Dynamic interaction system, interaction execution mechanism and intelligent equipment

Abstract

The invention discloses a dynamic interaction system, an interaction execution mechanism and intelligent equipment, and belongs to the technical field of dynamic interaction.The dynamic interaction system comprises the steps that user operation requirements are received, data are collected through a basic platform pressure sensor to generate a matrix, pressure stability is analyzed through a time window, data are processed only in the stable period, and a contact body contour is extracted; and synchronously judging a placement state and generating a calibration signal. The extracted contour is matched with a standard contour feature point, and a precise operation point coordinate and a control instruction are generated in combination with a geometrical relationship and elastic transformation. In the execution link, the execution unit is driven to operate the contact body, and the adjustable unit dynamically compensates the low-pressure area and the supporting area to ensure uniform stress and stability. And the interaction feedback module guides, adjusts and visualizes the state through a multi-modal interface. The scheme improves identification precision and operation stability through dynamic pressure analysis, intelligent contour verification and the like, is suitable for medical, industrial and other scenes, and provides an efficient intelligent solution.

Description

Technical Field

[0001] The present invention relates to the field of dynamic interaction technologies, and more particularly to a dynamic interaction system, an interaction execution mechanism, and an intelligent device. Background Art

[0002] In fields such as industrial manufacturing, medical assistance, and smart home, it is often necessary to perform positioning, fixing, and orienting operations on contact bodies such as workpieces and objects to be processed placed on an operation platform. In the prior art, relying on a fixed coordinate system or a preset model, it is required that the contact body be placed at a specific angular position, which is difficult to adapt to the differences in the size and shape of the contact body in practice and the randomness of user placement, resulting in large initial positioning deviations and cumbersome manual calibration, and insufficient operation efficiency and adaptability. In terms of fixing the contact body, it mostly relies on rigid jigs or fixed groove mechanical limits, lacking dynamic adaptation to the contour and stress state of the contact body, and easily causing deformation and damage of the contact body due to uneven pressure. For example, it may leave indentations on the surface of precision workpieces or cause wrinkles in flexible objects. At the same time, for contact bodies with different weights and hardnesses, the supporting force cannot be dynamically adjusted according to the real-time stress situation, and there is a risk of insufficient fixation causing displacement or pressure overload damaging the contact body. The operation parameter needs to be manually preset or rely on a fixed trajectory. When the shape of the contact body changes or the position shifts, it is difficult to automatically correct the target action point, and there are problems such as deviation of the action point and improper force, and the operation accuracy is limited in complex scenarios. In human-computer interaction, the system's feedback on the state of the contact body is simple, and the user needs to repeatedly try and error to adjust the placement position or parameters based on experience, lacking intelligent guidance and personalized adaptation capabilities. The deficiencies of the prior art in dynamic positioning, flexible fixing, precise operation, and interaction adaptation lead to the difficulty in meeting the requirements of operation efficiency and reliability in complex scenarios. Therefore, in order to overcome these limitations, the present invention proposes a dynamic interaction system, an interaction execution mechanism, and an intelligent device. Summary of the Invention

[0003] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide a dynamic interaction system, an interaction execution mechanism, and an intelligent device. Aiming at the problems of positioning deviation, uneven support, and inaccurate operation existing in the contact body during the dynamic interaction process, through dynamic analysis of the pressure matrix, multi-level contour verification, elastic coordinate correction, and differential support control, the precise positioning, stable support, and adaptive operation of the contact body are realized, and the reliability and interaction accuracy of the system in multiple scenarios such as medical rehabilitation and industrial automation are improved.

[0004] To achieve the above purpose, the present invention provides the following technical solutions:

[0005] A dynamic interaction system, comprising:

[0006] Receiving the operation requirement parameters and the contact body type identifier input by the user;

[0007] Collect the pressure distribution data of the contact surface between the acquisition contact body and the base platform to generate a pressure matrix. Determine whether the pressure matrix is in a stable period by the change of the pressure matrix within the time window. If it is in the stable period, perform fusion, feature enhancement, and binarization processing on the pressure matrix to divide, screen, and merge the connected regions. Extract the contour based on the foreground pixel points within the connected regions and calculate the similarity with the standard contour for contour verification, so as to output the contact body contour; at the same time, judge the placement state of the contact body through multiple thresholds and generate a position calibration signal.

[0008] Receive the position calibration signal, and output operation guidance and exception prompts to the user.

[0009] By extracting the contour feature points of the contact body contour and the standard contour, match the contact body contour and the standard contour. Based on the operation point information of the standard contour, deduce and correct the candidate coordinates of the operation points of the contact body contour through trigonometric function relationships to obtain the operation point coordinates of the contact body contour. Based on the operation point coordinates and the contact body contour, generate an execution unit control instruction and an adjustable unit control instruction.

[0010] Receive and parse the execution unit control instruction and the adjustable unit control instruction, drive the execution unit and the adjustable unit, and realize the operation and support fixation of the contact body.

[0011] Specifically, the specific steps of contact body contour extraction include:

[0012] Perform weighted average fusion on the pressure matrix during the stable period to obtain the target pressure matrix. After performing noise reduction processing and feature enhancement processing on the target pressure matrix, divide each pixel point in the target pressure matrix into foreground and background categories through an adaptive threshold, and convert the target pressure matrix into a binary image.

[0013] Configure the connected domain threshold, traverse the foreground pixels in the binary image, divide the connected regions, and extract the connected region features.

[0014] Judge whether the number of divided connected regions is greater than the connected domain threshold. If it is greater, screen the valid connected regions and perform connected region merging so that the number of connected regions is equal to the connected domain threshold; otherwise, judge whether the number of divided connected regions is less than the connected domain threshold. If it is less, generate a contact body placement signal; otherwise, extract the contour for each foreground pixel point within the connected region.

[0015] The specific steps of screening the valid connected regions and performing connected region merging include:

[0016] Calculate the distances from the centroid of each connected region to the centroids of the remaining connected regions, and combine with the areas of the connected regions to screen out the effective connected regions; successively merge the connected regions based on the overlapping areas of the circumscribed rectangles of the effective connected regions, the distances between the centroids, and the topological relationships until the number of connected regions is equal to the connected region threshold.

[0017] Specifically, the specific steps for extracting the contact body contour further include:

[0018] Configure a similarity threshold. After smoothing the contour extracted from the connected region, calculate the similarity between the contour within each connected region and the standard contour. If it is greater than the similarity threshold, the contour verification is successful, mark it as a valid contour, and output the contact body contour of the connected region that passes the contour verification.

[0019] Otherwise, the contour verification fails, start the contour correction operation, re-collect the pressure distribution data, extract the dynamic contour, and then perform the contour verification, and count the number of contour verification times; configure a verification threshold. If the number of contour verification times is greater than the verification threshold, generate a contour anomaly signal.

[0020] Specifically, the specific steps for obtaining the operating point coordinates of the contact body contour include:

[0021] Establish a local coordinate system for the contact body contour.

[0022] Extract the contour feature points of the contact body contour, and obtain the contact body operating points, contour feature points under the standard contour, and the supplementary attribute information of each operating point.

[0023] Align the contact body contour with the standard contour, and preliminarily match the contour feature points of the standard contour and the contact body contour according to the principle of the closest distance or the principle of contour feature point similarity.

[0024] Specifically, the specific steps for obtaining the operating point coordinates of the contact body contour further include:

[0025] According to the relative distance and angular relationship between the operating points and the contour feature points in the standard contour, and the coordinates of the preliminarily matched contour feature points in the contact body contour, deduce the candidate coordinates of the operating points of the contact body contour through trigonometric function relationships.

[0026] Calculate the relative distance difference and angular difference between the matching feature points in the contact body contour and the standard contour, and the shape descriptor difference between the region where the candidate coordinates of the operating point are located and the region corresponding to the operating point in the standard contour, and correct the candidate coordinates of the operating point through an elastic transformation model; output the operating point coordinates of each contour of the contact body.

[0027] Specifically, the specific steps for generating the control instructions for the execution unit include:

[0028] Receive the set of target operation points, the execution priority of the target operation points, the operation requirement parameters, and the operation body type identifier;

[0029] Convert the target operation point coordinates from the local coordinate system of the contact body to the global coordinate system, and obtain the current position and movement range of the execution unit;

[0030] Calculate the distance between each target operation point and each execution unit, obtain the target operation points within the movement range of the execution unit, and allocate the target operation points according to the priority and load balancing principle to form a subset of target operation points for the execution unit;

[0031] For the subset of target operation points of each execution unit, use a path optimization algorithm to plan a motion trajectory including path point coordinates, moving speed, acceleration, and dwell time, and output the control instruction for the execution unit.

[0032] Specifically, the specific steps for generating the control instruction for the adjustable unit include:

[0033] Obtain the pressure matrix and the effective contact body contour during the stable period, construct the three-dimensional surface equation of the contact surface between the contact body and the base platform, and extract the height extreme values of the three-dimensional surface;

[0034] Calculate the extreme value and the mean value of the pressure matrix, configure the compensation coefficient to generate the low pressure threshold, and divide the low pressure weak area by judging whether the pressure value within the contact body contour is less than the threshold;

[0035] Map the three-dimensional surface equation and the adjustable units of the support structure of the base platform into a grid, establish the correspondence between the adjustable unit coordinates and the surface height, screen the adjustable units in the low pressure weak area as compensation adjustable units, and screen the adjustable units in the non-contact body contour area as fixed adjustable units;

[0036] For the compensation adjustable units, calculate the compensation height based on the mean value of the pressure matrix and the pressure value of the adjustable unit; for the fixed adjustable units, set the compensation height to the height extreme value of the three-dimensional surface; generate the control instruction for the adjustable unit based on the compensation height, including the compensation height, the adjustment speed, and the adjustment method.

[0037] Specifically, the specific steps for judging whether the pressure matrix is in the stable period include:

[0038] Set the acquisition interval, acquire the pressure matrix, and calculate the element mean value of the pressure matrix and the change rate of the mean value of the adjacent pressure matrices;

[0039] Set the time window, and calculate the difference matrix of the corresponding matrix elements for the adjacent pressure matrices within the time window;

[0040] Set the mean change rate threshold and the difference norm threshold. When the change rates of the adjacent pressure matrix means within the time window are all less than the mean change rate threshold and the difference matrix norms are all less than the difference norm threshold, it is determined that the pressure matrix is in a stable period, and the contact body contour extraction operation is triggered.

[0041] An interactive execution mechanism includes a multi-modal driving component, an execution unit component, an adjustable support component, and a sensor feedback component;

[0042] The multi-modal driving component receives and parses the execution unit control instruction and the adjustable unit control instruction of the central control module in real time, converts them into corresponding hardware driving signals through the driving adaptation layer, dynamically allocates tasks based on the task priority and the load status, and ensures the multi-unit coordination and synchronization;

[0043] The execution unit component combines the built-in sensor feedback, precisely adjusts the position, speed, and output force through the closed-loop control algorithm, monitors the safety limit parameters in real time, and stops urgently and feeds back signals in case of anomalies;

[0044] The adjustable support component dynamically compensates the height of the low-pressure weak area through the grid adjustable unit based on the contact body surface equation and the pressure distribution data, and realizes the support and the execution unit action synchronization through the collaborative control mechanism;

[0045] The sensor feedback component collects the contact body and the execution mechanism state data in real time, constructs a control closed-loop, monitors the contact pressure, the motion trajectory deviation, and the contact force, configures the safety sensor to detect the abnormal state, and triggers the hardware-level emergency stop.

[0046] An intelligent device includes an intelligent bearing platform, an intelligent decision-making center, an operation support system, and a multi-modal human-machine interface;

[0047] The intelligent bearing platform has a pressure sensor array embedded on its surface, collects the pressure distribution data to generate a pressure matrix, determines the stable period by analyzing the mean, the mean change rate, and the difference matrix norm through the time window, preprocesses the data in the stable period, divides the connected regions, extracts the contact body contour, and generates the position calibration signal;

[0048] The intelligent decision-making center aligns the contact body contour with the standard contour, matches the feature points and calculates the similarity, establishes the local coordinate system to deduce the coordinates of the target operation point, and generates the execution unit control instruction including the coordinates of the target operation point and the adjustable unit control instruction including the height of the adjustable unit;

[0049] The operation support system integrates the execution unit and the adjustable unit. The execution unit adjusts the position and the function parameters based on the instruction, and the adjustable unit dynamically compensates the height of the low-pressure area to realize the support and the operation synchronization;

[0050] The multi-modal human-machine interface receives status signals and visually presents them, provides multi-modal guidance based on position calibration signals, and receives and formats user input parameters.

[0051] Advantages of the present invention:

[0052] Through multi-dimensional data processing and intelligent control, the present invention effectively solves problems such as inaccurate contour extraction, operation positioning deviation, uneven support, and low multi-unit cooperation efficiency in complex contact scenarios. By analyzing the stability of the pressure matrix through a time window, it ensures that the contour is extracted only when the contact is stable, avoiding misjudgment caused by motion interference and providing a reliable basis for subsequent operations; adopting multi-level conditional screening to merge connected regions, effectively dealing with the complex contours of human body surfaces or irregular workpieces, excluding noise regions and retaining key features, improving contour integrity; introducing a similarity verification and dynamic correction mechanism to solve contour deviations caused by noise or displacement, ensuring the effectiveness of the contour through multiple iterations; correcting the coordinates of the operation point based on the geometric relationship of feature points and shape descriptors to adapt to the individual differences of the contact body and achieve precise positioning; dividing regions according to the pressure distribution and differentially controlling the support units, dynamically compensating for gaps in low-pressure intervals and fixing the support to avoid contact failure or displacement and ensure uniform stress on the contact body; significantly enhancing the reliability and adaptability of the system in fields such as medical rehabilitation and industrial automation, and providing full-process intelligent support for the precise operation and stable support of the contact body. Description of the drawings

[0053] Figure 1 It is a schematic structural diagram of a dynamic interaction system of the present invention;

[0054] Figure 2 It is a flowchart of the specific steps for judging whether the pressure matrix is in a stable period of the present invention;

[0055] Figure 3 It is a flowchart of the specific steps for contour extraction of the contact body of the present invention;

[0056] Figure 4 It is a flowchart of the specific steps for calculating the coordinates of the operation point of the present invention;

[0057] Figure 5 It is a flowchart of the specific steps for generating control instructions for the execution unit of the present invention;

[0058] Figure 6 It is a flowchart of the specific steps for generating control instructions for the adjustable unit of the present invention. Detailed implementation manners

[0059] Example 1

[0060] Please refer to Figure 1, this embodiment introduces a dynamic interaction system, including a contact perception module, a central control module, an intelligent execution module, and an interaction feedback module;

[0061] The contact perception module generates a pressure matrix based on the pressure distribution data collected in real time, and determines whether the pressure matrix is in a stable period. If so, it triggers the contact body contour extraction operation, that is, performs data preprocessing on the pressure matrix, including noise reduction and feature enhancement, to obtain the preprocessed pressure matrix, and then identifies the contact body contour to provide reference data for subsequent target point positioning. At the same time, the placement state of the contact body is judged through multiple-level thresholds. When the effective contact area or the centroid position does not meet the preset standard, a position calibration signal is generated and transmitted to the interaction feedback module.

[0062] In this embodiment, pressure sensors are embedded on the surface of the basic platform to collect the pressure distribution data of the contact body in real time and generate a two-dimensional pressure matrix. In the present invention, the basic platform refers to a carrier structure for carrying the contact body and integrating the sensing function, and its form includes but is not limited to a flat plate body, a curved surface support table, a deformable contact layer, etc., and is applicable to multi-field scenarios such as medical rehabilitation, industrial automation, and smart home; the core function of the basic platform is to realize the real-time perception of the pressure distribution on the contact surface between the contact body and the basic platform through the built-in pressure sensors, so as to provide basic data support for subsequent contact body positioning, fixing, and operation. The contact body refers to any object that needs to be positioned, fixed, or operated, including but not limited to human body parts such as feet, hands, limbs, industrial workpieces, medical appliances, materials to be processed, etc. When the contact body is placed on the surface of the basic platform, the pressure sensors obtain a two-dimensional pressure matrix by detecting the pressure distribution on the contact surface, provide the original data input for the central control module, and then realize the shape recognition, target point calculation, and operation strategy generation of the contact body.

[0063] Please refer to Figure 2 , preferably, the specific steps for judging whether the pressure matrix is in a stable period include:

[0064] Set the acquisition interval for continuously collecting the pressure distribution matrix to generate a series of pressure matrices; for the time series of acquisitions as , then the corresponding pressure matrix is , is the length of the time series, and set a time window T for analyzing the change of the pressure matrix. At the th time point , then select the pressure matrix at the current time point and the previous T time for analysis; analyzing a single pressure matrix at a certain moment cannot determine whether it is stable, and the change trend of the pressure matrix needs to be observed for a period of time. By analyzing within the time window, the dynamic changes of the pressure matrix can be understood more comprehensively, so as to more accurately judge whether it is in a stable state.

[0065] For each pressure matrix, calculate the mean value of all its elements to reflect the overall pressure level of the pressure matrix at that moment. The mean value can comprehensively reflect the pressure conditions of each element in the pressure matrix and reflect the overall pressure magnitude when the contact body contacts the base platform. By judging the overall pressure level, it can be preliminarily determined whether the contact body and the base platform are in good contact. If the mean value is too low, it may indicate that the contact area between the contact body and the base platform is small or the contact is not tight; if the mean value is too high, it may indicate that the contact body exerts too much pressure on the base platform.

[0066] And within the time window, calculate the change rate of the mean values of adjacent pressure matrices, that is:

[0067]

[0068] where is the change rate of the mean value of the pressure matrix at the th time point , is the mean value of the pressure matrix at the th time point. When the mean value of the pressure matrix is 0, it indicates that there is no contact between the contact body and the base platform, then set the change rate of the mean value of the pressure matrix to an infinitely large number. The change rate of the mean value of the pressure matrix reflects the relative changes in the movement and contact depth of the contact body on the base platform, so as to further reflect whether the contact body is stably placed on the base platform to confirm whether operations such as contour extraction are to be performed;

[0069] For adjacent pressure matrices within the time window, calculate the difference matrix of the corresponding matrix elements, and measure the size of the difference matrix through matrix norms to reflect the overall change degree of the local elements of adjacent pressure matrices and reflect the contact depth of the contact body on the base platform. Matrix norms include: Frobenius norm, L1 norm, etc.;

[0070] Set the mean change rate threshold and the difference norm threshold to determine whether the movement of the contact body on the base platform and the contact depth are in a relatively stable state. The mean change rate threshold and the difference norm threshold are determined according to the actual application scenario and experience. For example, for some applications with high stability requirements, the mean change rate threshold and the difference norm threshold can be set smaller; for some applications with lower stability requirements, the thresholds can be appropriately relaxed. When the change rate of the adjacent pressure matrix means within the time window is less than the mean change rate threshold and the difference matrix norm is less than the difference norm threshold, it is determined that the pressure matrix is in a stable period, and the contact body contour extraction operation is triggered. When both the mean change rate and the difference matrix norm are small, it indicates that the position and contact depth of the contact body on the base platform are relatively stable. At this time, operations such as contact body contour extraction can obtain more accurate results. If the pressure matrix is unstable, the extracted contour may be affected by the movement of the contact body, resulting in inaccurate contours. When it is not determined that the pressure matrix is in a stable period, a position stability signal is generated and transmitted to the interactive feedback module to assist the user in quickly adjusting.

[0071] Please refer to Figure 3 , preferably, the specific steps of contact body contour extraction include:

[0072] Obtain the pressure matrix within the time window when the pressure matrix is in a stable period. In the stable state, the relative position and contact situation between the contact body and the base platform are relatively fixed. At this time, the pressure matrix can accurately reflect the actual pressure distribution of the contact body, laying a foundation for accurately extracting the contour subsequently.

[0073] Perform fusion processing on the pressure matrix within the time window to obtain the target pressure matrix. The weighted average method can be used, and different weights are assigned according to the chronological order of each pressure matrix collection. The matrix closer to the current time has a higher weight. To comprehensively consider the change of pressure within the time window, reduce the error caused by instantaneous fluctuations, and make the subsequently extracted contour more stable and accurate.

[0074] Perform data preprocessing on the target pressure matrix, including noise reduction processing and feature enhancement processing. Noise reduction processing refers to using filtering algorithms such as median filtering to remove impulse interferences such as salt-and-pepper noise in the target pressure matrix. Feature enhancement processing refers to using edge detection algorithms to refine the edges of the target pressure matrix, highlighting the areas with obvious pressure changes in the target pressure matrix, providing key information for subsequent contour recognition;

[0075] Convert the preprocessed target pressure matrix into a binary image. That is, by adopting an adaptive threshold, each pixel point in the target pressure matrix is divided into two categories: foreground and background, so as to maximize the between-class variance of the foreground and background, set the classification threshold, mark the pixels with pressure values greater than the classification threshold as the foreground, representing the contact body area, and mark the pixels with pressure values less than or equal to the classification threshold as the background. The binarization process can simplify the complex pressure distribution into clear contact body and non-contact body areas, providing a clear data basis for subsequent contour extraction.

[0076] Configure the connected component threshold, which is used to measure the number of standard contact areas between the contact body and the base platform. Traverse the foreground pixels in the binary image, divide the connected areas, and mark the mutually connected pixel areas as the same connected area; extract the basic features of each connected area, including area, centroid coordinates, and bounding rectangle.

[0077] If the number of divided connected areas is greater than the connected component threshold, then screen the valid connected areas and perform connected area merging so that the number of connected areas is equal to the connected component threshold, that is:

[0078] Calculate the distance from the centroid of each connected area to the centroids of the other connected areas, and combine the connected area area to screen the valid connected areas. That is, by configuring the area threshold and the centroid distance threshold, eliminate the connected areas with an area less than the area threshold and the distance from the centroid to the centroids of the other connected areas less than the centroid distance threshold.

[0079] For the valid connected areas, perform connected area merging successively based on the overlapping area of the bounding rectangles of the valid connected areas, the distance between centroids, and the topological relationship until the number of connected areas is equal to the connected component threshold;

[0080] Performing connected area merging based on the overlapping area of the bounding rectangles means that if there is an overlapping area between the bounding rectangles of two valid connected areas, then merge the two connected areas to form a new connected area and extract its basic features; in addition to the overlapping area of the bounding rectangles, the proportion of the overlapping area can also be considered. When the proportion of the overlapping area to the area of the smaller bounding rectangle is greater than the preset proportion threshold, merge the two connected areas;

[0081] Performing connected area merging based on the distance between centroids means calculating the distances between the centroids of all valid connected areas and constructing a distance matrix. Using a clustering algorithm, with the centroid distance as the metric standard, gradually merge the connected areas with closer distances.

[0082] Performing connected area merging based on the topological relationship means obtaining the topological relationship between the connected areas, including whether there is an inclusion relationship, an adjacent relationship, etc. For the connected areas with an inclusion relationship, merge the smaller connected area into the larger connected area; for the adjacent and complementary-shaped connected areas, also perform merging.

[0083] If the number of connected regions is less than the connected region threshold, a contact body placement signal is generated and transmitted to the interactive feedback module; the user is guided to adjust the placement state, reducing the cost of manual trial and error and improving the operation efficiency;

[0084] When the number of connected regions is equal to the connected region threshold, contour extraction is performed on the foreground pixel points in each connected region. The boundary tracking algorithm, such as the Moore neighborhood tracking algorithm, can be used to extract the boundary points of the connected region and form a closed contour.

[0085] Configure a similarity threshold. After smoothing the contours extracted from the connected regions, contour verification is performed based on prior knowledge, that is, calculate the similarity between the contour in each connected region and the standard contour in the prior database. If it is greater than the similarity threshold, the contour verification is successful, and the contour is marked as a valid contour, and the contact body contour of the connected region passing the contour verification is output; indicating that the currently extracted contact body contour conforms to the basic morphological characteristics of the corresponding object in the prior knowledge and can be used as a reliable basis for subsequent calculation of the pressure center, extraction of key feature points, and establishment of a local coordinate system. For example, in the medical rehabilitation scenario, if the foot contour meets the similarity standard with the standard foot contour in the prior database, it can be used for precise foot massage planning. The prior database is used to store the standard contours of the contact body and related prior knowledge. Its core function is to provide a benchmark reference for contact body contour verification to ensure that the extracted contours conform to the basic morphological characteristics of the target object. The database pre-stores the standard contour data of the contact body, the coordinates of the contour feature points, geometric attributes, and parameters related to the application scenario, forming a standardized contour verification benchmark library.

[0086] Otherwise, the contour verification fails, it is determined that there is a deviation in the currently extracted contour, the contour correction operation is started, the pressure distribution data is re-acquired, the dynamic contour is extracted, and then the contour verification is performed, and the number of contour verification times is counted; configure the verification threshold. If the number of contour verification times is greater than the verification threshold, a contour anomaly signal is generated and transmitted to the interactive feedback module; the user is guided to troubleshoot problems to avoid operation failures or safety risks caused by invalid contours;

[0087] The central control module deeply processes the contact body contour output by the contact perception module, matches the standard contour, and dynamically calculates the operation point coordinates in combination with the preset rule library. For different contact body types, the corresponding target operation points are matched through the strategy generation algorithm, and the execution unit control instruction and the adjustable unit control instruction are generated. When the contact body is the foot, the left and right foot contours of the foot are processed respectively, the standard contour is matched, and then the acupoint coordinates corresponding to the left and right foot contours are obtained to match the target acupoints to be operated, and the movement trajectory of operating on the target acupoints and the control instruction for supporting the foot according to the foot position are generated.

[0088] Please refer to Figure 4Preferably, the specific steps of calculating the coordinates of the operating point include:

[0089] The centroid of each contour of the contact body is used as the origin of the coordinate system, the long side direction of the external matrix of the contour is selected as the positive direction of the x-axis of the coordinate system, and the local coordinate system of each contour of the contact body is established based on the right-hand coordinate system rule;

[0090] Contour feature points are extracted for each contour of the contact body, including the use of curvature analysis method to extract contour feature points at the curvature mutation points, and the Harris corner detection algorithm is used to supplement the contour feature points of key corner points in the contour; for the sole contour, the curvature analysis method can be used to extract curvature mutation points such as toes and metatarsal protrusions as feature points. These parts show obvious shape changes on the sole contour, and the Harris corner detection algorithm can be used to supplement key corner points such as the ankle.

[0091] The contact body operation points and contour feature points under the standard contour are obtained from the prior database, as well as supplementary attribute information of each operation point, including the relative distance and angle between the operation point and the contour feature point, and the shape descriptor of the local area of ​​the standard contour where the operation point is located, including roundness and rectangularity. For example, when the contact body is a foot, the acupoints under the standard sole contour, i.e., the operation points, such as toe end points, metatarsal heads, ankle protrusions and other contour feature points are obtained from the prior database of foot acupoints and contour features.

[0092] The contact body contour is aligned with the standard contour by translation, rotation and scaling operations. First, the center of mass of the contact body contour is translated to coincide with the center of mass of the standard contour. Then, according to the proportional relationship between the long side lengths of the circumscribed matrix of the contact body contour and the standard contour, the standard contour is scaled so that the sizes of the two are in the same metric space.

[0093] According to the principle of the shortest distance or the principle of similarity of contour feature points, the contour feature points of the standard contour and the contact body contour are preliminarily matched to establish a preliminary correspondence relationship; for example, the principle of the shortest distance can be used to perform a preliminary match by calculating the Euclidean distance between each contour feature point detected on the contact body contour and the contour feature point of the standard contour, and after traversing all contour feature points, find the point with the shortest distance, and establish a preliminary correspondence relationship between the contact body contour point and the point with the shortest distance in the standard contour. If there are contour feature points with the same Euclidean distance, the principle of similarity of contour feature points is continued to be used for matching;

[0094] The contour feature point similarity principle includes: First, normalize the standard contour and the contact body contour to unify the size and direction. Then, calculate the local geometric descriptors of each contour feature point. For each feature point on the contact body contour, find the point on the standard contour with a similar geometric descriptor as the corresponding point. Exemplarily, when the contact body is a foot, if the local curvature and tangent direction of a certain corner point on the contact body contour have the highest similarity with the geometric descriptor of a certain feature point at the ankle on the standard contour, a preliminary correspondence relationship is established between the two.

[0095] According to the relative distance and angle relationship between the operating point and the contour feature point in the standard contour, and the coordinates of the preliminarily matched contour feature points in the contact body contour, the candidate coordinates of the operating point on the contact body contour are deduced through trigonometric function relationships. For example, if the operating point and the contour feature point in the standard contour have a distance of and an angle of , and the coordinates of the matched contour feature point in the contact body contour are , then the candidate coordinates of the operating point can be calculated through , .

[0096] Calculate the relative distance difference and angle difference between the matched feature points in the contact body contour and the standard contour, as well as the shape descriptor difference between the area where the candidate coordinates of the operating point are located and the area corresponding to the operating point in the standard contour, including circularity difference and rectangularity difference. Correct the candidate coordinates of the operating point through an elastic transformation model. Exemplarily, a distance correction coefficient, an angle correction coefficient, and a shape correction coefficient can be set, and the coordinate correction amount is calculated through the elastic transformation model to adjust the candidate coordinates; after correcting the candidate coordinates of the operating point, output the coordinates of the operating point for each contour of the contact body; when the contact body is a foot, output the precise coordinates of each acupoint on the sole of the foot in the established local coordinate system of the sole of the foot, providing accurate data support for subsequent operations such as positioning the massage head of a foot massage device and determining the training target point of a foot rehabilitation training device.

[0097] Please refer to Figure 5 , preferably, the specific steps for generating the execution unit control instruction include:

[0098] Receive the set of target operating points and the execution priority of the target operating points specified by the user from the interaction feedback module, and obtain the remaining operation requirement parameters and the operation body type identifier. Exemplarily, when the contact body is a foot, the target operating point is the operating acupoint located according to the user's needs, the operation requirement parameters include massage acupoints, massage intensity level, massage mode selection, etc., and the operation body type identifier is the left / right foot distinction information;

[0099] Convert the coordinates of the target operation points from the local coordinate system of the contact body to the global coordinate system to ensure the accurate positioning of the execution unit, obtain the position information of the execution unit, and divide the target operation point subset of each execution unit according to the movement range of the execution unit, that is:

[0100] For each target operation point, calculate its spatial distance from the current position of each execution unit respectively. Specifically, through the Euclidean distance formula in three-dimensional space, calculate the distance value between the coordinates of the target operation point and the position coordinates of the execution unit. After the calculation is completed, compare and judge this distance with the preset movement range of the execution unit. If the distance between the target operation point and the execution unit is less than or equal to the movement range radius of the execution unit, it is determined that the target operation point is within the movement range of the execution unit and is marked as reachable; otherwise, it is marked as unreachable, thereby obtaining the reachable target operation point list for each execution unit respectively.

[0101] In the reachable target operation point list of each execution unit, find the execution unit that matches the target operation point in turn, that is, give priority to processing the target operation point with the highest priority. In the reachable target operation point list of each execution unit, find the execution unit closest to this target operation point, assign this target operation point to this execution unit, and remove it from the reachable target operation point lists of other execution units to prevent repeated assignment.

[0102] For the remaining target operation points, if there are multiple execution units that can reach the same target operation point, first calculate the current load value of each execution unit. The load value comprehensively considers factors such as the number of assigned tasks, the estimated execution time, and the historical running duration, and is obtained through weighted calculation. Then select the execution unit with the smallest load value for assignment. If there are execution units with the same load value, select the unit closer to the target operation point for task assignment. Repeat the above steps until all target operation points are fully assigned, and finally generate a target operation point subset for each execution unit, including parameters such as the coordinates of the target operation point and the execution priority.

[0103] For a subset of target operation points for each execution unit, a path optimization algorithm is used to plan the movement trajectory of the execution unit, and at the same time, the path order is adjusted in combination with the execution priority to set the movement trajectory parameters of the execution unit, including: path point coordinates, moving speed, acceleration, and residence time; according to the operation requirement parameters, functional control parameters and safety limit parameters of the execution unit are generated, and an execution unit control instruction is generated; exemplarily, when the contact body is the foot, the operation requirement parameters include kneading frequency, vibration amplitude, and hot compress temperature, and the safety limit parameters include the movement range boundary and the functional control parameter threshold. The functional control parameters are set based on the operation requirements, the capabilities of the execution unit, and with reference to industry standards and historical experience for specific operation purposes and object characteristics; the safety limit parameters determine the movement range boundary and the functional control parameter threshold according to the physical limits of the execution unit, the working space limit, the equipment load capacity, and industry safety specifications to ensure operation safety.

[0104] Please refer to Figure 6 , preferably, the specific steps for generating the adjustable unit control instruction include:

[0105] Obtain the pressure matrix and the effective contact body contour during the stable period, construct the three-dimensional surface equation of the contact surface between the contact body and the base platform through the NURBS surface fitting or triangulation algorithm, and obtain the maximum height of the three-dimensional surface;

[0106] Calculate the extreme value and the mean value of the pressure matrix, and configure the compensation coefficient , whose value range is (0, 1). By quantifying the fluctuation degree of the pressure matrix, the precise division of the low-pressure weak area of the contact body is realized, providing a data basis for differential support control. Based on the compensation coefficient, a low-pressure threshold is generated , that is: , where is the mean pressure of the pressure matrix, is the maximum pressure of the pressure matrix, is the minimum pressure of the pressure matrix. By judging whether the pressure value within the contact body contour is less than the low-pressure threshold, the low-pressure weak area in the contact body contour is divided; the low-pressure weak area needs to be dynamically compensated to avoid contact failure. Exemplarily, when the contact body is the foot, the arch suspension area is divided to provide a basis for differential support control;

[0107] In order to achieve effective support for the contact body, the three-dimensional surface equation is mapped to the adjustable units of the base platform support structure in a grid manner, and a one-to-one correspondence between the coordinates of each adjustable unit and the corresponding surface height of the surface equation is established, and the adjustable units in the low-pressure weak area are obtained as the compensation adjustable units;

[0108] For each compensation adjustable unit, it is necessary to calculate its compensation height to ensure that the contact gap can be effectively filled and the contact stability can be improved. The compensation height of each compensation adjustable unit is calculated based on the mean value of the pressure matrix and the pressure value of the compensation adjustable unit, that is: , where is the compensation height of the th compensation adjustable unit, is the pressure-height compensation coefficient, is the compensation height of the th compensation adjustable unit. The pressure-height compensation coefficient is accurately calibrated through experiments or simulation analysis based on factors such as the material of the contact body, elastic modulus, and characteristics of the support structure.

[0109] To achieve uniform force on the contact body in the non-displacement state, according to the contact body contour, the adjustable units of the base platform support structure corresponding to the non-contact body contour are obtained as fixed adjustable units;

[0110] For each fixed adjustable unit, its compensation height is set to the maximum value of the three-dimensional surface height, so that the height of the adjustable unit of the base platform support structure corresponding to the non-contact body contour is consistent with the maximum height of the contact body, so as to fix the contact body when the contact body performs unit operations;

[0111] According to the calculated compensation height of each adjustable unit, corresponding adjustable unit control instructions are generated. The control instructions include compensation height, adjustment speed, and adjustment method; the setting of the adjustment speed needs to comprehensively consider the dynamic characteristics of the contact body and the response ability of the support structure. For situations that require rapid response, the adjustment speed can be appropriately increased; while for situations with higher stability requirements, the adjustment speed should be decreased. The adjustment method can be selected as continuous adjustment or step adjustment, depending on the driving method of the support structure and the control accuracy requirements.

[0112] The intelligent execution module is used to receive the execution unit control instructions and adjustable unit control instructions generated by the central control module, and achieve precise control of the execution unit and adjustable unit through the multi-modal drive interface;

[0113] Receive and parse the execution unit control instructions and adjustable unit control instructions in real time through the communication interface. The execution unit control instructions include the global coordinates of the target operation point, motion trajectory parameters, function control parameters, and safety limit parameters. The adjustable unit control instructions cover the target height, adjustment speed, and adjustment method of the support structure adjustable unit; the built-in drive adaptation layer uses the protocol conversion engine to parse the standardized instructions into control signals corresponding to the hardware, and is compatible with the control protocols of different actuators such as servo motors and pneumatic devices.

[0114] In terms of the execution unit control, based on the target operating point coordinates and motion trajectory parameters, combined with the feedback data of the built-in sensors of the execution unit, the position, speed, and functional parameters are adjusted in real time through a closed-loop control algorithm. Exemplarily, when the contact body is the foot, the massage head is controlled to locate the target acupoint and perform a kneading action at a preset frequency, while continuously monitoring the position of the execution unit and the safety limit parameters. When motion overrun or abnormal force is detected, an emergency stop is immediately triggered and an abnormal signal is fed back. For the adjustable unit control, the adjustable unit is driven to perform real-time adjustment according to the calculated compensation height to avoid local overload or contact failure, ensuring that the contact body realizes uniform force in a non-displacement state.

[0115] The interaction feedback module is used to receive the status signals from the contact perception module and the central control module in real time, and transmit operation guidance, status information, and abnormal prompts to the user through a multi-modal interaction interface. At the same time, it supports the user to input operation requirement parameters to form an interaction closed-loop. Specifically:

[0116] Build a multi-source signal receiving mechanism to parse the position calibration signals output by the contact perception module in real time, including position stability signals generated when the pressure matrix is unstable, contact body placement signals triggered when the number of connected regions is abnormal, and contour abnormal signals when the contour verification fails and the number of times exceeds the limit.

[0117] When the position stability signal is received, high-contrast prompt information is displayed on the touch screen, such as "Your position is not stable, please stay still", and a dynamic graphic guide is superimposed. Taking the medical foot massage scenario as an example, a semi-transparent standard foot contour template is synchronously displayed on the screen, with key load-bearing areas such as the heel and sole highlighted in green, and the current contact body position deviation area marked with a red flash, guiding the user to quickly adjust the posture through visual reference to ensure that the pressure distribution enters the stable period.

[0118] When the contact body placement signal is received, targeted prompts are generated according to the scenario requirements: in the double-foot massage scenario, "Please place your feet separately within the contour area" is displayed, and independent placement areas for the left and right feet are dynamically drawn at the bottom of the screen, with arrow animations indicating the correct displacement direction;

[0119] When the contour abnormal signal is received, a multi-level feedback mechanism is started, including popping up a full-screen warning window in the center of the touch screen, displaying "Please check the contact body placement posture or confirm the contact body type", and at the same time triggering a short beep alarm of the buzzer or a red flash of the device indicator light; if the user does not respond in time, further vibration feedback is provided, such as the vibration of the massage chair armrest, to enhance perception and ensure that the abnormal state is effectively recognized.

[0120] At the user input level, a touch screen is provided to receive operation requirement parameters, such as a set of target operation points, massage intensity levels, machining precision requirements, and contact body type identifiers, such as the left foot, right foot, workpiece A, and workpiece B. A slider control is set for numerical parameters and the input range is restricted. A drop-down menu is provided for enumerated parameters, and the input data is formatted to ensure the accuracy and reliability of the instruction parameters transmitted to the central control module.

[0121] The status visualization function realizes information presentation through a hierarchical interface. The bottom layer displays the result of the contact body contour extraction, and the contact interface is rendered in real time with a binary image or a three-dimensional mesh model, such as the edge contour line of the foot contour and the polygonal mesh of the industrial workpiece. The middle layer overlays a heat map of the pressure distribution, and the pressure matrix is dynamically mapped through color gradients. High pressure areas can be represented by red and low pressure areas by blue, supporting users to click to view local pressure values. The upper layer shows the operation progress, including the movement trajectory of the execution unit and the real-time height value of the adjustable unit.

[0122] Embodiment 2

[0123] This embodiment introduces an interactive execution mechanism for receiving the execution unit control instruction and the adjustable unit control instruction output by the central control module to achieve precise operation and stable support for the contact body. It includes a multi-modal drive component, an execution unit component, an adjustable support component, and a sensor feedback component.

[0124] The multi-modal drive component is used to receive and parse the standardized instructions output by the central control module in real time, including the execution unit control instruction, which contains the target operation point coordinates, movement trajectory parameters, function control parameters, and safety limit parameters, and the adjustable unit control instruction, which contains the target height of the adjustable unit, adjustment speed, and method. The instructions are converted into drive signals corresponding to the corresponding hardware through the drive adaptation layer, including servo motor pulse signals and pneumatic device pressure control signals, and are compatible with the control protocols of different execution mechanisms. Based on the task priority and the device load status, the operation tasks are dynamically allocated to the execution unit and the adjustable support unit to ensure the timing synchronization and resource balance of the multi-unit collaborative action.

[0125] The execution unit component is used to achieve precise operation of the contact body based on the control instruction output by the multi-modal drive component. Combining the built-in sensor feedback data, it precisely adjusts the position, speed, and output force through a closed-loop control algorithm. It also monitors the safety limit parameters in real time during the operation process, triggers an emergency stop when an abnormality occurs, and feeds back a signal.

[0126] The adjustable support assembly is used to provide differential support for the contact body based on the contact body surface equation and pressure distribution data generated by the central control module; through the grid-distributed adjustable units, the compensation adjustable units in the low-pressure weak areas are dynamically adjusted according to the calculated compensation height, filling the contact gap and avoiding contact failure; through the cooperative control mechanism, the support units and the execution units are ensured to act synchronously, realizing uniform fixed force on the contact body in the non-displacement state.

[0127] The sensor feedback assembly is used to collect the state data of the contact body and the actuator in real time to construct a control closed-loop; pressure sensors are integrated in the adjustable support units to monitor the contact pressure distribution and feedback it to the central control module for dynamic correction of the compensation height and contour verification; position sensors and force sensors are deployed in the execution units to feedback the motion trajectory deviation and contact force data in real time, supporting precision calibration and force-controlled compliant operation; safety monitoring sensors are configured to detect abnormal states in real time, trigger a hardware-level emergency stop and prompt the user through the interactive feedback module to ensure the safe and reliable operation of the system.

[0128] Embodiment 3

[0129] This embodiment introduces an intelligent device, which is a general intelligent terminal based on a dynamic interaction system, supporting the positioning, stable support and adaptive operation of the contact body; integrating four functions of contact perception, intelligent control, execution support and interactive feedback, applicable to multiple scenarios such as medical rehabilitation, industrial inspection, and smart home, including: an intelligent bearing platform, an intelligent decision-making center, an operation support system, and a multi-modal human-machine interface;

[0130] The intelligent bearing platform, as the physical bearing carrier and perception entrance of the contact body, realizes real-time perception of the contact state and extraction of contour features. The high-density pressure sensor array embedded on its surface collects the pressure distribution data of the contact surface between the contact body and the platform in real time, generates a two-dimensional pressure matrix, and analyzes the mean value of the pressure matrix, the change rate of adjacent mean values, and the norm of the difference matrix through a time window to determine whether the pressure matrix is in a stable period. After preprocessing such as noise reduction, feature enhancement, and binarization of the stable-period pressure matrix, the connected regions are divided, the effective regions are screened and merged, and the contact body contour is extracted. At the same time, the initial placement state of the contact body is evaluated through a multi-level threshold algorithm, and a position calibration signal is generated and transmitted to the multi-modal human-machine interface.

[0131] The intelligent decision-making center is used for data processing, strategy generation, and instruction output, realizing the intelligent mapping from perceived data to control instructions. It aligns the contact body contour extracted by the intelligent bearing platform with the standard contour in the prior database through translation, rotation, and scaling. It uses the curvature analysis method combined with Harris corner detection to match feature points and calculate the similarity. After passing the verification, a local coordinate system is established with the centroid of the contour as the origin. Based on the relative distance, angle, and shape descriptors of the operation points and feature points in the standard contour, the target operation point coordinates are generated through trigonometric function derivation and elastic transformation correction. Then, according to the execution priority, operation requirements, and device capabilities, two types of control instructions are generated. One is the execution unit control instruction that includes the target operation point coordinates, motion trajectory parameters, function control parameters, and safety limit parameters. The other is the adjustable unit control instruction that includes the target height, adjustment speed, and method of the adjustable unit.

[0132] The operation support system is used to realize the coordinated control of precise operation and dynamic support for the contact body. It integrates multi-modal execution units and an intelligent support matrix. The multi-modal execution units support different types of actuators such as rigid robotic arms and flexible actuators, and adapt various operation tools through end-effector quick-change interfaces. Based on the closed-loop control algorithm, the position, speed, and function parameters are adjusted in real time. The intelligent support matrix is composed of grid-distributed adjustable units. The compensation adjustable units for low-pressure weak areas calculate the compensation height based on the mean value of the pressure matrix and the unit pressure value and dynamically adjust to fill the contact gap. The fixed adjustable units are set to the extreme values of the surface height to provide rigid support. And through the coordinated control mechanism, it ensures that the support units and the execution units act synchronously, first completing the support adjustment and then starting the operation, realizing the uniform force on the contact body in a non-displacement state.

[0133] The multi-modal human-machine interface is used to realize the two-way interaction of status visualization, operation guidance, and user input. It receives the status signals from the contact perception module and the central control module in real time, and presents information such as the contact body contour extraction result, pressure distribution heat map, motion trajectory of the execution unit, and real-time height of the adjustable unit in a hierarchical interface. When receiving the position calibration signal, it guides the user to adjust the position of the contact body through multi-modal methods such as touch screen graphic guidance, voice prompts, and vibration feedback. When receiving signals such as contour anomalies, it starts a multi-level feedback mechanism, including popping up a warning window, triggering sound and light alarms, etc. At the same time, it provides a visual operation interface to receive the operation requirement parameters and contact body type identification input by the user, sets the sliding bar limit range for numerical parameters, provides a drop-down menu for enumeration parameters, and formats the input data to ensure the accuracy and reliability of the instruction parameters transmitted to the central control module.

[0134] Working principle and its effects:

[0135] The dynamic interaction system of the present invention realizes adaptive operation and stable support for the contact body through a closed loop: the pressure sensor collects the pressure matrix in real time, extracts the contour only when the contact is stable through time window analysis, and accurately processes complex contours such as human limbs and industrial workpieces by noise reduction, binarization and connected region screening to exclude noise interference. The extracted contour is matched and verified with the standard contour feature points, and when it does not meet the standard, it automatically restarts and corrects to ensure the validity of the contour; a local coordinate system is established based on the effective contour, and the coordinates are deduced and elastically corrected in combination with the geometric relationship of the standard operation points to achieve millimeter-level positioning and adapt to individual differences.

[0136] The low-pressure weak area is dynamically divided according to the pressure distribution. The compensation adjustable unit calculates the height to fill the gap according to the pressure difference, and the fixed unit provides rigid support with the surface extreme value to ensure uniform force and no displacement of the contact body; multiple execution units allocate tasks according to the distance to the target point, the movement range and the load balance, optimize the path and improve the cooperation efficiency. The interaction module receives signals in real time and guides the adjustment through multi-modal methods such as graphic guidance and voice prompts, visualizes the pressure distribution and the execution state, and forms a human-computer interaction closed loop.

[0137] It solves the problems of inaccurate contour extraction, positioning deviation, uneven support and low cooperation efficiency in the traditional dynamic interaction system, realizes the full-process intelligence from contact perception to precise operation, significantly improves the reliability and operation accuracy of equipment in fields such as medical rehabilitation and industrial automation, provides an efficient solution for intelligent contact control, and promotes the automation upgrade of multiple scenarios.

[0138] The above is only the preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those of ordinary skill in the art, several improvements and refinements made without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A dynamic interaction system, characterized in that, Including: Receiving the operation requirement parameters and the contact type identifier input by the user; Collecting the pressure distribution data of the contact surface between the contact body and the base platform to generate a pressure matrix, judging whether the pressure matrix is in a stable period by the change of the pressure matrix within the time window. If it is in the stable period, the pressure matrix is fused, feature enhanced and binarized to divide, screen and merge the connected regions, extracting the contour based on the foreground pixel points within the connected regions and calculating the similarity with the standard contour for contour verification to output the contact body contour; at the same time, judging the placement state of the contact body through multi-level thresholds and generating a position calibration signal; Receiving the position calibration signal and outputting operation guidance and exception prompts to the user; By extracting the contour feature points of the contact body contour and the standard contour, matching the contact body contour and the standard contour, based on the operation point information of the standard contour, deriving and correcting the candidate coordinates of the operation points of the contact body contour through trigonometric function relationships to obtain the operation point coordinates of the contact body contour, and generating an execution unit control instruction and an adjustable unit control instruction based on the operation point coordinates and the contact body contour; Receiving and parsing the execution unit control instruction and the adjustable unit control instruction, driving the execution unit and the adjustable unit to realize the operation and support fixation of the contact body.

2. The dynamic interaction system according to claim 1, characterized in that, The specific steps of the contact body contour extraction include: Performing weighted average fusion on the pressure matrix in the stable period to obtain a target pressure matrix, performing noise reduction processing and feature enhancement processing on the target pressure matrix, and then dividing each pixel point in the target pressure matrix into two categories, foreground and background, by an adaptive threshold, and converting the target pressure matrix into a binary image; Configuring a connected domain threshold, traversing the foreground pixels in the binary image, dividing the connected regions and extracting the features of the connected regions; Judging whether the number of divided connected regions is greater than the connected domain threshold. If it is greater, screening the effective connected regions and merging the connected regions so that the number of connected regions is equal to the connected domain threshold; otherwise, judging whether the number of divided connected regions is less than the connected domain threshold. If it is less, generating a contact body placement signal; otherwise, extracting the contour for each foreground pixel point within the connected regions; The specific steps of screening the effective connected regions and merging the connected regions include: Calculating the distance from the centroid of each connected region to the centroids of the other connected regions, and combining the area of the connected regions to screen the effective connected regions; successively merging the connected regions based on the overlapping area of the circumscribed rectangles of the effective connected regions, the distance between the centroids, and the topological relationship until the number of connected regions is equal to the connected domain threshold.

3. A dynamic interaction system according to claim 2, characterized in that, The specific steps of the contact body contour extraction further include: Configuring a similarity threshold, smoothing the contour extracted from the connected region, calculating the similarity between the contour within each connected region and the standard contour. If it is greater than the similarity threshold, the contour verification is successful, and it is marked as an effective contour, and the contact body contour of the connected region passing the contour verification is output; Otherwise, the contour verification fails, the contour correction operation is started, the pressure distribution data is recollected, the dynamic contour is extracted, and then the contour verification is performed, and the number of contour verification times is counted; a verification threshold is configured. If the number of contour verification times is greater than the verification threshold, a contour anomaly signal is generated.

4. A dynamic interaction system according to claim 1, characterized in that, The specific steps for obtaining the operating point coordinates of the contact body contour include: Establish a local coordinate system for the contact body contour; Extract the contour feature points of the contact body contour, and obtain the operating points, contour feature points of the contact body under the standard contour, and the supplementary attribute information of each operating point; Align the contact body contour with the standard contour, and preliminarily match the contour feature points of the standard contour and the contact body contour according to the principle of the closest distance or the similarity principle of contour feature points.

5. A dynamic interaction system according to claim 4, characterized in that, The specific steps for obtaining the operating point coordinates of the contact body contour further include: According to the relative distance and angle relationship between the operating points and the contour feature points in the standard contour, and the coordinates of the preliminarily matched contour feature points in the contact body contour, deduce the candidate coordinates of the operating points of the contact body contour through trigonometric function relationships; Calculate the relative distance difference and angle difference between the matching feature points of the contact body contour and the standard contour, and the shape descriptor difference between the area where the candidate coordinates of the operating points are located and the area corresponding to the operating points in the standard contour, and correct the candidate coordinates of the operating points through an elastic transformation model; output the operating point coordinates of each contour of the contact body.

6. The dynamic interaction system according to claim 1, characterized in that, The specific steps for generating the control instruction of the execution unit include: Receive the target operating point set, the execution priority of the target operating points, the operation requirement parameters, and the operation body type identifier; Convert the target operating point coordinates from the local coordinate system of the contact body to the global coordinate system, and obtain the current position and movement range of the execution unit; Calculate the distance between each target operating point and each execution unit, obtain the target operating points within the movement range of the execution unit, and allocate the target operating points according to the priority and load balancing principle to form the target operating point subset of the execution unit; For the target operating point subset of each execution unit, use a path optimization algorithm to plan a motion trajectory including path point coordinates, moving speed, acceleration, and staying time, and output the control instruction of the execution unit.

7. A dynamic interaction system according to claim 1, characterized in that, The specific steps for generating the control instruction of the adjustable unit include: Obtain the pressure matrix and the effective contact body contour during the stable period, construct a three-dimensional surface equation of the contact surface between the contact body and the base platform, and extract the height extreme values of the three-dimensional surface; Calculate the extreme value and the mean value of the pressure matrix, configure a compensation coefficient to generate a low pressure threshold, and divide the low pressure weak area by judging whether the pressure value within the contact body contour is less than the threshold; Perform a grid mapping of the three-dimensional surface equation and the adjustable units of the support structure of the base platform, establish the corresponding relationship between the adjustable unit coordinates and the surface height, screen the adjustable units in the low pressure weak area as compensation adjustable units, and screen the adjustable units in the non-contact body contour area as fixed adjustable units; For the compensation adjustable units, calculate the compensation height based on the mean value of the pressure matrix and the pressure value of the adjustable units; for the fixed adjustable units, set the compensation height to the height extreme value of the three-dimensional surface; generate the control instruction of the adjustable unit based on the compensation height, including the compensation height, the adjustment speed, and the adjustment method.

8. The dynamic interaction system according to claim 1, characterized in that, The specific steps for determining whether the pressure matrix is in a stable period include: Set the acquisition interval, acquire the pressure matrix, and calculate the change rate of the element mean of the pressure matrix and the mean of the adjacent pressure matrix; Set a time window, and for adjacent pressure matrices within the time window, calculate the difference matrix of the corresponding matrix elements; Set a mean change rate threshold and a difference norm threshold. When the change rates of the means of adjacent pressure matrices within the time window are all less than the mean change rate threshold and the norms of the difference matrices are all less than the difference norm threshold, it is determined that the pressure matrix is in a stable period, and the contact body contour extraction operation is triggered.

9. An interactive execution mechanism is used to receive the execution unit control instruction and the adjustable unit control instruction of a dynamic interactive system according to any one of claims 1-8, and operate and stably support the contact body. It is characterized in that It includes a multi-modal drive component, an execution unit component, an adjustable support component, and a sensor feedback component; The multi-modal drive component receives and parses the execution unit control instruction and the adjustable unit control instruction of the central control module in real time, converts them into corresponding hardware drive signals through the drive adaptation layer, dynamically allocates tasks based on the task priority and load status, and ensures the coordination and synchronization of multiple units; The execution unit component combines the built-in sensor feedback, precisely adjusts the position, speed, and output force through the closed-loop control algorithm, monitors the safety limit parameters in real time, and stops urgently and feeds back signals in case of abnormalities; The adjustable support component dynamically compensates the height of the low-pressure weak area through the grid adjustable unit based on the contact body surface equation and the pressure distribution data, and realizes the synchronization of the support and the execution unit actions through the collaborative control mechanism; The sensor feedback component collects the state data of the contact body and the actuator in real time, constructs a control closed-loop, monitors the contact pressure, the motion trajectory deviation, and the contact force, configures safety sensors to detect abnormal states, and triggers a hardware-level emergency stop.

10. An intelligent device, which is an intelligent terminal for implementing a dynamic interaction system according to any one of claims 1-8, characterized in that, It includes an intelligent bearing platform, an intelligent decision-making center, an operation support system, and a multi-modal human-machine interface; The intelligent bearing platform has a pressure sensor array embedded on its surface, acquires pressure distribution data to generate a pressure matrix, determines the stable period by analyzing the mean, mean change rate, and difference matrix norm through a time window, preprocesses the data in the stable period, divides the connected regions, extracts the contact body contour, and generates a position calibration signal; The intelligent decision-making center aligns the contact body contour with the standard contour, matches the feature points and calculates the similarity, establishes a local coordinate system to deduce the coordinates of the target operation point, and generates an execution unit control instruction containing the coordinates of the target operation point and an adjustable unit control instruction containing the height of the adjustable unit; The operation support system integrates the execution unit and the adjustable unit. The execution unit adjusts the position and functional parameters based on the instruction, and the adjustable unit dynamically compensates the height of the low-pressure area to achieve the synchronization of support and operation; The multi-modal human-machine interface receives the status signal and presents it visually, provides multi-modal guidance according to the position calibration signal, and receives and formats the user input parameters.

Images