A control system for a massage device and a massage device
Patent Information
- Application Number
- CN202611141753.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-30
- Publication Date
- 2026-10-23
- Estimated Expiration
- 2046-07-30
AI Technical Summary
[0002]传统的按摩设备采用预设按摩程序,按照固定的轨迹对用户进行按摩,由于不同用户的体型、身高、肩颈腰背肌肉分布差异,在更换新用户的情况下,如果还是按照固定的按摩路径进行按摩,就会导致按摩位置偏移的问题,影响用户的按摩体验
本申请的技术方案,控制器基于检测图像确定按摩路径。当更换为新的用户时,就可以根据新的用户的检测图像确定该新的用户的按摩路径。提高了自适应性,有利于提高用户的按摩体验。
Smart Images

Figure CN122653038B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of massage equipment technology, specifically to a control system and massage equipment for a massage device. Background Technology
[0002] Traditional massage devices use preset massage programs to massage users along a fixed path. However, due to differences in body shape, height, and muscle distribution in the shoulders, neck, waist, and back among different users, if the massage is performed along the same path when a new user is introduced, the massage position will shift, affecting the user's massage experience. Summary of the Invention
[0003] To overcome the problems existing in the related technologies, this disclosure provides a control system for a massage device and a massage device.
[0004] Firstly, a control system for a massage device includes at least: A controller, and a camera device and a robotic arm drive device connected to the controller respectively; The camera device is used to acquire a detection image and send the detection image to the controller; The controller is used to determine the massage path corresponding to each part of the human body based on the detected image; A first control command is generated based on the massage path, and the first control command carries the massage path. Send the first control command to the robotic arm drive device; The robotic arm drive device is used to drive the robotic arm to massage the user along the massage path according to the massage path in the first control command received.
[0005] It also includes a torque sensor, which is used to detect the torque of the massage head and send the torque to the controller; The massage head is located at the end of the robotic arm and is used to massage the user. The controller is used to generate a second control command based on the torque of the massage head and a predetermined torque threshold; the second control command is used to control the robotic arm drive device to adjust the massage intensity of the massage head.
[0006] In some embodiments, the controller is further configured to, in response to receiving a first massage type instruction, send a third control instruction to the robotic arm, the third control instruction being configured to control the robotic arm to perform the first massage type action; The massage head is of type 1; The controller is further configured to, in response to receiving a second massage type instruction, send a fourth control instruction to the robotic arm, the fourth control instruction being used to control the robotic arm to perform the second massage type action; The massage head is of type two. The controller is also configured to, in response to receiving a third massage type instruction, send a fifth control instruction to the robotic arm, the fifth control instruction being used to control the robotic arm to perform the third massage type action; The massage head is of type three. The controller is also configured to, in response to receiving a fourth massage type instruction, send a sixth control instruction to the robotic arm, the sixth control instruction being used to control the robotic arm to perform the fourth massage type action; The massage head is of type four.
[0007] In some embodiments, the first massage type is point pressing, and the robotic arm performs the action of the first massage type by feeding linearly along the normal direction of the user's contact surface to achieve point pressing; The second massage type is pushing. The robotic arm performs the second massage type action by moving in a straight line along the user's muscle direction to achieve pushing and pressing. The third massage type is kneading. The robotic arm performs the third massage type action by executing a circular or elliptical motion trajectory to achieve kneading. The fourth massage type is rubbing. The robotic arm performs the fourth massage type by performing high-frequency, small-amplitude reciprocating linear motion to rub the user's muscles.
[0008] In some embodiments, the controller is further configured to determine one or more candidate bounding boxes in the detected image, wherein each candidate bounding box includes at least a detection confidence score; The target human body bounding box is determined based on the detection confidence level; Determine the key points of the human body within the target human body frame; Determine body parts based on the set of key points on the human body; Identify the target massage points for each part of the body; Massage paths are generated for each body part based on the massage target points.
[0009] In some embodiments, the controller is further configured to determine whether the back of the human body is facing the camera device based on a preset first judgment condition: It is also used to generate a stop massage command based on a preset second judgment condition.
[0010] In some embodiments, the controller is further configured to determine massage target points for each body part when the massage area is the spine, including: Determine the center points of the shoulder and hip respectively; Construct the longitudinal axis of the human body from the center of the shoulder to the center of the hip; Multiple massage target points are generated by interpolation at preset intervals along the vertical axis; Set the hovering height and posture parameters for each massage target point.
[0011] Secondly, this application also proposes a massage device, including a control system for the massage device described in any of the above claims, and further comprising: A robotic arm mounting base, used to fix the robotic arm; The bottom support is provided with a seat cushion. A reclining support, wherein the first end of the reclining support is rotatably connected to the bottom support; and the second end is connected to the first end of the neck brace. Armrest, the first end of which is connected to the backrest support; When the user is getting a massage, they sit on the cushion with their back facing the robotic arm and their arms resting on the armrest. The camera device at the front end of the robotic arm takes a picture of the user's back; The massage head at the end of the robotic arm comes into contact with the user's back and body for massage.
[0012] In some embodiments, a reclining backrest adjuster is further included, the reclining backrest adjuster being used to adjust the tilt angle of the reclining backrest; The first end of the reclining back adjuster is fixedly connected to the reclining back bracket; The backrest adjuster is arc-shaped; The reclining back adjuster is provided with multiple adjustment holes; When adjusting the tilt angle of the backrest bracket, replace the adjustment hole connected to the bottom bracket to adjust the angle of the backrest bracket; The multiple adjustment holes are evenly distributed.
[0013] In some embodiments, a neck brace height adjuster is also included. The neck brace height adjuster is used to adjust the height of the neck brace; The neck brace height adjuster includes at least: a first fixing plate, a first adjusting plate, and adjusting bolts; The first adjustment plate is used to connect the neck brace; The first fixing plate is fixedly mounted on the backrest bracket, on the side opposite to the robotic arm; The first fixing plate has multiple adjustment holes along the height direction; The first adjusting plate is provided with at least one through hole; The adjusting bolt is used to pass through the through hole on the first adjusting plate and into the adjusting hole on the first fixing plate to fix the first adjusting plate. Before adjusting the height of the neck brace, the adjusting bolt is inserted into the first adjusting hole; After adjusting the height of the neck brace, the adjusting bolt is inserted into the second adjusting hole.
[0014] The technical solutions provided by the embodiments of this disclosure may include the following beneficial effects: The technical solution of this application involves a controller determining the massage path based on detected images. When a new user is switched to a new user, the massage path for that new user can be determined based on their detected images. This improves adaptability and enhances the user's massage experience. Attached Figure Description
[0015] Figure 1 This is a block diagram of a control system for a massage device according to an embodiment of this application; Figure 2 This is a side view of a massage device according to an embodiment of this application; Figure 3 This is a front view of a massage device according to an embodiment of this application; Figure 4 This is a front view of a fastener according to an embodiment of this application. Detailed Implementation
[0016] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.
[0017] See appendix Figure 1 This application proposes a control system for a massage device, comprising at least: The controller 001, the camera device 002, and the robotic arm drive device 003 are respectively connected to the controller.
[0018] In this embodiment, the controller 001 may further include a processor 0010 and a memory 0011. The processor 0010 is used to execute the steps of the control method. The memory 0011 is used to store the relevant program of the control method.
[0019] The camera device 002 is used to acquire detection images and send the detection images to the controller.
[0020] In this embodiment, the imaging device can be a camera. Preferably, it is a depth camera / structure camera. The image captured by the depth camera is a depth image. The controller controls the start and stop of the camera.
[0021] The controller 001 is used to determine the massage path corresponding to each part of the human body based on the detected image.
[0022] In this embodiment, the detected image is a depth image, in which each pixel has depth data.
[0023] In some embodiments, the controller is further configured to generate a stop massage command when the following conditions occur in multiple consecutive image frames; A first control command is generated based on the massage path, and the first control command carries the massage path. The first control command is sent to the robotic arm drive device 003.
[0024] The robotic arm drive device 003 is used to drive the robotic arm 005 to massage the user along the massage path according to the massage path in the first control command received.
[0025] In this embodiment, the robotic arm is a 6 / 7-axis lightweight collaborative robotic arm. A massage head is provided at the front end of the robotic arm, and the massage head provides a massage to the user under the action of the robotic arm.
[0026] The technical solution of this application involves a controller determining the user's local body regions based on images of those regions, and then determining the corresponding massage paths. When a new user is switched to a different user, the controller can determine the massage path for that new user based on images of their local body regions. This improves adaptability and enhances the user's massage experience.
[0027] In some embodiments, the control system of the massage device described above further includes a torque sensor 004, which is used to detect the torque of the massage head and send the torque to the controller.
[0028] The massage head is located at the end of the robotic arm and is used to massage the user. The controller is used to generate a second control command based on the torque of the massage head and a predetermined torque threshold; the second control command is used to control the robotic arm drive device to adjust the massage intensity of the massage head.
[0029] In this embodiment, the process of generating a second control command based on the torque of the massage head and a predetermined torque threshold may further include the following steps: When the torque of the massage head is greater than a predetermined torque threshold compared to the torque threshold, a second control command is generated to reduce the torque of the massage head. When the torque of the massage head is less than a predetermined torque threshold, and the difference between the torque threshold and the torque of the massage head is greater than a torque difference threshold, a second control command is generated to increase the torque of the massage head.
[0030] In this embodiment, the torque difference threshold can be flexibly set.
[0031] In some embodiments, when the torque of the massage head is less than a predetermined torque threshold compared to a predetermined torque threshold, and the difference between the torque threshold and the torque of the massage head is less than or equal to a torque difference threshold, no torque adjustment is required, and no second control command needs to be generated.
[0032] In some embodiments, the controller is further configured to, in response to receiving a first massage type instruction, send a third control instruction to the robotic arm, the third control instruction being configured to control the robotic arm to perform the first massage type action; The massage head is of type 1; The controller is further configured to, in response to receiving a second massage type instruction, send a fourth control instruction to the robotic arm, the fourth control instruction being used to control the robotic arm to perform the second massage type action; The massage head is of type two. The controller is also configured to, in response to receiving a third massage type instruction, send a fifth control instruction to the robotic arm, the fifth control instruction being used to control the robotic arm to perform the third massage type action; The massage head is of type three. The controller is also configured to, in response to receiving a fourth massage type instruction, send a sixth control instruction to the robotic arm, the sixth control instruction being used to control the robotic arm to perform the fourth massage type action; The massage head is of type four.
[0033] In some embodiments, the first massage type is point pressing, and the robotic arm performs the action of the first massage type by feeding linearly along the normal direction of the user's contact surface to achieve point pressing; The second massage type is pushing. The robotic arm performs the second massage type action by moving in a straight line along the user's muscle direction to achieve pushing and pressing. The third massage type is kneading. The robotic arm performs the third massage type action by executing a circular or elliptical motion trajectory to achieve kneading. The fourth massage type is rubbing. The robotic arm performs the fourth massage type by performing high-frequency, small-amplitude reciprocating linear motion to rub the user's muscles.
[0034] In some embodiments, the controller is further configured to determine one or more candidate bounding boxes in the detected image, wherein each candidate bounding box includes at least a detection confidence score.
[0035] The target human body bounding box is determined based on the detection confidence level; Determine the key points of the human body within the target human body frame.
[0036] Among them, the preferred key points of the human body include: left shoulder key point, right shoulder key point; left hip key point, right hip key point; head key point / face key point; trunk key point; upper back key point, lower back key point.
[0037] Filter based on the visibility and confidence of each keypoint to remove invalid keypoints.
[0038] In this embodiment, the two-dimensional coordinates of key human body points within the target human body frame can also be determined and transformed into three-dimensional coordinates. After the coordinate transformation, it is easier to control the movement of the robotic arm based on the three-dimensional coordinates.
[0039] Specifically, the process of converting two-dimensional coordinates to three-dimensional coordinates includes the following steps: For each key point or region, extract the depth value of the corresponding pixel from the depth map.
[0040] To avoid single-point depth noise, one of the following methods is preferred: Sample multiple pixels within the neighborhood of the key point and calculate the average value; After extracting the effective depth within the region, calculate the median. Cluster the regions by depth and take the depth of the foreground main cluster; Distance is calculated after removing zero-value depths and outliers.
[0041] By combining the camera intrinsic parameters, the depth of each pixel and the two-dimensional pixel coordinates are converted into three-dimensional coordinates to obtain the position of the human body part in the camera coordinate system.
[0042] The body parts are determined by the set of key points on the human body.
[0043] In this embodiment, semantic recognition of human body parts can be performed based on a set of key human body points. A preferred recognition method is as follows: Shoulder area: Composed of key points on the left and right shoulders; The center line of the back is approximately obtained by connecting the center of the shoulder and the center of the hip. Upper back area: Located within a preset proportion below the center of the shoulder; Lower back / lumbar region: The lower section of the body located between the center of the shoulder and the center of the hip; Midline of the spine: obtained by interpolation fitting of the center of the shoulder, center of the trunk, and center of the hip; Local massage points: Extract specific points along the center line of the back.
[0044] In some embodiments, considering the case of missing key points, if some key points are missing, the following degradation estimation scheme is adopted: The center of the human body frame is used to approximate the center of the torso. In this embodiment, the center of the human body frame can be used to approximate the center of the torso because the center of the human body frame is easier to calculate and can improve computational efficiency.
[0045] The longitudinal axis of the human body is fitted using the remaining visible shoulder / hip points. In this embodiment, the remaining visible shoulder / hip key points can be used for fitting to obtain the longitudinal axis of the human body.
[0046] The missing key points in the current image frame are smoothly compensated by combining key points from historical image frames. In this embodiment, historical image frames can be fully utilized to compensate for the lack of key points in the current image frame.
[0047] Identify the massage target points for each body part. In this embodiment, for example, if the body part is the shoulder area, the massage target points are the key points of the shoulder; if the body part is the waist area, the massage target points are the key points of the waist; if the body part is the spine area, the massage target points are multiple key points on the spine.
[0048] Massage paths are generated for each body part based on its massage target points. Once the massage target points for each body part have been determined, massage paths for each body part can be generated based on the coordinates of these target points.
[0049] In this embodiment, a muscle recognition algorithm can be used to determine the spinal position, muscle boundaries, and massage target points to generate a massage path. The aforementioned muscle recognition algorithm, based on image / depth sensing data, automatically segments key muscle groups in the neck, shoulders, waist, and back, extracts muscle contours, and classifies and identifies multiple muscle groups. By fusing constraints from key points of the human skeleton through a lightweight feature extraction network, it distinguishes target muscles such as the trapezius, latissimus dorsi, erector spinae, and scapular muscles, suppressing recognition interference caused by clothing obscuring the image, body shape deformation, and uneven lighting. It outputs muscle masks, region coordinates, and visualized muscle partitioning results, supporting downstream applications such as posture assessment, muscle strain localization, and quantitative analysis of rehabilitation therapy.
[0050] Different massage types may use different path generation methods: Pushing: generates a continuous linear path along the longitudinal axis; Kneading: generates a small circular or elliptical trajectory around a local target point; Rubbing: generates a high-frequency back-and-forth short path in a local area; Pointing: selects one or more fixed target points.
[0051] In some embodiments, the controller is further configured to determine whether the back of the human body is facing the camera device based on a preset first judgment condition.
[0052] In this embodiment, the first judgment condition includes the following, which are calculated based on the following features: The number of missing facial key points; determining the number of missing facial key points; in response to the number of missing facial key points being less than a predetermined threshold for the number of missing facial key points, determining that the back of the human body is facing the camera device.
[0053] The key points on the left and right shoulders are relatively stable, but the key points on the face are not visible; in this embodiment, the absence of visible facial key points indicates that the user is facing away from the camera device. The relative stability of the key points on the left and right shoulders indicates that the user's posture remains stable.
[0054] Integrity of the torso region; in this embodiment, image features of the torso region can be identified, and the integrity of the torso region can be determined based on the image features.
[0055] Orientation score of historical continuous image frames: In this embodiment, the orientation score of each image frame in the historical continuous image frames can be calculated. When the orientation scores of multiple consecutive image frames are higher than the orientation score threshold, it is determined that the human body is in a state where its back is facing the camera device.
[0056] In some embodiments, the controller is further configured to generate a stop massage command based on a preset second judgment condition.
[0057] In this embodiment, the second judgment condition mentioned above includes, but is not limited to, the following: The main target in the image frame is missing; the main target in the image frame has turned around; the offset distance of each body part in the image frame exceeds the offset distance threshold for each body part; body parts in the image frame cannot be identified. Depth data in the image frame is invalid; for example, the depth map displays blank or black areas, or the depth value is an extremely large negative number (0). Visual service malfunctions, such as not receiving images, camera device failure, or inability to capture images.
[0058] In some embodiments, the controller is further configured to determine the massage target points for each body part when the massage area is the spine, and may further include the following steps: Determine the center points of the shoulder and hip respectively. Construct a longitudinal axis of the human body from the center of the shoulder to the center of the hip. Generate multiple massage target points on the longitudinal axis by interpolation at preset intervals. The preset intervals can be flexibly set or determined based on historical experience data. Set the hovering height and posture parameters for each massage target point. The hovering height and posture parameters for each massage target point can be determined based on historical experience data to verify the massage effect of that massage target point.
[0059] In some embodiments, the controller is also used to interact with an external APP client; it is equipped with a touch panel or a dedicated handle as the primary control method, and some support remote setting programs via the APP. This enables fine-grained parameter settings and mode storage.
[0060] The app can send the following control commands to the controller: massage area selection (cervical spine / upper back / lower back / waist); technique selection; intensity level setting; working time setting; custom massage trajectory editing; start / stop / reset.
[0061] After receiving the command, the controller parses it into motion control parameters and torque parameters, drives the robotic arm to perform the corresponding action, and at the same time sends the current status (position, force, running time) back to the APP for display.
[0062] In some embodiments, the system further includes a voice interaction module 007 and a communication module 006. The voice interaction module 007 is used to implement voice interaction and voice control functions. Users can adjust the massage state in real time without manual operation. The voice interaction module includes a microphone array, a voice recognition chip, and a voice command parsing unit.
[0063] A microphone array is used to collect voice commands; a voice recognition chip is used to convert voice into text commands; and a voice command parsing unit is used to parse commands, including: start / pause massage; increase / decrease intensity; switch massage position; switch between push / knead / rub / press techniques; and adjust massage speed.
[0064] The controller updates motion parameters and torque thresholds in real time according to instructions, enabling real-time voice control.
[0065] Communication module 006 is used for wireless communication with external devices. Communication module 006 can be a wireless communication module, employing WiFi or Bluetooth communication units to establish a wireless connection with a mobile terminal app.
[0066] In some embodiments, the controller is also used to perform a visual review of the process. During the massage, the latest image frames are continuously acquired to review the main target and body parts.
[0067] The review may include, but is not limited to: whether the main target trajectory remains consistent; whether the back orientation is maintained; whether the target part is still within the allowable deviation range; whether the human body has left the frame; and whether the camera recognition results are continuously abnormal.
[0068] In some embodiments, the controller is also used for torque control and flexible feedback control. This ensures moderate massage intensity and avoids rigid impacts and overload damage. The torque sensor collects the end-effector output pressure and torque in real time and converts the analog signal into a digital signal, which is then fed back to the controller.
[0069] The controller is equipped with a dual closed-loop control system. The position loop is responsible for the precise execution of the massage trajectory; the torque loop is responsible for comparing the actual torque with the set torque in real time. When the actual torque exceeds the threshold, the controller immediately reduces the feed speed of the robotic arm or reverses it to achieve flexible force control.
[0070] When at least one of the following conditions is detected, the controller immediately triggers safety protection, controlling the robotic arm to stop feeding and maintain flexible avoidance, thereby achieving safe torque limiting and real-time compliant control. These conditions include, but are not limited to: sudden torque changes, sudden pressure increases, and abnormal resistance.
[0071] In some embodiments, the controller is further configured to maintain information for each target, including but not limited to the following: trajectory number; number of consecutive occurrences; number of consecutive missing frames; number of consecutive back frames; number of consecutive non-back frames; and whether it is a primary target. The system preferentially selects targets with high back scores, stable depth, and long continuous durations as primary targets.
[0072] In some embodiments, the controller is also used for multi-frame target tracking and primary target locking. Considering that the human body may move, occlude, or turn, a trajectory is established for each human body, and cross-frame matching is performed. The trajectory matching is preferably based on, but is not limited to, the following: bounding box center distance; depth difference; change in body part position; continuity of back state; and historical confidence.
[0073] In some embodiments, the controller is also used to provide multi-level safety protection functions, preventing safety risks caused by malfunctions, excessive torque, or out-of-range movement. Safety protection is achieved through a combination of multiple hardware and software components. Torque overload protection: If the torque exceeds the set threshold, the vehicle will immediately decelerate and retract. Mechanical limit protection: Physical limit switches are installed on the joints of the robotic arm to prevent it from exceeding the safe working range.
[0074] Abnormal emergency stop protection: The controller monitors the motor current and attitude in real time for abnormalities, and immediately cuts off the drive output when an abnormality is detected; Skeletal avoidance protection: When the visual system identifies the spine or skeletal area, the controller automatically reduces the force or avoids the area.
[0075] All safety logic is executed independently by the controller, without relying on external devices, ensuring safe operation. The workflow and action relationships are explained below: 1. User in position. The user lies prone on the main body of the massage stool, with the neck, back, and waist naturally exposed to the working area of the robotic arm.
[0076] 2. Visual Recognition and Positioning. The controller activates the visual recognition camera to capture images of the back, and uses a muscle recognition algorithm to determine the position of the spine, muscle boundaries, and target massage points, generating a massage path.
[0077] 3. Initial positioning of the robotic arm. The controller drives the robotic arm to move according to the target coordinates, so that the massage head moves above the massage starting position and maintains a safe distance.
[0078] 4. Massage execution. The controller outputs the motion trajectory according to the selected technique (push / knead / rub / press), and simultaneously activates torque closed-loop control to adjust the intensity in real time.
[0079] 5. Real-time interactive adjustment. Users can adjust the position, intensity, and technique in real time via voice or an app, and the controller will respond instantly and update the motion parameters.
[0080] 6. Torque Feedback and Safety Protection. The torque sensor continuously provides pressure feedback, and the controller dynamically adjusts the feed rate to ensure gentle force without exceeding limits.
[0081] 7. End and Reset. After the massage is complete, the controller guides the robotic arm to return to the initial standby position along a safe path, and the device enters standby mode.
[0082] This application also proposes a control method for a massage device, applied to the aforementioned controller, or processor, or processor within a controller, the method further comprising the following steps: 1. Stable identification of human targets from detected images; 2. Further identify body parts such as the neck, shoulders, back, and waist; 3. Obtain the three-dimensional spatial position of the aforementioned human body parts; 4. Automatically generate massage target points and massage paths based on body parts; 5. Maintain target stability when the human body moves, turns, or is partially obscured; 6. Continuously verify the effectiveness of the target and trigger the stop protection during the massage process.
[0083] Input data may include: color images; depth images; camera intrinsic parameters; historical target tracking status.
[0084] The output results may include: the current human target; the position of human body parts such as neck, shoulders, back, and waist; the three-dimensional coordinates of human body parts; the back orientation determination result; the main target trajectory number; the massage target point and massage path; and the target validity status during the execution period.
[0085] Secondly, see the appendix. Figure 2-3 This application proposes a massage device, including a control system for the massage device described in any of the above claims, and further comprising: Robotic arm mounting base 016, the robotic arm mounting base is used to fix the robotic arm 005.
[0086] In this embodiment, the robotic arm mounting base is also called a rear mounting bracket.
[0087] Bottom support 014; a seat cushion 015 is provided on the bottom support.
[0088] In this embodiment, the bottom support can also be referred to as a tilt massage stool.
[0089] The backrest support 011 has a first end that is rotatably connected to the bottom support; and a second end that is connected to the first end of the neck brace.
[0090] Armrest 012, the first end of which is connected to the backrest support. During a massage, the user sits on the cushion with their back facing the robotic arm and their arms resting on the armrest. A camera device 002 at the front end of the robotic arm captures a video of the user's back; a massage head 008 at the end of the robotic arm makes contact with and massages the user's back.
[0091] The controller 001 is mounted on the side of the robotic arm mounting base 016; the controller 001 is electrically connected to the aforementioned robotic arm drive device 003.
[0092] In some embodiments, the robotic arm drive device 003 can be a motor; the motor can be installed at the joint positions of the robotic arm 005.
[0093] For example, robotic arm 005 includes a first robotic arm segment and a second robotic arm segment; The first motor can be set at the position of the first joint connecting the first segment of the robotic arm and the second segment of the robotic arm. The second motor can be installed at the position of the second joint where the first segment of the robotic arm connects to the robotic arm mounting base 016.
[0094] In some embodiments, the aforementioned prone massage stool body is designed for a user to lie prone, with the stool surface ergonomically configured to provide support areas for the neck, chest, and lower back. A rear mounting bracket is fixedly mounted on the rear side of the massage stool for mounting a multi-degree-of-freedom robotic arm, ensuring the robotic arm's workspace covers the user's cervical spine, back, and lower back areas. The fixed end of the multi-degree-of-freedom robotic arm is connected to the rear mounting bracket, while its movable end is connected to the multi-shaped massage head's actuator. A visual recognition camera is mounted on the robotic arm's end or above the bracket to capture images of the user's back area and transmit them to the controller module. A torque sensor is connected in series between the robotic arm's end and the massage head's actuator to collect massage pressure and torque signals in real time.
[0095] The controller is electrically connected to the robotic arm drive unit, visual recognition camera, torque sensor, voice interaction module, and wireless communication module.
[0096] In some embodiments, the massage device further includes a backrest adjuster 013, which is used to adjust the tilt angle of the backrest support; the first end of the backrest adjuster is fixedly connected to the backrest support; the backrest adjuster is arc-shaped; the backrest adjuster is provided with a plurality of adjustment holes; when adjusting the tilt angle of the backrest support, the adjustment hole connected to the bottom support is replaced to adjust the angle of the backrest support.
[0097] In some embodiments, the plurality of adjustment holes are evenly distributed.
[0098] In some embodiments, a neck brace height adjuster 010 is further included; the neck brace height adjuster is used to adjust the height of the neck brace 009. The neck brace height adjuster 010 includes at least: a first fixing plate 020, a first adjusting plate 021, and an adjusting bolt 018; the first adjusting plate is used to connect to the neck brace. In this embodiment, the first adjusting plate can be directly connected to the neck brace, or it can be connected to the neck brace via a connector.
[0099] See appendix Figure 3 The first adjustment plate is connected to the neck brace 009 via a connector 017. The neck brace has a curved surface that matches the user's neck and supports the neck. The curved surface is perpendicular to the inner and outer sides of the neck brace. The inner side of the neck brace faces the robotic arm, and the outer side faces away from the robotic arm.
[0100] The outer side of the neck brace is fixedly connected to the connecting piece. Alternatively, the outer side of the neck brace is directly connected to one end of the first adjusting plate.
[0101] The first fixing plate is fixedly mounted on the backrest bracket, on the side opposite to the robotic arm.
[0102] In this embodiment, the first fixing plate can be fixed to the backrest bracket by means of screws.
[0103] In some embodiments, the thickness of the first fixing plate needs to be set to be greater than a predetermined thickness threshold.
[0104] See appendix Figure 4 The first fixing plate 020 has a plurality of adjustment holes 0201 along the height direction; The first adjustment plate has at least one through hole.
[0105] The adjusting bolt 018 is used to pass through the through hole on the first adjusting plate and into the adjusting hole on the first fixing plate to fix the first adjusting plate. Before adjusting the height of the neck brace, the adjusting bolt is inserted into the first adjusting hole; after adjusting the height of the neck brace, the adjusting bolt is inserted into the second adjusting hole.
[0106] In this embodiment, the height of the neck brace can be adjusted by adjusting the first adjusting plate. The first adjusting plate is connected to the first fixed plate via an adjusting bolt, which is flexibly and detachably connected. Before adjusting the position, the adjusting bolt passes through the through hole on the first adjusting plate and into the first adjusting hole on the first fixed plate. After moving the first adjusting plate upwards, the adjusting bolt passes through the through hole on the first adjusting plate and into the second adjusting hole on the first fixed plate.
[0107] The first adjustment plate can also be equipped with a hollow structure to reduce weight. As can be seen from the figure, the hollow shape is triangular, and through the hollow triangular area, multiple adjustment holes on the first fixing plate can be seen.
[0108] Compared with existing solutions that rely solely on fixed trajectories or simple human detection, the technical solution of this application has the following advantages: Improved detection accuracy allows for the identification of specific massage areas; deep fusion provides 3D coordinates of body parts, enhancing positioning accuracy; it can determine back orientation, reducing accidental massage; it maintains stability of the primary target during body movement; it automatically generates massage paths adapted to body shape; and it continuously verifies the target during execution, improving safety. Through human body detection, key point estimation, deep fusion, body part identification, back assessment, multi-frame tracking, and massage path generation, it achieves automatic identification, stable positioning, and closed-loop monitoring of the neck, shoulders, back, and waist target areas.
[0109] This invention uses visual recognition technology to automatically locate human muscle areas, solving the problem of inaccurate positioning in traditional massage devices; By using a multi-degree-of-freedom robotic arm and multi-shaped massage heads, simulated techniques such as pushing, kneading, rubbing, and pressing are achieved, solving the problems of monotonous and stiff techniques. The torque sensor and dual closed-loop feedback control enable gentle massage, solving the problems of uncontrollable intensity and discomfort. Multi-mode intelligent control is achieved through voice and wireless communication modules, improving ease of use; A multi-level software and hardware security protection mechanism ensures the safety and reliability of the massage process.
[0110] The overall device has a simple structure and is easy to set up, enabling precise, safe, and intelligent massage of the neck, shoulders, and waist while in a prone position.
[0111] The various techniques described herein can be implemented in combination with hardware or software, or a combination thereof. The methods and apparatus of the present invention, or certain aspects or portions thereof, may take the form of program code (i.e., instructions) embedded in a tangible medium, such as a removable hard disk, USB flash drive, floppy disk, CD-ROM, or any other machine-readable storage medium, wherein when the program is loaded into and executed by a machine such as a computer, the machine becomes an apparatus for practicing the present invention.
[0112] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of this disclosure. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.
[0113] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.
Claims
1. A control system for a massage device, characterized in that, At least including: A controller, and a camera device and a robotic arm drive device connected to the controller respectively; The camera device is used to acquire a detection image and send the detection image to the controller; The controller is used to determine the massage path corresponding to each part of the human body based on the detected image; A first control command is generated based on the massage path, and the first control command carries the massage path. Send the first control command to the robotic arm drive device; The robotic arm drive device is used to drive the robotic arm to massage the user along the massage path according to the massage path in the first control command received; It also includes a torque sensor, which is used to detect the torque of the massage head and send the torque to the controller; The massage head is located at the end of the robotic arm and is used to massage the user. The controller is configured to generate a second control command based on the torque of the massage head and a predetermined torque threshold; the second control command is configured to control the robotic arm drive device to adjust the massage intensity of the massage head. The controller is further configured to determine whether the back of the human body is facing the camera device based on a preset first judgment condition: The first determination condition includes: the number of missing facial key points; in response to the number of missing facial key points being less than a predetermined threshold for the number of missing facial key points, it is determined that the back of the human body is facing the camera device; The key points on the left and right shoulders are relatively stable, but the key points on the face are not visible. Orientation score of historical continuous image frames; calculate the orientation score of each image frame in the historical continuous image frames, and determine that the human body is in a state where its back is facing the camera device when the orientation scores of multiple consecutive image frames are higher than the orientation score threshold; It is also used to generate a stop massage command based on a preset second judgment condition; The second judgment conditions include: the main target in the image frame is lost; the main target in the image frame turns around; the offset distance of each human body part in the image frame is greater than the offset distance threshold of each human body part; the human body part in the image frame cannot be identified; the depth data in the image frame is invalid; and the visual service is abnormal.
2. The control system of the massage device according to claim 1, characterized in that, The controller is further configured to, in response to receiving a first massage type instruction, send a third control instruction to the robotic arm, the third control instruction being used to control the robotic arm to perform the first massage type action; The massage head is of type 1; The controller is further configured to, in response to receiving a second massage type instruction, send a fourth control instruction to the robotic arm, the fourth control instruction being used to control the robotic arm to perform the second massage type action; The massage head is of type two. The controller is also configured to, in response to receiving a third massage type instruction, send a fifth control instruction to the robotic arm, the fifth control instruction being used to control the robotic arm to perform the third massage type action; The massage head is of type three. The controller is further configured to, in response to receiving an instruction of a fourth massage type, send a sixth control instruction to the robotic arm, where the sixth control instruction is used to control the robotic arm to perform the action of the fourth massage type; The type of the massage head adopts a fourth type.
3. The control system for a massage device according to Claim 2, characterized in that, The first massage type is point pressing, and the action performed by the robotic arm for the first massage type is that the robotic arm feeds linearly in the normal direction of the contact surface of the user to realize fixed-point pressing; The second massage type is pushing, and the action performed by the robotic arm for the second massage type is that the robotic arm moves linearly along the muscle direction of the user to realize pushing massage; The third massage type is kneading, and the action performed by the robotic arm for the third massage type is that the robotic arm follows a circular or elliptical motion trajectory to realize kneading; The fourth massage type is rubbing, and the action performed by the robotic arm for the fourth massage type is that the robotic arm performs high-frequency small-amplitude reciprocating linear motion to rub the user's muscles.
4. The control system for a massage device according to Claim 1, characterized in that, The controller is further configured to determine one or more candidate human body frames in the detected image, where each candidate human body frame at least includes a detection confidence; Determine a target human body frame based on the detection confidence; Determine human body key points in the target human body frame; Determine human body parts according to the set of human body key points; Determine a massage target point for each human body part; Generate a massage path for each human body part according to the massage target point of the human body part.
5. The control system of the massage device as described in claim 1, characterized in that, The controller is further configured to, when the massage part is the spine, determining the massage target point for each human body part includes: Determine the center point of the shoulder and the center point of the hip respectively; Construct a longitudinal human axis from the shoulder center to the hip center; Generate a plurality of massage target points by interpolation on the longitudinal axis at preset intervals; Set a hovering height and attitude parameter for each massage target point.
6. A massage device, characterized in that, A massage device comprising the control system for a massage device according to any one of Claims 1 to 5, further comprising: A robotic arm fixing base, configured to fix the robotic arm; A bottom support; a seat cushion is arranged on the bottom support; A prone back support, a first end of the prone back support is rotatably connected to the bottom support; and a second end of the prone back support is connected to a first end of a neck support; An arm support, a first end of the arm support is connected to the prone back support; When receiving a massage, the user sits on the seat cushion, with the back facing the robotic arm, and both arms placed on the arm support; The camera device at the front end of the robotic arm captures an image of the back of the user; The massage head at the end of the robotic arm is in contact with the back of the user for massage.
7. The massage device according to Claim 6, characterized in that, The massage device further comprises a prone back adjuster, which is configured to adjust the inclination angle of the prone back support; A first end of the prone back adjuster is fixedly connected to the prone back support; The prone back adjuster is in an arc shape; A plurality of adjusting holes are arranged on the prone back adjuster; When adjusting the inclination angle of the prone back support, the adjusting hole connected to the bottom support is replaced to adjust the angle of the prone back support; The multiple adjustment holes are evenly distributed.
8. The massage device as described in claim 6, characterized in that, It also includes a neck brace height adjuster. The neck brace height adjuster is used to adjust the height of the neck brace; The neck brace height adjuster includes at least: a first fixing plate, a first adjusting plate, and adjusting bolts; The first adjustment plate is used to connect the neck brace; The first fixing plate is fixedly mounted on the backrest bracket, on the side opposite to the robotic arm; The first fixing plate has multiple adjustment holes along the height direction; The first adjusting plate is provided with at least one through hole; The adjusting bolt is used to pass through the through hole on the first adjusting plate and into the adjusting hole on the first fixing plate to fix the first adjusting plate. Before adjusting the height of the neck brace, the adjusting bolt is inserted into the first adjusting hole; After adjusting the height of the neck brace, the adjusting bolt is inserted into the second adjusting hole.
Citation Information
Patent Citations
Massage robot
CN115847393A
Human body posture key point identification method and device for auxiliary rehabilitation exercise
CN118570838A
Massage robot based on dual-electrode massage head and massage method thereof
CN118986704A