquire facial images in real time from the 3D camera of the physiotherapy robot; Point cloud data is generated based on the facial image; The point cloud data is superimposed with a unit vector of a preset multiple to obtain backdated point cloud data, wherein the distance between the backdated point cloud data and the target facial region is greater than a preset safe distance; Generate target trajectory data based on the reverted point cloud data; Based on the target trajectory data, the physiotherapy head of the physiotherapy robot is controlled to move in at least three degrees of freedom to perform physiotherapy on the target facial area.

5. The physiotherapy robot as described in claim 3, characterized in that, The types of therapy heads include: moxibustion therapy, red light therapy, and microwave therapy; among them, the retraction parameters are different for different therapy heads.

6. The physiotherapy robot as described in claim 4, characterized in that, The physiotherapy robot is used for: Based on the reverted point cloud data, the target trajectory data generated includes: Calculate the normal vector of the local plane corresponding to the backed-up point cloud data; The target trajectory data is calculated based on the normal vector.

7. The physiotherapy robot as described in claim 4, characterized in that, The target trajectory data includes a first deflection angle, a second deflection angle, and a third deflection angle; wherein the axes corresponding to the first deflection angle, the second deflection angle, and the third deflection angle are mutually perpendicular; the physiotherapy robot is used for: Based on satisfying the first deflection angle constraint, the second deflection angle constraint, and the third deflection angle constraint, the corresponding target trajectory data is calculated and generated, wherein... The first deflection angle constraint conditions include: First formula: And, the second formula: Wherein, the normal vector of the local plane RX represents the first deflection angle; The second deflection angle constraint includes: Third formula: And, the fourth formula: Where RY represents the second deflection angle; The third deflection angle constraint condition is: the third deflection angle is equal to 180°.

8. The physiotherapy robot as described in claim 4, characterized in that, The physiotherapy robot is also used for: Based on the physiotherapy plan selected by the user, the regression parameters are confirmed, wherein the regression parameters for facial physiotherapy are greater than those for back physiotherapy.

9. The physiotherapy robot as described in claim 1, characterized in that, The host computer is equipped with a main control board, and the therapy head is equipped with an auxiliary control board. The identification of the type of therapy head includes: The auxiliary control board encrypts the random code based on a preset algorithm and sends it to the main control board; The main control board decrypts the random code based on the preset algorithm and then confirms the type of the physiotherapy head.

Images