Full-automatic infrared moxibustion physiotherapy system based on image processing and positioning method

Abstract

The invention discloses a full-automatic infrared moxibustion physiotherapy system based on image processing and a positioning method, and relates to a full-automatic infrared moxibustion physiotherapy system and a positioning method. The invention aims to solve the problems of heat concentration, lack of personalized treatment on different parts, inaccurate acupuncture point positioning and slow response in the prior art. The system comprises an image processing module, a mechanical arm control module and a man-machine interaction module. The image processing module is composed of a camera and an image processing platform. The mechanical arm control module is composed of a main controller, a mechanical arm, a temperature sensor, a distance sensor and an infrared moxibustion tail end. The man-machine interaction module is an interaction system with a graphical interface, the interaction system with the graphical interface integrates a real-time image loading function and an automatic acupoint positioning method function, communicates with the mechanical arm main controller through a wireless serial port, sends a G code control instruction and receives and displays position information of various sensors and the mechanical arm. The invention belongs to the technical field of medical instruments.

Description

Technical Field

[0001] The invention relates to a full-automatic infrared moxibustion therapy system and a positioning method, belonging to the technical field of medical devices. Background Art

[0002] Traditional Chinese medicine's moxibustion therapy can effectively help patients recover quickly from chronic diseases by applying heat stimulation to certain acupuncture points; however, the implementation of moxibustion therapy is inconvenient due to its complicated operation and the generation of smoke during the process.

[0003] Zhang Deyi and others from Jiangsu University of Science and Technology designed an intelligent moxibustion box, which aims to solve the problems of uneven heat, low thermal efficiency, and cumbersome operation in the existing moxibustion process, so as to achieve the purpose of efficient use of heat and medicinal smoke. This moxibustion device includes a combustion, evaporation, collection and control system, which can realize automatic smoke collection, precise and intelligent temperature control, and simplified moxibustion operation process.

[0004] Li Min and others from Nanjing University of Chinese Medicine designed a fully automatic temperature-controlled moxibustion box, which is characterized by being able to automatically control the temperature inside the moxibustion box between 38 and 60°C, aiming to solve problems such as low-temperature burns caused by inaccurate temperature control in traditional moxibustion. The working principle of the moxibustion box is mainly to adjust the opening and closing of the pores in the moxibustion box through the expansion and contraction of a thermal spring, thereby controlling the internal temperature. One end of the thermal spring is connected to the side wall of the lower box body, and the other end is connected to the heat-conducting layer. The increase in temperature inside the moxibustion box will cause the thermal spring to stretch, pushing the heat-conducting layer forward, causing the pores to gradually close; when the temperature drops, the thermal spring contracts, the heat-conducting layer returns to its original position, and the pores reopen, allowing air to enter and promoting the burning of moxa sticks. This design realizes the automatic regulation of the temperature inside the moxibustion box.

[0005] Zhao Guoyou and others from Sichuan University designed a six-degree-of-freedom moxibustion robot arm. This moxibustion robot arm combines cutting-edge control systems and mechanical designs to automate the moxibustion process and perform a variety of moxibustions. The moxibustion robot arm is designed with two modes: manual and automatic. In manual mode, the operator can personally control the movement of the robot arm and perform moxibustion treatment, such as adjusting the position, depth and duration of moxibustion; in automatic mode, the moxibustion robot arm will automatically complete the moxibustion process according to the pre-set process.

[0006] Yao Yong and others from Shanghai University of Traditional Chinese Medicine designed a three-degree-of-freedom moxibustion robot arm. The system uses the three-degree-of-freedom robot arm to implement precise motion control, enabling it to perform specific moxibustion techniques on different acupuncture points on the human body, such as spiral moxibustion and sparrow pecking moxibustion. The monocular camera, as the "eye" of the system, captures the image of the acupuncture point label with a blue filter patch to accurately locate the acupuncture point. In addition, the system's built-in Bluetooth module enables communication between the microcontroller and the smartphone, allowing the operator to conveniently control the robot arm and receive treatment feedback in real time. The system is equipped with a temperature sensor and a distance sensor. The former monitors the moxibustion temperature to prevent overheating, and the latter ensures that the robot arm maintains an appropriate distance from the skin to avoid contact injury. The introduction of a high-pressure purification device effectively handles the smoke generated during the moxibustion process.

[0007] Wang Cong from Beijing University of Posts and Telecommunications first used the Kinect depth camera to obtain human body images and depth information, especially for more regular body parts such as arms; by using a modified curvature analysis technology, the paper successfully identified the position of the horizontal lines on the elbow, which is a prerequisite for locating the acupoints on the hands; then the OpenPose network was used to accurately locate the acupoints on the hands; in order to track the changes in limbs during movement, the Kalman filter algorithm was used to predict the movement trajectory of the limbs; in addition, the optical flow method was used to track the real-time position of the target acupoints to ensure that the acupoints can be accurately identified under dynamic conditions.

[0008] Although the above technical solution can solve the problems of cumbersome operation and smoke generation during real-time moxibustion, there are problems of heat concentration, lack of personalized treatment for different parts, inaccurate acupoint positioning and slow response. Summary of the invention

[0009] The present invention aims to solve the problems of heat concentration, lack of personalized treatment for different parts, inaccurate acupoint positioning and slow response in the prior art, and further proposes a fully automatic infrared moxibustion therapy system and positioning method based on image processing.

[0010] The technical solution adopted by the present invention to solve the above problems is: the fully automatic infrared moxibustion therapy system based on image processing described in the present invention comprises an image processing module, a mechanical arm control module and a human-computer interaction module;

[0011] The image processing module consists of a camera and an image processing platform. The camera collects real-time images of the back of a human body placed on the working plane and transmits them to the human-machine interaction module and the robotic arm control module through a wireless serial port for subsequent processing and analysis.

[0012] The robotic arm control module is composed of a main controller, a robotic arm, a temperature sensor, a distance sensor and an infrared moxibustion terminal. The main controller receives instructions from the human-computer interaction page, and the driving module of the robotic arm receives the G code control instructions from the main controller to drive the infrared moxibustion terminal to move, thereby realizing the human body's gentle moxibustion action. While the robotic arm is running, the temperature sensor and the distance sensor respectively collect the surface temperature of the moxibustion site and the distance between the robotic arm terminal and the surface of the moxibustion site. The temperature sensor and the distance sensor upload the surface temperature and distance to the main controller, and the robotic arm terminal clamps the infrared therapy lamp for heat therapy.

[0013] The human-computer interaction module includes an interactive system with a graphical interface, which integrates functions such as real-time image loading and automatic acupoint positioning methods; the human-computer interaction page communicates with the main controller and the drive module of the robotic arm through the wireless serial port, and is used to send G-code control instructions, receive and display temperature sensor, distance sensor, and robotic arm position information.

[0014] The steps of the fully automatic infrared moxibustion therapy positioning method based on image processing described in the present invention include:

[0015] Step 1: Establish a heat transfer model to predict surface temperature changes under infrared therapy conditions;

[0016] First, the time, distance and temperature data under different conditions are collected and recorded through cameras, temperature sensors, distance sensors and HMI. Then, the heat transfer model is corrected and optimized based on the actual collected data to ensure that the simulation results are as close and accurate as possible to the actual situation.

[0017] Secondly, a long short-term memory network-convolutional neural network deep learning model is used to train and learn the time, distance and temperature sequence data obtained from the thermal field simulation model; the goal of this step is to establish a model that can accurately predict the surface temperature changes during physical therapy, providing a basis for subsequent physical therapy control strategies;

[0018] Step 2: Identify the moxibustion site, accurately locate the moxibustion site under different lighting and surface type conditions, and provide precise guidance and positioning for subsequent operations;

[0019] Use machine vision module to collect and preprocess images, obtain images of moxibustion sites and extract features of moxibustion sites;

[0020] Step 3: Establish the forward and inverse kinematics models of the robot arm;

[0021] Step 4: Integrate the heat transfer model, moxibustion site identification and robotic arm kinematics model;

[0022] The established temperature prediction model and moxibustion site identification are integrated into a reinforcement learning simulation training environment; this environment is designed to simulate and optimize the temperature control strategy during physical therapy. The reinforcement learning algorithm will adjust the operation of the infrared physical therapy device that fits the thermal effect of moxibustion based on real-time feedback from the temperature prediction model and acupoint identification system, making it more accurate and effective.

[0023] Furthermore, step 1 specifically includes:

[0024] Based on the Pennes heat transfer equation, the human body heat transfer equation was established, and the ANSYS human body heat transfer model was established to test the correctness of the model. By correcting the temperature changes of the human body under moxibustion and infrared therapy, the foundation for the next step of temperature prediction was laid:

[0025] The human body heat transfer equation is established based on the Pennes heat transfer equation:

[0026]

[0027] In the above formula, ρ represents density, C p represents heat capacity, k represents thermal conductivity;

[0028] Transfer Term Q perf =ρ b C p,b ω b (T b -T) represents the convective cooling of blood perfusion, where ρ b represents blood density, C p,b represents blood volume, T b represents blood temperature, ω b Indicates blood perfusion rate;

[0029] The steps to establish the heat transfer model include:

[0030] Step 101, collecting time series temperature data, including heat transfer model data and experimental acquisition data, covering temperature change information at different altitudes, to establish a time-temperature-altitude relationship;

[0031] Step 102: pre-process the data to ensure time sequence and continuity, and perform spatial standardization so that the model can effectively process temperature data at different heights;

[0032] When designing the model, the LSTM network is used to capture long-term temporal dependencies, while the CNN network is used to process spatial features. In the model training phase, the training set is used for joint training, and the validation set is used to evaluate the performance and generalization ability of the model. In the process of optimizing the model, the parameters and structure of the model are adjusted to improve the accuracy of temperature prediction and effectively deal with possible overfitting.

[0033] Step 103: Use the optimized model to predict the temperature at future time points and different altitudes;

[0034] Guided by the data of human body temperature changes under moxibustion, the model selects the appropriate height to achieve the effect that the infrared therapy device can fit the thermal effect of moxibustion.

[0035] Furthermore, step 2 specifically includes:

[0036] Step 201, the system uses the sensor.snapshot() method to capture the current frame of the real-time video stream of the sensor camera and saves it as an image for subsequent processing;

[0037] Step 202, the system uses the find_blobs function to accurately detect the marked area in the image based on the preset LAB color space threshold;

[0038] This process not only identifies the image area that meets the color threshold, but also provides detailed location, size and shape information, which provides an important basis for subsequent acupoint identification;

[0039] Step 203: The system inputs the captured image into YOLO and uses deep learning technology to perform target detection and positioning;

[0040] YOLO can quickly and accurately identify key parts of the human body and other targets in the image and return their precise location information; through these key points, the system can accurately define and divide different limb areas of the human body;

[0041] Step 204: After obtaining the output of find_blobs and YOLO, the system fuses the information of the two. This step quickly locks the target area based on color information, and further verifies and refines the position and range of the area in combination with the human posture estimation result of YOLO. The selected limbs are segmented and marked according to the identified limb features, and the area where the acupoints are located is clarified, thereby establishing the coordinates of the acupoints.

[0042] Step 205: The system uses the draw_point technology to draw a point diagram on the image to clearly mark the location of the detected color block.

[0043] Furthermore, step 3 specifically includes:

[0044] Step 301: Kinematic analysis:

[0045] Use inverse kinematics to control the robot arm in robotics technology. According to the predetermined position of the end effector of the robot arm, reversely analyze the angles of each joint required to achieve this position. Use the thinking method of tracing back from the result to the cause. Know the target position of the end and find out the joint configuration that needs to be adjusted to achieve this position.

[0046] The specific formula is as follows:

[0047] Planar and stereo distance calculations:

[0048] Top View Radius:

[0049]

[0050] This is the straight-line distance from the center of rotation of the robot arm to the end effector in the xy plane;

[0051] Side View Radius:

[0052]

[0053] This is the actual three-dimensional distance from the center of rotation of the robot arm to the end effector, taking into account the height, i.e., the z-coordinate;

[0054] Joint angle calculation:

[0055] Rotation Angle:

[0056]

[0057] This angle determines how much the base of the robot arm should rotate to align the end effector with the target's x and y position;

[0058] High Angle Stepper Motor:

[0059]

[0060] This angle is based on the length from the "shoulder" to the "wrist" of the robot arm, which is actually measured to be 120mm. It is calculated to calculate the angle that the robot arm shoulder needs to bend in order for the end effector to reach the specified z coordinate;

[0061] Low angle stepper motor:

[0062]

[0063] This angle determines how much the "elbow" of the robot arm needs to rotate to support the position of the "shoulder" to ensure the accurate position of the end effector;

[0064] Step 302: Path planning and interpolation:

[0065] The target position coordinates are received through the setInterpolation function. These coordinates are set as a series of points, including x, y, and z coordinates. Then, the system calculates the Euclidean distance between the starting point and the end point. The formula is as follows:

[0066]

[0067] This distance not only reflects the straight-line distance between the two points, but also serves as the basis for the subsequent calculation of the time multiplier, which is a key parameter for determining the interpolation speed. After determining the distance, the next step is to adjust the time multiplier according to the preset speed v. If the speed is not explicitly given, the system will automatically set a suitable speed value based on the distance to ensure the continuity and smoothness of the robot's movement.

[0068] The cosine interpolation function was chosen based on the smooth start and stop characteristics it provides; the specific form of the cosine interpolation is:

[0069] progress=-cos(π×t×tmul)×0.5+0.5 (37)

[0070] Where t represents the time elapsed from the start of interpolation, tmul is the time multiplier calculated based on speed and distance, and the progress value generated by this expression smoothly transitions from 0 to 1, representing the entire moving process from the starting point to the end point;

[0071] In each use of the device, according to the progress value calculated at the previous time point, the real-time position of the robot arm will be adjusted and updated according to a specific formula, as follows:

[0072] current position=start position+progress×delta (38)

[0073] Where delta is the difference between the target position and the starting position.

[0074] Furthermore, step 4 specifically includes:

[0075] Step 401, the robot arm system receives and processes the coordinate information of the acupoints through the command module. This process is the starting point of the entire mechanical operation; the acupoint coordinate information can be provided by the image processing system or directly input by the user; this information is transmitted to the robot arm control system through serial communication in G-code format; whenever a new input character is received, the system will check whether a complete command line is formed;

[0076] Step 402, the complete command line is sent to the processMessage function for parsing; the core task of the processMessage function is to convert the G-code command in text format into instructions that the system can understand and execute; it first determines the type of command, i.e., G-code or M-code; after parsing the command type, the function further extracts and parses subsequent parameters, such as target coordinates and moving speed; this step is completed by finding each parameter identifier and the value following it, ensuring that each parameter is read and set correctly;

[0077] Step 403, after receiving the coordinates, the RobotGeometry module calculates the joint angles required for the robot arm to reach these coordinates; this includes detailed angle adjustments for each joint to ensure that the robot arm can be accurately aligned to the target position; the Interpolation module generates a smooth motion path based on the calculated start and end positions;

[0078] Step 404, the movement execution is controlled by the RampsStepper module; this module receives the interpolation points from the Interpolation module and drives the corresponding motors to gradually adjust the robot arm to each predetermined position;

[0079] Step 405, the RampsStepper module sends a signal through the set stepper motor control pin to drive the joints of the robot arm to move along a predetermined path; ensure that the stepper motor is enabled, and then gradually adjust the number of steps of the motor according to the target position calculated by the interpolation module to achieve the transition from the current position to the target position; the movement direction of the motor is determined by the relative relationship between the target position and the current position, and the specific stepping action is achieved by finely controlling the direction pin and step pin of the motor.

[0080] The beneficial effects of the present invention are:

[0081] The present invention utilizes infrared radiation technology combined with acupoint recognition and robotic arm control to achieve precise moxibustion-like thermal effects; firstly, by establishing an accurate heat transfer model, we deeply understand the influence mechanism of infrared radiation on human body temperature, so as to effectively predict and control the change of surface temperature; secondly, we develop high-precision acupoint recognition technology, which can accurately locate the target area under different lighting and skin type conditions, and provide precise guidance and positioning for subsequent operations; finally, we design and implement a robotic arm control system to realize automated operation based on the acupoint recognition results, ensuring that the robotic arm can move to the predetermined position accurately and smoothly during the process, so as to realize personalized and precise use. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 It is a system diagram of infrared therapy device fitting the thermal effect of moxibustion;

[0083] Figure 2 It is a diagram of the positioning method of the infrared physiotherapy device fitting the thermal effect of moxibustion;

[0084] Figure 3 It is a step-by-step diagram for identifying moxibustion treatment sites;

[0085] Figure 4 It is the hardware block diagram of the infrared therapy device fitting the thermal effect of moxibustion;

[0086] Figure 5 It is the control flow chart of the robot arm;

[0087] Figure 6 It is a graph of moxibustion;

[0088] Figure 7 This is a schematic diagram of the thermal effect of a 100W infrared therapy device at 38cm;

[0089] Figure 8 It is a schematic diagram of the HMI user page;

[0090] Fig. 9 This is a schematic diagram of the HMI acupoint identification page;

[0091] Fig.10 This is a schematic diagram of the robot arm control page;

[0092] Fig.11 This is a schematic diagram of traditional moxibustion thermal imaging;

[0093] Fig.12 It is a thermal imaging schematic diagram of an infrared therapy device system that fits the thermal effect of moxibustion. DETAILED DESCRIPTION

[0094] Specific implementation method 1: A fully automatic infrared moxibustion therapy system based on image processing, including an image processing module, a robotic arm control module and a human-computer interaction module;

[0095] The image processing module consists of a camera and an image processing platform. The camera collects real-time images of the back of a human body placed on the working plane and transmits them to the human-machine interaction module and the robotic arm control module through a wireless serial port for subsequent processing and analysis.

[0096] The robotic arm control module is composed of a main controller, a robotic arm, a temperature sensor, a distance sensor and an infrared moxibustion terminal. The main controller receives instructions from the human-computer interaction page, and the driving module of the robotic arm receives the G code control instructions from the main controller to drive the infrared moxibustion terminal to move, thereby realizing the human body's gentle moxibustion action. While the robotic arm is running, the temperature sensor and the distance sensor respectively collect the surface temperature of the moxibustion site and the distance between the robotic arm terminal and the surface of the moxibustion site. The temperature sensor and the distance sensor upload the surface temperature and distance to the main controller, and the robotic arm terminal clamps the infrared therapy lamp for heat therapy.

[0097] The human-computer interaction module includes an interactive system with a graphical interface, which integrates functions such as real-time image loading and automatic acupoint positioning methods; the human-computer interaction page communicates with the main controller and the drive module of the robotic arm through the wireless serial port, and is used to send G-code control instructions, receive and display temperature sensor, distance sensor, and robotic arm position information.

[0098] The present invention successfully constructs a fully automatic infrared moxibustion therapy instrument by using advanced image processing technology and robot arm control algorithm, which has the ability to accurately identify and determine acupoints; this innovative measure is to solve the various challenges faced by traditional moxibustion technology, including but not limited to the formation of harmful smoke, high-intensity labor burden, lack of professional moxibustion practitioners, and low accuracy of acupoint selection based on image processing due to differences in human characteristics; the advanced intelligent moxibustion instrument not only improves the accuracy and convenience of moxibustion therapy, but can also be widely used in the elderly, presenting them with a safer and more efficient use option; this not only significantly improves the quality of life of the elderly, but also provides new possibilities for the modern promotion of traditional Chinese medicine moxibustion.

[0099] One end of the robotic arm is fixed on a mobile or fixed platform, and the other end of the robotic arm (the end of the robotic arm) is equipped with infrared moxibustion, temperature sensor, and distance sensor. The driving module of the robotic arm is used to convert control instructions into the key task of mechanical movement, including a stepper motor and a driver chip. The driver chip communicates with the human-computer interaction page, and the driver chip realizes the posture adjustment of the robotic arm through the stepper motor.

[0100] The robot arm is a three-degree-of-freedom robot arm; the stepper motor adopts a 42-type stepper motor, which has stable performance and excellent control accuracy, and provides precise motion control. The stepper motor converts electrical signals into angles or linear displacements, has fast response capabilities, can start or stop quickly, and also shows good adaptability in switching the direction and speed of operation. Its self-locking characteristics are particularly in line with the requirements of long-term stable operation of moxibustion. The proportional linear relationship between the pulse signal and the step angle ensures that no matter how the load changes, the movement of the stepper motor can maintain a high degree of accuracy, and changing the order of the pulse signal can simply achieve steering adjustment; the drive core The chip uses the A4988 driver chip, which supports multiple modes including but not limited to forward and reverse, sleep and reset. By adjusting the levels of the MS1, MS2 and MS3 pins, it can achieve a micro-step drive mode from full step to one-sixth step, thereby achieving the required motion accuracy and smoothness. The A4988 chip not only meets the precise control requirements of motor motion, but also has a variety of safety protection mechanisms, such as over-current protection, thermal shutdown and logic level under-voltage protection. These functions ensure the safety and stability of the system under various operating conditions. The A4988 driver module ensures that the moxibustion robot arm can move accurately and smoothly during use. The robot moves to the designated acupuncture points on the back for precise moxibustion. At the same time, this drive design ensures economic benefits and system reliability. The distance sensor uses a ToF laser ranging sensor based on the VL53LOX chip. It is small in size, low in power consumption, and can maintain stable accuracy under a variety of reflectivity conditions. It is very suitable for controlling the distance during moxibustion. It sends energy beams and detects their reflected waves. It estimates the distance to the target by calculating the round-trip time. It is responsible for providing accurate information on the distance between the robotic arm and the moxibustion site. The temperature sensor is responsible for monitoring the epidermal temperature during the process to ensure safe and effective results. The temperature sensor uses MLX9 The 0614ESF-DCC infrared temperature sensor module has the advantages of small size, integrated low-noise amplifier, high-precision ADC and DSP unit, and precise calibration specifically for medical scenarios in terms of non-contact skin temperature monitoring. This design enables it to directly provide calibrated temperature data, greatly simplifying the operation process. The sensor's built-in 81101 thermoelectric element is responsible for temperature detection. Through the precise voltage output formula, it can convert the temperature difference between the target object and the sensor itself into a clear digital output, making temperature monitoring accurate and convenient. The main controller uses the Arduino Mega 2560 control module to ensure the coordinated work of the system. It connects 42 stepper motors with the corresponding drive device A4988. These stepper motors are responsible for ensuring the precise positioning of the robotic arm, and then accurately moxibustion the designated back acupuncture points. In addition, with the help of the integrated sensor module, the system can monitor the distance to the surface and the surface temperature in real time, thereby ensuring the safety and effectiveness of the solution.In terms of hardware settings, the VL53LOX laser distance measurement sensor and the MLX90614 infrared temperature sensor provide the system with key environmental feedback information, and these data are directly transmitted to the main control device via the I2C bus; in addition, the system also integrates key power management components to ensure that all components operate in a stable and safe voltage environment; infrared moxibustion uses Philips R95 100W infrared lamps, and the testo 869 infrared thermal imager is used to perform thermal imaging analysis of traditional moxibustion and Philips R95 100W infrared lamps, such as; Figure 11-12 As shown, it is just to verify that the two have similarities in hot spot distribution. Traditional moxibustion is shown as local very high temperature hot spots in thermal imaging, and these hot spots directly correspond to the positions where the moxa sticks are placed; while the back irradiated with Philips R95 100W infrared lamp presents a wider and more uniform heat distribution; in terms of temperature range, the temperature near the traditional moxibustion point is very high, which can reach 39.7℃ to 46.9℃, showing the high temperature characteristics of moxibustion, and the heat distribution is uneven; while the temperature range of the Philips R95100W infrared lamp irradiation area is from 38.4℃ to 45.7℃, providing a mild and stable thermal effect, and the heat provided is evenly distributed, reducing the risk of local overheating, suitable for long-term use, and has strong deep penetration ability; the present invention shows superiority in maintaining temperature stability and uniformity of heat distribution in the moxibustion site. The device is not only safer and more comfortable, but also has significantly improved durability, which is very suitable for long-term use by users.

[0101] Specific implementation method 2: Figures 1 to 12 As shown, a fully automatic infrared moxibustion therapy positioning method based on image processing specifically includes:

[0102] Step 1: Establish a heat transfer model to predict surface temperature changes under infrared therapy conditions;

[0103] First, the time, distance and temperature data under different conditions are collected and recorded through cameras, temperature sensors, distance sensors and HMI. Then, the heat transfer model is corrected and optimized based on the actual collected data to ensure that the simulation results are as close and accurate as possible to the actual situation.

[0104] Secondly, a long short-term memory network-convolutional neural network deep learning model is used to train and learn the time, distance and temperature sequence data obtained from the thermal field simulation model; the goal of this step is to establish a model that can accurately predict the surface temperature changes during physical therapy, providing a basis for subsequent physical therapy control strategies;

[0105] Step 2: Identify the moxibustion site, accurately locate the moxibustion site under different lighting and surface type conditions, and provide precise guidance and positioning for subsequent operations;

[0106] Use machine vision module to collect and preprocess images, obtain images of moxibustion sites and extract features of moxibustion sites;

[0107] Aiming at the problem of acupuncture point identification and positioning on the back of the human body, a high-definition camera is used to capture and track the surface image of the back, and then 43 acupuncture points on the back are accurately positioned according to a specific acupuncture point coordinate table; through real-time image analysis, it has the function of automatically marking the location of acupuncture points. Its back tracking and identification accuracy rate is as high as 96%, and the accuracy rate of acupuncture point positioning technology is 97%, which fully meets the use requirements;

[0108] Step 3: Establish the forward and inverse kinematics models of the robot arm;

[0109] Based on the features of step 2, a supervised learning model is established to train the system to accurately identify and locate different moxibustion treatment sites in real time;

[0110] The robotic arm control system uses kinematic analysis technology to convert the two-dimensional coordinates provided by the image processing system into the three-dimensional space of the robotic arm. The system uses a high-precision stepper motor to drive the robotic arm, and uses cosine interpolation technology to accurately plan the path of the end action of the robotic arm and accurately push it to the designated acupuncture point. Through these means, the accuracy of acupuncture point tracking is as high as 97%, which fully meets the expected requirements of the design. The integration of temperature sensors and distance sensors can monitor the treatment environment in real time, thereby ensuring the safety of users.

[0111] Step 4: Integrate the heat transfer model, moxibustion site identification and robotic arm kinematics model;

[0112] The established temperature prediction model and moxibustion site identification are integrated into a reinforcement learning simulation training environment; this environment is designed to simulate and optimize the temperature control strategy during physical therapy. The reinforcement learning algorithm will adjust the operation of the infrared physical therapy device that fits the moxibustion thermal effect based on real-time feedback from the temperature prediction model and acupoint identification system, making it more accurate and effective.

[0113] Finally, the trained reinforcement learning model is deployed on the actual hardware platform of the robotic arm of the infrared physiotherapy device that fits the moxibustion thermal effect to verify the hardware platform. This step will verify the practicality and superiority of the algorithm in actual operation.

[0114] The present invention performs well in the identification and precise positioning of acupoints or other acupuncture points, with an error limited to within 1.5 cm, thus ensuring its accuracy in clinical applications; this innovative medical device successfully combines image processing technology, robotic arm control and moxibustion therapy, providing users with a more comfortable method of use.

[0115] This implementation method transforms traditional dimensional method into modern coordinate calculation, uses a high-definition camera to capture human body images, and uses image processing software to identify the main characteristic areas of the human body, such as the spine line and shoulder peak on the surface; this process involves converting traditional measurement units into precise pixel values ​​so that acupuncture points can be accurately marked in digital images;

[0116] In the modern medical environment, digital technology equipment is used for precise operation. According to the National Standard of the People's Republic of China GB-12346-90 Acupuncture Points and related image processing technologies, a set of calculation schemes for acupoint coordinates using digital measurement methods is designed, with the goal of improving the positioning accuracy of acupoints and converting traditional "inches" and "minutes" into millimeters or pixels. The width and height of the identified color block are divided into 16 equal parts, each representing the basic distance unit between acupoints.

[0117] Then we can get the Y of the acupoint 穴位 The coordinate calculation formula is as follows:

[0118]

[0119] Among them, Y 基准 is the Y coordinate of the regional feature value point, n is the number of inches down or up, converted into the corresponding pixel value;

[0120] For the side acupuncture points, the calculation formula for the X coordinate is:

[0121]

[0122] Among them, X 基准 is the X coordinate of the regional feature value point, m is the number of inches away from the center, and the positive or negative value depends on the position of the acupoint relative to the key point.

[0123] Specific implementation method three: Figures 1 to 12 As shown, step 1 specifically includes:

[0124] Based on the Pennes heat transfer equation, the human body heat transfer equation was established, and the ANSYS human body heat transfer model was established to test the correctness of the model. By correcting the temperature changes of the human body under moxibustion and infrared therapy, the foundation for the next step of temperature prediction was laid:

[0125] The human body heat transfer equation is established based on the Pennes heat transfer equation:

[0126]

[0127] In the above formula, ρ represents density, C p represents heat capacity, k represents thermal conductivity;

[0128] Transfer Term Qperf =ρ b C p,b ω b (T b -T) represents the convective cooling of blood perfusion, where ρ b represents blood density, C p,b represents blood volume, T b represents blood temperature, ω b Indicates blood perfusion rate;

[0129] First, the Pennes biological heat transfer equation is used, which describes the basic process of temperature distribution inside biological tissues and takes into account the effect of blood on temperature transfer;

[0130] Secondly, the process of using ANSYS to establish a human body heat transfer model includes geometric modeling, setting boundary conditions, defining material properties, meshing and numerical solution; in the geometric modeling stage, select the specific part to be simulated, such as arms or legs, and ensure the accuracy and realism of the model; when setting boundary conditions, it is necessary to consider factors such as surface temperature, radiation conditions and possible heat input; define the physical properties of the tissue such as density, specific heat capacity and thermal conductivity, which directly affect the thermal conduction behavior of the model; then meshing is performed to ensure the accuracy and stability of the simulation results, and finally the thermal conduction module of ANSYS is used to numerically solve the Pennes equation to obtain the temperature distribution in time and space;

[0131] The process of model verification includes comparing the simulation results with the experimental data, especially verifying the temperature changes under moxibustion and infrared therapy. If the simulation results do not match the experimental data, the simulation accuracy can be optimized by adjusting the model parameters, such as adjusting the thermal conductivity of the tissue or the blood flow rate. Finally, the simulation parameters or working mode of the infrared therapy device are adjusted according to the verification results to ensure that it can accurately simulate the temperature changes of the human body during moxibustion. The accuracy and applicability of the simulation are evaluated by comparing the corrected simulation results with the actual temperature change data during moxibustion.

[0132] The steps to establish the heat transfer model include:

[0133] Step 101, collecting time series temperature data, including heat transfer model data and experimental acquisition data, covering temperature change information at different altitudes, to establish a time-temperature-altitude relationship;

[0134] Step 102: pre-process the data to ensure time sequence and continuity, and perform spatial standardization so that the model can effectively process temperature data at different heights;

[0135] When designing the model, the LSTM network is used to capture long-term temporal dependencies, while the CNN network is used to process spatial features. In the model training phase, the training set is used for joint training, and the validation set is used to evaluate the performance and generalization ability of the model. In the process of optimizing the model, the parameters and structure of the model are adjusted to improve the accuracy of temperature prediction and effectively deal with possible overfitting.

[0136] Step 103: Use the optimized model to predict the temperature at future time points and different altitudes;

[0137] Guided by the data of human body temperature changes under moxibustion, the model selects the appropriate height to achieve the effect that the infrared therapy device can fit the thermal effect of moxibustion.

[0138] Specific implementation method four: Figures 1 to 12 As shown, step 2 specifically includes:

[0139] Step 201, the system uses the sensor.snapshot() method to capture the current frame of the real-time video stream of the sensor camera and saves it as an image for subsequent processing;

[0140] Step 202, the system uses the find_blobs function to accurately detect the marked area in the image based on the preset LAB color space threshold;

[0141] This process not only identifies the image area that meets the color threshold, but also provides detailed location, size and shape information, which provides an important basis for subsequent acupoint identification;

[0142] Step 203: The system inputs the captured image into YOLO and uses deep learning technology to perform target detection and positioning;

[0143] YOLO can quickly and accurately identify key parts of the human body and other targets in the image and return their precise location information; through these key points, the system can accurately define and divide different limb areas of the human body;

[0144] Step 204: After obtaining the output of find_blobs and YOLO, the system fuses the information of the two. This step quickly locks the target area based on color information, and further verifies and refines the position and range of the area in combination with the human posture estimation result of YOLO. The selected limbs are segmented and marked according to the identified limb features, and the area where the acupoints are located is clarified, thereby establishing the coordinates of the acupoints.

[0145] Step 205: The system uses the draw_point technology to draw a point diagram on the image to clearly mark the location of the detected color block.

[0146] Among them, after capturing the real-time image, the system uses the find_blobs function (find color block function) to detect the specific color area; the input parameter of this function is a color threshold, which is set to green_threshold (green threshold) here to identify the dark green marked area; the threshold setting is based on the LAB color space, where L represents brightness, and A and B represent color opposition dimensions; the specific thresholds (47, 94, -19, 6, -8, 11) are adjusted according to the lighting conditions and background colors in the experimental environment, so that the color consistency and recognition accuracy can be maintained under changing lighting conditions, so as to ensure that the background and mark colors can be effectively distinguished;

[0147] The find_blobs function not only returns the image area that meets the color threshold, but also provides detailed area attributes such as location, size, and shape, which are crucial for subsequent acupoint identification;

[0148] Compared with other techniques such as color histogram matching, the find_blobs method shows a higher judgment accuracy when processing scenes with clear color markings; although color histogram matching has some advantages in processing complex color distributions, it is not as real-time as the threshold-based strategy; especially when the color of the marked area is uniform and clear, using find_blobs can quickly lock the target area, reduce the consumption of computing resources, and simplify the processing flow; in order to more clearly reflect the color threshold advantage of LAB, Table 1 presents an in-depth comparative analysis;

[0149] Table 1 Comparison of color block detection methods

[0150]

[0151] Once the system successfully identifies the dark green color area, it will enter the acupuncture point identification and marking stage; at the same time, the system will input the captured image into YOLO, using deep learning technology for target detection and positioning. YOLO (You Only Look Once) can quickly and accurately identify key parts of the human body and other targets in the image and return their precise location information. Through these key points, the system can accurately define and divide different limb areas of the human body, such as arms, legs, etc.

[0152] After obtaining the output of find_blobs and YOLO, the system fuses the information of the two. This step quickly locks the target area based on color information, and further verifies and refines the location and range of the area in combination with the human posture estimation results of YOLO. The selected limbs are segmented and marked according to the identified limb features, and the area where the acupoints are located is clarified, thereby establishing the coordinates of the acupoints.

[0153] Finally, the system uses the draw_point technique to draw a dot map on the image, clearly marking the location of the detected color blocks, so that users or subsequent processing can intuitively view and analyze the recognition results.

[0154] Specific implementation method five: Figures 1 to 12 As shown, step 3 specifically includes:

[0155] Step 301: Kinematic analysis:

[0156] Use inverse kinematics to control the robot arm in robotics technology. According to the predetermined position of the end effector of the robot arm, reversely analyze the angles of each joint required to achieve this position. Use the thinking method of tracing back from the result to the cause. Know the target position of the end and find out the joint configuration that needs to be adjusted to achieve this position.

[0157] The specific formula is as follows:

[0158] Planar and stereo distance calculations:

[0159] Top View Radius:

[0160]

[0161] This is the straight-line distance from the center of rotation of the robot arm to the end effector in the xy plane;

[0162] Side View Radius:

[0163]

[0164] This is the actual three-dimensional distance from the center of rotation of the robot arm to the end effector, taking into account the height, i.e., the z-coordinate;

[0165] Joint angle calculation:

[0166] Rotation Angle:

[0167]

[0168] This angle determines how much the base of the robot arm should rotate to align the end effector with the target's x and y position;

[0169] High Angle Stepper Motor:

[0170]

[0171] This angle is based on the length from the "shoulder" to the "wrist" of the robot arm, which is actually measured to be 120mm. It is calculated to calculate the angle that the robot arm shoulder needs to bend in order for the end effector to reach the specified z coordinate;

[0172] Low angle stepper motor:

[0173]

[0174] This angle determines how much the "elbow" of the robot arm needs to rotate to support the position of the "shoulder" to ensure the accurate position of the end effector;

[0175] Step 302: Path planning and interpolation:

[0176] The target position coordinates are received through the setInterpolation function. These coordinates are set as a series of points, including x, y, and z coordinates. Then, the system calculates the Euclidean distance between the starting point and the end point. The formula is as follows:

[0177]

[0178] This distance not only reflects the straight-line distance between the two points, but also serves as the basis for the subsequent calculation of the time multiplier, which is a key parameter for determining the interpolation speed. After determining the distance, the next step is to adjust the time multiplier according to the preset speed v. If the speed is not explicitly given, the system will automatically set a suitable speed value based on the distance to ensure the continuity and smoothness of the robot's movement.

[0179] The cosine interpolation function was chosen based on the smooth start and stop characteristics it provides; the specific form of the cosine interpolation is:

[0180] progress=-cos(π×t×tmul)×0.5+0.5 (37)

[0181] Where t represents the time elapsed from the start of interpolation, tmul is the time multiplier calculated based on speed and distance, and the progress value generated by this expression smoothly transitions from 0 to 1, representing the entire moving process from the starting point to the end point;

[0182] In each use of the device, according to the progress value calculated at the previous time point, the real-time position of the robot arm will be adjusted and updated according to a specific formula, as follows:

[0183] current position=start position+progress×delta (38)

[0184] Where delta is the difference between the target position and the starting position.

[0185] This method ensures that the robot moves smoothly along the predetermined path, and also makes the entire control system well adaptable to various dynamic changes; by implementing this cosine interpolation-based algorithm, the motion trajectory of the robot is optimized.

[0186] Specific implementation method six: Figures 1 to 12 As shown, step 4 specifically includes:

[0187] Step 401, the robot arm system receives and processes the coordinate information of the acupoints through the command module. This process is the starting point of the entire mechanical operation; the acupoint coordinate information can be provided by the image processing system or directly input by the user; this information is transmitted to the robot arm control system through serial communication in G-code format; whenever a new input character is received, the system will check whether a complete command line is formed;

[0188] Step 402, the complete command line is sent to the processMessage function for parsing; the core task of the processMessage function is to convert the G-code command in text format into instructions that the system can understand and execute; it first determines the type of command, i.e., G-code or M-code; after parsing the command type, the function further extracts and parses subsequent parameters, such as target coordinates and moving speed; this step is completed by finding each parameter identifier and the value following it, ensuring that each parameter is read and set correctly;

[0189] Step 403, after receiving the coordinates, the RobotGeometry module calculates the joint angles required for the robot arm to reach these coordinates; this includes detailed angle adjustments for each joint to ensure that the robot arm can be accurately aligned to the target position; the Interpolation module generates a smooth motion path based on the calculated start and end positions;

[0190] Step 404, the movement execution is controlled by the RampsStepper module; this module receives the interpolation points from the Interpolation module and drives the corresponding motors to gradually adjust the robot arm to each predetermined position;

[0191] Step 405, the RampsStepper module sends a signal through the set stepper motor control pin to drive the joints of the robot arm to move along a predetermined path; ensure that the stepper motor is enabled, and then gradually adjust the number of steps of the motor according to the target position calculated by the interpolation module to achieve the transition from the current position to the target position; the movement direction of the motor is determined by the relative relationship between the target position and the current position, and the specific stepping action is achieved by finely controlling the direction pin and step pin of the motor.

[0192] These control strategies enable the robot arm to locate acupuncture points in an efficient and accurate manner while also ensuring the safety of its operation process; Figure 5 As shown, the control flow block diagram of the robotic arm is shown.

[0193] The above is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this profession can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement made to the above embodiments without departing from the content of the technical solution of the present invention, based on the technical essence of the present invention, within the spirit and principles of the present invention, still fall within the protection scope of the technical solution of the present invention.

Claims

1. A fully automatic infrared moxibustion therapy system based on image processing, characterized in that: It includes image processing module, robotic arm control module and human-computer interaction module; The image processing module consists of a camera and an image processing platform. The camera collects real-time images of the back of a human body placed on the working plane and transmits them to the human-machine interaction module and the robotic arm control module through a wireless serial port for subsequent processing and analysis. The robotic arm control module is composed of a main controller, a robotic arm, a temperature sensor, a distance sensor and an infrared moxibustion terminal. The main controller receives instructions from the human-computer interaction page, and the driving module of the robotic arm receives the G code control instructions from the main controller to drive the infrared moxibustion terminal to move, thereby realizing the human body's gentle moxibustion action. While the robotic arm is running, the temperature sensor and the distance sensor respectively collect the surface temperature of the moxibustion site and the distance between the robotic arm terminal and the surface of the moxibustion site. The temperature sensor and the distance sensor upload the surface temperature and distance to the main controller, and the robotic arm terminal clamps the infrared therapy lamp for heat therapy. The human-computer interaction module includes an interactive system with a graphical interface, which integrates functions such as real-time image loading and automatic acupoint positioning methods; the human-computer interaction page communicates with the main controller and the drive module of the robotic arm through the wireless serial port, and is used to send G-code control instructions, receive and display temperature sensor, distance sensor, and robotic arm position information.

2. A fully automatic infrared moxibustion therapy positioning method based on image processing, characterized in that: Specifically include: Step 1: Establish a heat transfer model to predict surface temperature changes under infrared therapy conditions; First, the time, distance and temperature data under different conditions are collected and recorded through cameras, temperature sensors, distance sensors and HMI. Then, the heat transfer model is corrected and optimized based on the actual collected data to ensure that the simulation results are as close and accurate as possible to the actual situation. Secondly, a long short-term memory network-convolutional neural network deep learning model is used to train and learn the time, distance and temperature sequence data obtained from the thermal field simulation model; the goal of this step is to establish a model that can accurately predict the surface temperature changes during physical therapy, providing a basis for subsequent physical therapy control strategies; Step 2: Identify the moxibustion site, accurately locate the moxibustion site under different lighting and surface type conditions, and provide precise guidance and positioning for subsequent operations; Use machine vision module to collect and preprocess images, obtain images of moxibustion sites and extract features of moxibustion sites; Step 3: Establish the forward and inverse kinematics models of the robot arm; Step 4: Integrate the heat transfer model, moxibustion site identification and robotic arm kinematics model; The established temperature prediction model and moxibustion site identification are integrated into a reinforcement learning simulation training environment; this environment is designed to simulate and optimize the temperature control strategy during physical therapy. The reinforcement learning algorithm will adjust the operation of the infrared physical therapy device that fits the thermal effect of moxibustion based on real-time feedback from the temperature prediction model and acupoint identification system, making it more accurate and effective.

3. The fully automatic infrared moxibustion therapy apparatus based on image processing according to claim 2 is characterized in that: Step 1 specifically includes: Based on the Pennes heat transfer equation, the human body heat transfer equation was established, and the ANSYS human body heat transfer model was established to test the correctness of the model. By correcting the temperature changes of the human body under moxibustion and infrared therapy, the foundation for the next step of temperature prediction was laid: The human body heat transfer equation is established based on the Pennes heat transfer equation: In the above formula, ρ represents density, C p represents heat capacity, k represents thermal conductivity; Transfer Term Q perf =ρ b C p,b ω b (T b -T) represents the convective cooling of blood perfusion, where ρ b represents blood density, C p,b represents blood volume, T b represents blood temperature, ω b Indicates blood perfusion rate; The steps to establish the heat transfer model include: Step 101, collecting time series temperature data, including heat transfer model data and experimental acquisition data, covering temperature change information at different altitudes, to establish a time-temperature-altitude relationship; Step 102: pre-process the data to ensure time sequence and continuity, and perform spatial standardization so that the model can effectively process temperature data at different heights; When designing the model, the LSTM network is used to capture long-term temporal dependencies, while the CNN network is used to process spatial features. In the model training phase, the training set is used for joint training, and the validation set is used to evaluate the performance and generalization ability of the model. In the process of optimizing the model, the parameters and structure of the model are adjusted to improve the accuracy of temperature prediction and effectively deal with possible overfitting. Step 103: Use the optimized model to predict the temperature at future time points and different altitudes; Guided by the data of human body temperature changes under moxibustion, the model selects the appropriate height to achieve the effect that the infrared therapy device can fit the thermal effect of moxibustion.

4. The fully automatic infrared moxibustion therapy apparatus based on image processing according to claim 2, characterized in that: Step 2 specifically includes: Step 201, the system uses the sensor.snapshot() method to capture the current frame of the real-time video stream of the sensor camera and saves it as an image for subsequent processing; Step 202, the system uses the find_blobs function to accurately detect the marked area in the image based on the preset LAB color space threshold; This process not only identifies the image area that meets the color threshold, but also provides detailed location, size and shape information, which provides an important basis for subsequent acupoint identification; Step 203: The system inputs the captured image into YOLO and uses deep learning technology to perform target detection and positioning; YOLO can quickly and accurately identify key parts of the human body and other targets in the image and return their precise location information; through these key points, the system can accurately define and divide different limb areas of the human body; Step 204: After obtaining the output of find_blobs and YOLO, the system fuses the information of the two. This step quickly locks the target area based on color information, and further verifies and refines the position and range of the area in combination with the human posture estimation result of YOLO. The selected limbs are segmented and marked according to the identified limb features, and the area where the acupoints are located is clarified, thereby establishing the coordinates of the acupoints. Step 205: The system uses the draw_point technology to draw a point diagram on the image to clearly mark the location of the detected color block.

5. The fully automatic infrared moxibustion therapy apparatus based on image processing according to claim 2, characterized in that: Step 3 specifically includes: Step 301: Kinematic analysis: Use inverse kinematics to control the robot arm in robotics technology. According to the predetermined position of the end effector of the robot arm, reversely analyze the angles of each joint required to achieve this position. Use the thinking method of tracing back from the result to the cause. Know the target position of the end and find out the joint configuration that needs to be adjusted to achieve this position. The specific formula is as follows: Planar and stereo distance calculations: Top View Radius: This is the straight-line distance from the center of rotation of the robot arm to the end effector in the xy plane; Side View Radius: This is the actual three-dimensional distance from the center of rotation of the robot arm to the end effector, taking into account the height, i.e., the z-coordinate; Joint angle calculation: Rotation Angle: This angle determines how much the base of the robot arm should rotate to align the end effector with the target's x and y position; High Angle Stepper Motor: This angle is based on the length from the "shoulder" to the "wrist" of the robot arm, which is actually measured to be 120mm. It is calculated to calculate the angle that the robot arm shoulder needs to bend in order for the end effector to reach the specified z coordinate; Low angle stepper motor: This angle determines how much the "elbow" of the robot arm needs to rotate to support the position of the "shoulder" to ensure the accurate position of the end effector; Step 302: Path planning and interpolation: The target position coordinates are received through the setInterpolation function. These coordinates are set as a series of points, including x, y, and z coordinates. Then, the system calculates the Euclidean distance between the starting point and the end point. The formula is as follows: This distance not only reflects the straight-line distance between the two points, but also serves as the basis for the subsequent calculation of the time multiplier, which is a key parameter for determining the interpolation speed. After determining the distance, the next step is to adjust the time multiplier according to the preset speed v. If the speed is not explicitly given, the system will automatically set a suitable speed value based on the distance to ensure the continuity and smoothness of the robot's movement. The cosine interpolation function was chosen based on the smooth start and stop characteristics it provides; the specific form of the cosine interpolation is: progress=-cos(π×t×tmul)×0.5+0.5 (37) Where t represents the time elapsed from the start of interpolation, tmul is the time multiplier calculated based on speed and distance, and the progress value generated by this expression smoothly transitions from 0 to 1, representing the entire moving process from the starting point to the end point; In each use of the device, according to the progress value calculated at the previous time point, the real-time position of the robot arm will be adjusted and updated according to a specific formula, as follows: current position = start position + progress × delta (38) where delta is the difference between the target position and the start position.

6. The fully automatic infrared moxibustion therapy apparatus based on image processing according to claim 2, characterized in that: Step 4 specifically includes: Step 401, the robot arm system receives and processes the coordinate information of the acupoints through the command module. This process is the starting point of the entire mechanical operation; the acupoint coordinate information can be provided by the image processing system or directly input by the user; this information is transmitted to the robot arm control system through serial communication in G-code format; whenever a new input character is received, the system will check whether a complete command line is formed; Step 402, the complete command line is sent to the processMessage function for parsing; the core task of the processMessage function is to convert the G-code command in text format into instructions that the system can understand and execute; it first determines the type of command, i.e., G-code or M-code; after parsing the command type, the function further extracts and parses subsequent parameters, such as target coordinates and moving speed; this step is completed by finding each parameter identifier and the value following it, ensuring that each parameter is read and set correctly; Step 403, after receiving the coordinates, the RobotGeometry module calculates the joint angles required for the robot arm to reach these coordinates; this includes detailed angle adjustments for each joint to ensure that the robot arm can be accurately aligned to the target position; The Interpolation module generates a smooth motion path based on the calculated start and end positions; Step 404, the movement execution is controlled by the RampsStepper module; this module receives the interpolation points from the Interpolation module and drives the corresponding motors to gradually adjust the robot arm to each predetermined position; Step 405, the RampsStepper module sends a signal through the set stepper motor control pin to drive the joints of the robot arm to move along a predetermined path; ensure that the stepper motor is enabled, and then gradually adjust the number of steps of the motor according to the target position calculated by the interpolation module to achieve the transition from the current position to the target position; the movement direction of the motor is determined by the relative relationship between the target position and the current position, and the specific stepping action is achieved by finely controlling the direction pin and step pin of the motor.

Images