Optical positioning and vision modeling-based implant point optimization system and method

Abstract

The application discloses an implantation point optimization system and method based on optical positioning and visual modeling, belongs to the field of precise implantation and optical navigation, and aims to solve the problems of insufficient positioning accuracy of an existing implantation point, disconnection between planning and operation guidance, high operation risk, and dependence on manual experience without standardized quantitative control. The application constructs a three-dimensional entity model of an operation object through non-contact optical scanning, completes standardized positioning of an implantation point, multi-dimensional compliance quantitative evaluation, and hierarchical safety path planning, and realizes accurate target visualization guidance of a target node in combination with real-time optical dynamic registration. The scheme can significantly improve the positioning accuracy and operation safety of precise implantation operation, and effectively reduce human errors and operation risks.

Description

Technical Field

[0001] This invention relates to the fields of precision implantation and optical navigation technology, and more specifically to an implantation point optimization system and method based on optical positioning and visual modeling. Background Technology

[0002] With the rapid development of precision medical technology and high-end equipment manufacturing industry, precision implantation has been widely used in clinical medicine, aerospace, precision mechanical assembly, microelectromechanical system integration and other fields. This has placed extremely stringent requirements on the positioning accuracy of implantation sites, the standardized control of operation processes, and the safety and stability of operation processes. The accuracy of implantation sites directly determines the final implantation effect, service reliability and safety. Deviations at the millimeter or even micrometer level may lead to implantation failure, structural damage or even serious safety accidents.

[0003] Currently, the entire process of precision implantation operations relies heavily on the professional experience and subjective judgment of operators. A unified, standardized management system covering the entire chain from implantation site planning and quality control to implementation has not yet been established. This results in a lack of unified standards for implantation site positioning, difficulty in consistently ensuring site accuracy and repetitiveness with similar operations, a lack of objective and quantifiable evaluation and judgment criteria for the usability of implantation sites, and an inability to effectively screen out unqualified sites beforehand. Furthermore, the connection between operation path planning and preliminary site planning is insufficient, and the operation execution process lacks precise and intuitive full-process guidance. Deviations between actual operation and preset technical requirements are prone to occur, significantly reducing operational efficiency and easily leading to implantation failure, component damage, and even safety accidents in high-risk scenarios. This makes it difficult to meet the core development requirements of high precision, high reliability, and high safety in current precision implantation operations, and severely restricts the large-scale promotion and application of precision implantation technology in more demanding scenarios. Therefore, to overcome these limitations, this invention proposes an implantation point optimization system and method based on optical positioning and visual modeling. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide an implantation point optimization system and method based on optical positioning and visual modeling, which solves the technical problems of how to improve the accuracy of implantation point positioning, the level of standardized control of implantation process and the safety of operation in precision implantation, and reduce human error and operational risks in implantation operations.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] Implantation point optimization methods based on optical positioning and visual modeling include:

[0007] A full-view optical 3D scan is performed on the target area of ​​the object to be implanted, generating a 3D solid model of the object's physical structure. Based on the preset implantation operation specifications, the positioning reference structure matching the implantation operation is extracted from the 3D solid model, a standard spatial rectangular coordinate system for the implantation operation is established, a set of pose parameters corresponding to the standard implantation point is generated, and the standard implantation point of the 3D solid model is located and marked.

[0008] A quantitative evaluation model for implantation point quality is constructed. Based on the set of pose parameters corresponding to the standard implantation point, the comprehensive quality score of the standard implantation point is calculated to determine whether the standard implantation point meets the compliance requirements.

[0009] If the standard implantation point meets the compliance requirements, the set of auxiliary operation points that match the operation type of the standard implantation point is retrieved and mapped and marked in the 3D solid model. The target space coordinates of each auxiliary operation point are obtained to construct the full domain space of the implantation operation and divide it into grid cells. By calculating the node spatial distribution density value of each grid cell, the minimum safe avoidance distance is allocated. Combined with the obstacle space contour of the 3D solid model, the implantation path is generated.

[0010] Based on the implantation path, the target node is updated in real time, real-time optical 3D data of the actual working area is collected, dynamic registration is performed with the 3D entity model, it is determined whether the effective acquisition space range of the current actual working area completely covers the current target node, and targeted visualization guidance is provided for the target node.

[0011] Specifically, the steps for establishing a standard spatial rectangular coordinate system for implantation operations include:

[0012] Multimodal optical 3D data of the target area to be implanted is collected, and spatial dimension is uniformly processed. Global registration is performed according to the spatial position correspondence, and the data is fused into a unified data set under a unified spatial coordinate system.

[0013] Anomalies are identified and removed from the unified dataset, and spatial grids are generated to construct an initial grid structure that matches the surface morphology of the work object.

[0014] The initial mesh structure is surface-fitted and smoothly reconstructed to generate a three-dimensional solid model of the object's solid structure.

[0015] Pre-defined implantation operation specifications include structural adaptation requirements, pose accuracy requirements, safety obstacle avoidance requirements, and requirements for selecting positioning reference structures required for implantation operations.

[0016] Based on the preset implantation operation specifications, the global positioning reference structure is extracted from the three-dimensional solid model, and structural regions that meet the requirements for the selection of positioning reference structures in the implantation operation specifications are selected.

[0017] The selected structural regions are analyzed for geometric features to extract their geometric center, contour boundaries, and normal directions. A standard spatial rectangular coordinate system for the implantation operation is established, and the standard spatial rectangular coordinate system is calibrated axially.

[0018] Specifically, the steps for locating and marking standard implantation points on a 3D solid model include:

[0019] Based on the structural adaptation requirements and pose accuracy requirements in the preset implantation operation specifications, the target area to be implanted in the three-dimensional solid model is partitioned and analyzed in a standard spatial rectangular coordinate system to identify the target point area.

[0020] Collect spatial coordinate information of each point within the target location area, filter out the core point set of the target location area through coordinate clustering analysis, and calculate the average coordinate of the core point set as the target spatial coordinates corresponding to the target location area;

[0021] Based on the surface morphology of the target location area, the normal direction of the core point set of the target location area is calculated, which is used to solve for the implantation normal attitude angle information corresponding to the target location area;

[0022] Measure the continuous surface area of ​​the target location area, select the continuous surface within the target location area whose flatness meets the implantation operation specifications as the effective working surface; and determine the boundary range and geometric parameters of the effective working surface to form the effective working surface parameter information corresponding to the target location area.

[0023] The target spatial coordinates, implantation normal attitude angle information and effective working surface parameter information corresponding to the target point area are associated and integrated to form a set of pose parameters corresponding to the standard implantation point.

[0024] In the 3D solid model, based on the set of pose parameters corresponding to the standard implantation point, the spatial point corresponding to the standard implantation point is located, and a unique positioning mark is drawn at that spatial point.

[0025] Specifically, the steps to determine whether a standard implantation site meets compliance requirements include:

[0026] Based on the pre-defined implantation operation specifications, a multi-dimensional evaluation index system for implantation site quality assessment is determined, and a quantitative evaluation model for implantation site quality is constructed; corresponding quantitative assignment rules and weight coefficients are set for each evaluation index in the multi-dimensional evaluation index system.

[0027] According to the quantitative assignment rules corresponding to each evaluation indicator, the collected pose parameter set is quantitatively converted to obtain the standardized quantitative score corresponding to each evaluation indicator; the standardized quantitative score corresponding to each evaluation indicator is weighted and calculated with the corresponding weight coefficient, and the comprehensive quality score corresponding to the standard implantation point is obtained by weighted summation.

[0028] If the overall quality score is greater than or equal to the passing threshold, the standard implantation point is deemed to meet compliance requirements, a compliance pass instruction is generated, and the subsequent implantation path planning process is triggered. If the overall quality score is less than the passing threshold, the standard implantation point is deemed to not meet compliance requirements, a compliance fail instruction is generated, and the subsequent alternative implantation point planning and anomaly warning process is triggered.

[0029] Specifically, the steps for generating the implantation path include:

[0030] The target spatial coordinates corresponding to the standard implantation point are taken as the end point of the implantation path, and the initial entry point corresponding to the implantation operation is taken as the starting point of the implantation path; and the set of auxiliary operation points that match the operation type of the standard implantation point are retrieved.

[0031] Using the standard Cartesian coordinate system corresponding to the 3D solid model as a reference, and combining the process space mapping rules of the auxiliary operation point set, each auxiliary operation point in the auxiliary operation point set is mapped to the 3D solid model to obtain the target space coordinates of each auxiliary operation point.

[0032] Based on the process execution order of each auxiliary operation point, the starting point of the implantation path, each auxiliary operation point, and the ending point of the implantation path are sorted to form a node sequence for implantation path planning.

[0033] Based on the geometric structure data of the 3D solid model, the 3D isosurface boundary of the work object is extracted as the obstacle space contour;

[0034] The two adjacent nodes in the node sequence are divided into an independent path planning unit. The obstacle space contour of the 3D solid model is used as the spatial constraint. The path planning is performed in combination with the minimum safe avoidance distance of each grid unit to generate the sub-path corresponding to each path planning unit. The sub-paths of all path planning units are connected in the order of the node sequence to obtain the implanted path.

[0035] Specifically, the steps of constructing the entire space for the implantation operation and dividing it into grid cells, and allocating the minimum safe avoidance distance by calculating the node spatial distribution density value of each grid cell, include:

[0036] A cuboid space covering the starting point of the implantation path, each auxiliary operation point, the ending point of the implantation path, and the preset operation range is constructed as the implantation operation global space; the implantation operation global space is discretized into three-dimensional equal-size rasterization, and the implantation operation global space is decomposed into several raster units;

[0037] Based on the target spatial coordinates of each auxiliary operation point and the standard implantation point, combined with the side length of a single grid cell and the average spacing between adjacent nodes in the node sequence, the calculation kernel width is determined. The geometric center of each grid cell in the entire implantation operation space is used as the calculation base point. The Gaussian kernel function is used to traverse and calculate the density contribution value of all nodes in the node sequence to the grid cell, and the node spatial distribution density value of each grid cell is calculated.

[0038] Based on the preset quantile threshold of the global grid cell density value, all grid cells are divided into operation space clusters, and the minimum safe avoidance distance of each operation space grid cell is assigned based on the positive correlation of the node spatial distribution density.

[0039] Specifically, the steps for obtaining the target space coordinates of each auxiliary operation point include:

[0040] Read the preset relative position parameters between each auxiliary operation point and the standard implantation point from the process space mapping rules built into the auxiliary operation point set, including relative coordinate offset, relative angle constraint, and relative distance range;

[0041] Based on the target spatial coordinates and implantation normal attitude angle of the standard implantation point, the initial three-dimensional spatial coordinates of each auxiliary operation point in the current three-dimensional solid model are calculated in the standard spatial rectangular coordinate system according to the relative position parameters of each auxiliary operation point.

[0042] Obtain the structural features of the positioning reference structure associated with each auxiliary operation point, and use the initial three-dimensional spatial coordinates of each auxiliary operation point as the center of the sphere, combined with its structural features, to set the basic search radius and delineate the spherical search area.

[0043] Within the spherical search area, the actual structural features of each auxiliary operation point are extracted point by point, and the similarity between the actual structural features and the structural features of the corresponding associated positioning reference structure is calculated. This is used to filter candidate points within the spherical search area, and the three-dimensional spatial coordinates of the candidate points are used as the target spatial coordinates of the auxiliary operation point.

[0044] Specifically, the steps for collecting real-time optical 3D data of the actual work area and performing dynamic registration with the 3D solid model include:

[0045] Preprocess the optical 3D data of the current actual working area acquired in real time, extract real-time surface features, and construct a real-time local surface model;

[0046] Extract the standard surface model of the 3D solid model, perform point-by-point geometric analysis, extract the edge feature points and curvature extreme points of the standard surface model to form a registration reference feature set; and perform coordinate normalization processing on the standard surface model to construct its axis-aligned bounding box and mark the geometric center.

[0047] Calculate the curvature value of each sampling point of the real-time local surface model, extract real-time curvature feature points, perform curvature similarity matching between the real-time curvature feature points and the curvature extreme points in the registration reference feature set, and screen effective curvature feature points.

[0048] A spherical extended verification range is defined with the effective curvature feature points as the center and half the length of the spatial diagonal of the standard surface model axis-aligned bounding box as the radius. The matching density of the effective curvature feature points within the extended verification range is statistically analyzed, and potential registration regions are selected.

[0049] Within the selected potential registration area, based on the matched effective curvature feature points, the edge feature points of the mapped standard surface model are searched, and the mapped edge points are connected in the edge order of the standard surface model to form a closed real-time operation area that matches the contour of the standard surface model.

[0050] Specifically, the steps for targeted visual guidance of target nodes include:

[0051] Within the standard Cartesian coordinate system, the effective coverage area of ​​the target node is defined by taking the target spatial coordinates of the current target node as the geometric center and combining the preset operation tolerance range.

[0052] The closed real-time operation area matching the current actual operation area is compared with the effective coverage determination area of ​​the target node. If the closed real-time operation area completely contains the effective coverage determination area of ​​the target node, it is determined that the target node has been covered; otherwise, it is determined that the target node has not been covered.

[0053] If the target node is determined to be uncovered, the geometric centroid of the actual working area closed real-time operation area that matches the contour of the standard surface model is calculated, and the three-dimensional spatial position deviation between the geometric centroid of the target node effectively covered judgment area is calculated, and a standardized displacement guidance command is generated.

[0054] If the target node is determined to be covered, the target node on the surface of the actual work area is located, and the target node is visualized and the work is guided.

[0055] An implantation point optimization system based on optical positioning and visual modeling includes a model building module, a compliance judgment module, an intelligent planning module, and a visualization guidance module.

[0056] The model building module is used to generate a three-dimensional solid model of the object's physical structure, extract the positioning reference structure that matches the implantation operation, establish a standard spatial rectangular coordinate system for the implantation operation, generate a set of pose parameters corresponding to the standard implantation point, and locate and mark the standard implantation point of the three-dimensional solid model.

[0057] The compliance assessment module is used to build a quantitative evaluation model for implantation site quality and determine whether standard implantation sites meet compliance requirements.

[0058] The intelligent planning module is used to retrieve a set of auxiliary operation points that match the operation type of the standard implantation point, map and annotate them into the 3D solid model, and generate an implantation path by combining the obstacle space contour of the 3D solid model.

[0059] The visualization guidance module is used to collect real-time optical 3D data of the actual work area, perform dynamic registration with the 3D entity model, determine whether the effective collection space of the current actual work area completely covers the current target node, and provide targeted visualization guidance for the target node.

[0060] The beneficial effects of this invention are:

[0061] This application constructs a 3D solid model of the work object through full-view optical 3D scanning. It then extracts a matching positioning reference structure based on standardized implantation operation specifications and establishes a unified standard spatial rectangular coordinate system. This avoids systematic deviations caused by subjective selection of positioning references from the outset. Through target point area analysis, coordinate clustering screening, and multi-dimensional parameter integration, it accurately calculates the pose parameters of the standard implantation points and completes standardized positioning marking, improving the positioning accuracy and repeatability consistency of the implantation points. Furthermore, by constructing a multi-dimensional implantation point quality quantitative evaluation model, it objectively determines the compliance of implantation points using standardized rules and scientific weight allocation, replacing traditional manual experience-based judgment. This proactively screens out unqualified points, avoids safety risks such as implantation failure, establishes unified quality control standards, and improves the overall process. The process achieves standardized management and control; for compliant implantation points, a set of standardized auxiliary operation points corresponding to the operation type is matched and precise spatial mapping calibration is completed. Through full-domain spatial rasterization, node density calculation and allocation of graded safety avoidance distances, and combined with obstacle contours, the implantation path planning is optimized, ensuring the adaptability of the path and operation procedures and the safety of the operation; finally, the target nodes are updated in real time based on the planned path, and coordinate calibration is completed through real-time optical data acquisition and model dynamic registration. Combined with coverage judgment, targeted visualization guidance is achieved, realizing precise linkage between operation execution and pre-planning, reducing human operation deviation, improving the accuracy and efficiency of precision implantation operations, and can be adapted to various scenarios such as medical implantation and precision mechanical assembly, with good applicability and promotion value. Attached Figure Description

[0062] Figure 1This is a flowchart of the implantation point optimization method based on optical positioning and visual modeling of the present invention;

[0063] Figure 2 This is a flowchart illustrating the process by which the present invention determines whether a standard implantation point meets compliance requirements.

[0064] Figure 3 A flowchart for generating the implantation path in this invention;

[0065] Figure 4 This is a flowchart illustrating the targeted visualization guidance for target nodes according to the present invention. Detailed Implementation

[0066] Please see Figure 1 This embodiment introduces an implantation point optimization method based on optical positioning and visual modeling, including:

[0067] Step S1: Using a non-contact optical scanning device, a full-view optical 3D scan of the target area to be implanted is performed on the object, acquiring multimodal optical 3D data of the target area. The acquired multimodal optical 3D data is then processed sequentially with point cloud denoising, point cloud registration, outlier removal, mesh generation, and surface reconstruction to generate a 3D solid model of the object's physical structure, providing a unified digital reference carrier for subsequent implantation point planning and operation guidance. Based on preset implantation operation specifications, a positioning reference structure matching the implantation operation is extracted from the 3D solid model. Using the extracted positioning reference structure as the origin, a standard spatial rectangular coordinate system dedicated to the implantation operation is established, clarifying the axial rules and scale accuracy of the coordinate system. Based on preset implantation operation parameter requirements, the 3D spatial coordinates, implantation normal attitude angle, and effective working surface parameters corresponding to the standard implantation point are calculated in the standard spatial rectangular coordinate system, marking the standard implantation point location on the 3D solid model.

[0068] In this embodiment, the work object refers to a physical object requiring precision implantation or precision machining, including at least one of human biological tissue, precision mechanical parts, and micro-assembly components. For example, when the work object is human temporal bone tissue used for cochlear implantation surgery, the non-contact optical scanning device refers to an optical positioning device used for intraoperative and preoperative three-dimensional morphological acquisition, including at least one of: a structured light 3D scanner, a laser 3D scanner, a binocular vision positioning system, and an intraoperative optical navigation camera. The positioning reference structure refers to a landmark structure within the target area of ​​the work object that has a fixed geometric position, stable morphological characteristics, and is repeatedly identifiable. For example, when the work object is human temporal bone tissue, the positioning reference structure includes at least one of: a round window niche, a conical eminence, an incus-stapedius joint, and a facial nerve recess.

[0069] Step S1's core design addresses the technical problems in existing implantation point localization methods, including insufficient 3D modeling accuracy, subjective and disordered selection of localization reference structures, inaccurate standard implantation point pose analysis, lack of unified standards for localization markers, and poor integration with subsequent operations. Existing technologies often employ contact measurement or single-dimensional data acquisition methods to obtain data on the target area of ​​the work object, which easily generates data noise and fails to accurately reconstruct the object's physical structure. The lack of clear standards for selecting localization reference structures leads to localization reference deviations, and the analysis of implantation point pose parameters is incomplete and lacks standardized verification. Furthermore, the localization markers are disconnected from parameter information, thus affecting the accuracy of subsequent implantation path planning and intraoperative guidance. Therefore, this step uses non-contact optical scanning to acquire multimodal optical 3D data, combined with multi-step data purification and reconstruction processing, to ensure that the 3D solid model accurately reflects the object's physical structure, providing reliable data for subsequent localization. Digitalized reference; by pre-setting standardized implantation operation specifications, the specific requirements for the selection of positioning reference structures, pose accuracy, and structural adaptation are clearly defined, avoiding the subjectivity of positioning reference structure selection. Through global positioning reference structure extraction, standardized coordinate system establishment, comprehensive pose parameter analysis, and consistency verification, the pose parameters of standard implantation points are accurately obtained. Then, through the binding and storage of exclusive positioning identifiers and pose parameter sets, standardized positioning marks are completed. This ensures the accuracy, standardization, and repeatability of standard implantation point positioning, and achieves smooth connection between positioning marks and subsequent implantation path planning and intraoperative operation guidance. At the same time, it takes into account the diversity of operation objects, adapting to the implantation positioning needs of different types of operation objects such as human biological tissues and precision mechanical parts. This lays the foundation for improving the accuracy of the entire implantation point optimization method and effectively avoids problems such as implantation operation errors and low operation efficiency caused by positioning deviations.

[0070] Preferably, the specific steps for locating and marking standard implantation points on a 3D solid model include:

[0071] A full-view optical 3D scan is performed on the target area of ​​the object to be implanted, and multimodal optical 3D data of the target area is acquired. The multimodal optical 3D data includes at least one of high-density 3D point cloud data, surface texture feature data, and geometric contour feature data.

[0072] The acquired multimodal optical 3D data undergoes spatial dimension unification processing to remove discrete noise points and redundant data, completing the initial purification of the multimodal optical 3D data. The preliminarily purified multimodal optical 3D data is then globally registered according to spatial positional correspondences, fusing multimodal optical 3D data acquired from different perspectives into a unified dataset under a unified spatial coordinate system. Anomalies are then identified and removed from the globally registered unified dataset to eliminate abnormal data that deviates from the true structural distribution, ensuring the structural authenticity of the 3D data.

[0073] Spatial meshing is performed on the complete dataset after removing outliers to construct an initial mesh structure that matches the surface morphology of the work object. The initial mesh structure is then subjected to surface fitting and smooth reconstruction to generate a 3D solid model that accurately reflects the physical structure of the work object. This 3D solid model serves as a unified digital benchmark for subsequent implantation point planning and work guidance.

[0074] The pre-defined implantation operation specifications are standardized technical guidelines for implantation operations on the target object. These specifications include structural adaptation requirements, pose accuracy requirements, and requirements for selecting positioning reference structures. Structural adaptation requirements are quantitative constraints on the geometry and spatial structure of the implantation point area, specifically including: maintaining a constant surface curvature in the implantation point area; the height difference between adjacent sampling points within the area being within a pre-defined controllable range, which is the allowable fluctuation range based on historical qualified samples, with the 95th percentile as the upper limit; exceeding this range indicates geometric discontinuity in the area; there are no geometric discontinuities within the area; and the relative coordinates of each spatial sub-unit within the area follow a regular arrangement rule. There is no coordinate overlap or spatial path interference between sub-units; specifically, historical qualified samples are full-process operation data that have been verified as qualified in the final operation effect, without safety accidents, and in compliance with industry standards under the corresponding operation scenario; for each quantitative constraint parameter embedded in the operation specification, the actual detection value of the parameter in all historical qualified samples is extracted to form a sample dataset; using the Grubbs criterion, outlier data with a significance level below 0.05 is removed, and valid historical qualified sample data is retained; the valid historical qualified sample data are sorted in ascending order, the 95th quantile position is calculated, and the 95th quantile value is calculated using linear interpolation. The calculated 95th quantile is set as the upper limit of the allowable fluctuation range of the parameter. When the actual parameter value is ≤95th quantile, it is judged to meet the specification requirements. The pose accuracy requirements are quantitative constraints on the stability of spatial coordinate analysis, pose calculation, and parameter extraction in the implantation operation specifications. Specifically, these include: the spatial coordinate analysis deviation of the standard implantation point is within a preset controllable range; the repetition deviation of coordinate information extraction remains constant; the calculation deviation of the implantation normal pose angle meets preset constraints; and the deviation between the pose analysis result and the actual spatial position is within an allowable range. The positioning reference structure selection requirements are quantitative constraints on the spatial and geometric attributes of the positioning reference structure area in the implantation operation specifications. Specifically, these include: the spatial position repetition recognition deviation of the positioning reference structure area is within a preset controllable range; the repetition rate of geometric contour feature extraction meets a preset standard, which is a threshold for repetition recognition rate or feature similarity, set based on historical successful positioning case statistics; and the geometric and textural differentiation between the positioning reference structure and the surrounding area meets preset judgment conditions, which are dual threshold constraints of geometric shape difference and texture feature difference, ensuring that the reference structure can be accurately distinguished from the surrounding area.

[0075] For example, this implantation procedure specification is designed for cochlear implantation surgery on human temporal bone tissue. All parameters are determined based on clinical anatomical standards, industry standards for cochlear implantation surgery, and statistical analysis of qualified clinical surgical samples. Specific contents include:

[0076] ①Structural adaptation requirements: The surface curvature change rate of the implantation site area is ≤0.2mm / mm, and the height difference between adjacent sampling points is ≤0.1mm; there are no geometric discontinuities or cortical bone defects in the area, and the bone wall thickness is ≥1.5mm; the minimum safe distance between the implantation channel and the facial nerve, horizontal semicircular canal, and round window is ≥2.0mm, and there is no spatial path interference.

[0077] ②Position accuracy requirements: standard implantation point spatial coordinate resolution deviation ≤ 0.3mm, coordinate extraction repetition deviation ≤ 0.15mm; implantation normal attitude angle calculation deviation ≤ 1°, deviation from actual spatial position ≤ 0.3mm; implantation path axial angle deviation from the normal of the circular window niche plane ≤ 2°.

[0078] ③ Safety and obstacle avoidance requirements: The minimum safe avoidance distance between the implantation path and the facial nerve recess and chorda tympani nerve is ≥1.0mm; the minimum safe distance between the drilling area and the inner ear structure is ≥1.5mm; there are no important anatomical structures obstructing the working area, and the requirements for the degree of freedom of operation of surgical instruments are met.

[0079] ④ Requirements for selecting the positioning reference structure: The spatial position repetition error of the reference structure is ≤0.2mm, and the contour feature extraction repetition rate is ≥99%; the difference in geometric shape with the surrounding area is ≥0.5mm, and the difference in texture features is ≥30 gray values; round window niches, conical protrusions, and anvil-stirrup joints are preferred as reference structures.

[0080] This specification for the installation of precision mechanical fasteners is designed for aerospace precision mechanical titanium alloy fastener installation and assembly scenarios. All parameters are determined based on assembly industry standards and statistical analysis of qualified assembly samples. Specific details include:

[0081] ①Structural adaptation requirements: The surface curvature change rate of the implantation site area is ≤0.05mm / mm, and the height difference between adjacent sampling points is ≤0.02mm; there are no geometric breaks, surface scratches, or stress concentration areas in the area, and the uniformity deviation of the substrate wall thickness is ≤5%; the coaxiality deviation between the implantation hole and the surrounding assembly hole is ≤0.01mm, and there is no spatial interference.

[0082] ②Position accuracy requirements: standard implantation point spatial coordinate resolution deviation ≤ 0.02mm, coordinate extraction repeatability deviation ≤ 0.01mm; implantation normal attitude angle calculation deviation ≤ 0.5°, deviation from actual spatial position ≤ 0.02mm; fastener implantation axial perpendicularity deviation from assembly reference surface ≤ 0.03mm / m.

[0083] ③ Safety and obstacle avoidance requirements: The minimum safe clearance between the implantation path and surrounding precision components is ≥0.1mm; the minimum safe distance between the pressing operation area and the sealing surface and mating surface is ≥0.2mm; there is no assembly interference in the operation area, and the motion freedom requirements of the pressing tool are met.

[0084] ④ Requirements for selecting positioning reference structures: The spatial position repetition recognition deviation of the reference structure is ≤0.01mm, and the contour feature extraction repetition recognition rate is ≥99.5%; the difference in geometric shape with the surrounding area is ≥0.05mm, and the difference in texture features is ≥20 gray values; the assembly reference hole, positioning pin hole, and reference end face are preferred as reference structures.

[0085] Based on the preset implantation operation specifications, the global positioning reference structure is extracted from the 3D solid model. The global positioning reference structure extraction involves analyzing the surface geometric contour, texture distribution, structural transitions, and concave and convex features of the 3D solid model point by point. By comparing the global positioning reference structure point by point, structural regions that meet the requirements for the selection of positioning reference structures in the implantation operation specifications are selected. The structural regions are continuous regions in the 3D solid model with fixed spatial positions, significant geometric distinctions, and stable shapes. Their geometric parameters and texture features are unique in the 3D solid model and can be used as references for implantation positioning.

[0086] The selected structural regions are analyzed for geometric features to extract their geometric center, contour boundary, and normal direction. The geometric center of the structural region is determined as the spatial reference origin, and a standard spatial rectangular coordinate system is established for the implantation operation. The standard spatial rectangular coordinate system is a dedicated coordinate system used only for the positioning of the implantation point, path planning, and operation guidance in this operation, and its spatial orientation is consistent with the implantation operation logic.

[0087] The established standard spatial rectangular coordinate system is calibrated axially to clarify the spatial orientation of the X, Y, and Z axes. The Z axis is aligned with the preset implantation direction, the X axis is aligned with the tangent direction of the surface of the target area to be implanted, and the Y axis is derived from the X and Z axes based on the right-hand coordinate system rule. At the same time, the scale accuracy of the standard spatial rectangular coordinate system is calibrated to ensure that the positioning accuracy of the coordinate system meets the pose accuracy requirements in the implantation operation specifications.

[0088] Based on the pre-defined structural adaptation and pose accuracy requirements in the implantation operation specifications, the target area of ​​the 3D solid model to be implanted is partitioned and analyzed within a standard spatial rectangular coordinate system. Target point areas that meet the structural adaptation and pose accuracy requirements are identified. These target point areas are candidate areas in the 3D solid model that meet the conditions for implantation operations, and their geometric dimensions and surface morphology match the operational requirements of the implantation operation. Specifically, using the standard spatial rectangular coordinate system as the analytical reference, the target area to be implanted is spatially partitioned. Structural continuity is detected in each partitioned sub-region, and sub-regions that meet the surface continuity conditions are retained. These retained sub-regions are then determined as the target point areas.

[0089] A full-area scan and analysis of the target point area is performed to collect the spatial coordinate information of each point within the target point area. Coordinate clustering analysis is used to eliminate edge points deviating from the core range of the target point area, thus filtering out the core point set of the target point area. For example, supplementary quantitative constraints are added: K-means coordinate clustering analysis is used to calculate the core cluster center of the target point area. The core range is defined with the core cluster center as the origin and a radius of 1.5 times the allowable deviation of pose accuracy in the embedded operation specifications as the radius. Edge points exceeding the core range are eliminated to obtain the core point set of the target point area. The average coordinates of this core point set are calculated as the target spatial coordinates corresponding to the target point area.

[0090] The normal direction of the core point set of the target point area is calculated based on the surface morphology of the target point area. Combined with the pose accuracy requirements specified in the implantation operation specifications, the angle of this normal direction is calibrated to obtain the implantation normal attitude angle information corresponding to the target point area. This implantation normal attitude angle information is used to determine the tilt direction and angle of the execution end during the implantation operation. The normal direction of the core point set of the target point area is the surface geometric normal, an objective geometric attribute calculated solely based on the surface morphology. It only reflects the local spatial orientation of the surface and does not consider the actual engineering constraints of the implantation operation. The implementation of the implantation operation not only requires matching the surface geometry but also must meet the safety obstacle avoidance, operational accessibility, structural adaptation, assembly, and anatomical function requirements in the implantation operation specifications. The original geometric normal often cannot directly adapt to these constraints. For example, in the cochlear implantation scenario, the geometric normal may point to key anatomical structures such as the facial nerve or inner ear; direct implantation along this direction would cause serious medical risks. In precision assembly scenarios, the geometric normal may not match the perpendicularity requirements of the assembly reference plane; direct implantation would lead to fastener misalignment and assembly stress concentration. Therefore, the angle calibration of the present invention is based on the original geometric normal and uses the quantitative constraints of the implantation operation specifications as boundary conditions to make compliant corrections to the normal direction, ensuring that the corrected implantation normal simultaneously meets the requirements of surface fitting, operation safety, and accuracy, and has clear physical meaning and engineering application value.

[0091] Specifically, the steps for calculating the normal direction of the core point set of the target point area based on the surface morphology of the target point area, and then calibrating the angle of this normal direction in accordance with the pose accuracy requirements specified in the implantation operation specifications, include:

[0092] Within the established standard spatial rectangular coordinate system, principal component analysis is performed on the core point set of the target point region to complete the calculation and standardization of the original geometric normal, ensuring that the initial value of the normal accurately reflects the objective geometric properties of the surface. The specific execution process is as follows:

[0093] Suppose the core point set contains m valid sampling points, and the three-dimensional coordinates of each point in the standard Cartesian coordinate system are: ,in, Construct an m×3 coordinate matrix M, where each row of the matrix corresponds to the three-dimensional coordinates of a sampling point;

[0094] Calculate the mean coordinates of the core point set. ,in For all sampling points The arithmetic mean of the coordinates. , Similarly, subtract the mean point from each row of the coordinate matrix M. The corresponding coordinates are used to obtain the decentralized coordinate matrix. This eliminates the interference of overall coordinate offset on normal calculation;

[0095] Calculate the decentralized coordinate matrix covariance matrix For the covariance matrix Eigenvalue decomposition yields three eigenvalues ​​sorted by their numerical values. and the orthogonal unit eigenvectors corresponding to each eigenvalue. , , ;

[0096] Minimum eigenvalue Corresponding feature vector That is, the normal vector of the local tangent plane of the surface where the core point set is located, denoted as the original geometric normal. ;right After normalization, the original geometric normal vector of the unit is obtained. Satisfying the modulus length constraint ;

[0097] Based on the implantation operation reference direction specified in the implantation operation specifications, the direction of the original geometric normal is checked. If the normal direction is opposite to the preset implantation operation direction, then... Inverting the normal ensures that the initial orientation of the normal is consistent with the operational logic of the implantation operation, thus completing the standardization of the original geometric normal.

[0098] Based on the pre-defined implantation operation specifications, structural adaptation requirements, pose accuracy requirements, and positioning reference structure selection requirements are extracted to construct constraints, ensuring that the calibration process fully meets the safety and accuracy requirements of the operation. For example, the minimum safe avoidance distance between the implantation path and the obstacle spatial contour is extracted from the implantation operation specifications. Construct constraints: Modified normal The minimum spatial distance between the corresponding straight implantation path and the marked obstacle spatial contour in the 3D solid model. ;

[0099] Extract the maximum allowable angle between the implantation normal and the operation reference axis from the implantation operation specifications. And the maximum solution deviation limit of the implanted normal attitude angle, and construct the constraint conditions: the corrected normal Spatial angle with the implantation reference axis ;

[0100] Extract the maximum normal correction angle limit corresponding to the pose accuracy from the implantation operation specifications. Construct constraints: Modified normal With the original geometric normal spatial angle This ensures that the calibrated normal conforms to the original geometric properties of the surface to the greatest extent possible, avoiding surface fit deviation and implantation stability issues caused by over-correction.

[0101] With the goal of minimizing the angular deviation from the original geometric normal, a constrained minimum angular deviation calibration model is constructed based on the constraints established in step S12. The optimal corrected normal is then solved. The specific process is as follows:

[0102] With the core objective of minimizing the spatial angle between the corrected normal and the original geometric normal, an optimization objective function is constructed as follows:

[0103]

[0104] in, Let be the unit calibration normal vector to be solved, satisfying the unit magnitude constraint. ;

[0105] The constrained minimum angle deviation calibration model is solved, and the iteration convergence threshold is set to 0.1 times the allowable deviation of pose accuracy in the implantation operation specification to ensure that the solution accuracy meets the operation requirements. During the iteration process, each step checks whether the current solution meets all constraints, and invalid solutions that do not meet the constraints are directly eliminated.

[0106] When the iterative process meets the convergence threshold and the obtained normal vector satisfies all constraints, the unit normal vector is determined as the final calibrated normal vector. If a valid solution satisfying all constraints is not obtained after iterating to the preset maximum number of times, the alternative insertion point re-screening process is immediately triggered to prevent invalid insertion points from entering the subsequent process.

[0107] Within a standard Cartesian coordinate system, based on the calibrated unit normal vector The specific formula for calculating the implanted normal attitude angle information is as follows:

[0108] Pitch angle , used to characterize the tilt angle of the implantation execution end in the vertical plane;

[0109] Yaw angle , used to characterize the deflection angle of the implantation execution end in the horizontal plane;

[0110] Among them, pitch angle With yaw angle Together, they constitute complete implantation normal attitude angle information, defining the spatial tilt direction and angle of the implantation operation execution end.

[0111] The normal direction of the core point set is projected into a standard spatial rectangular coordinate system. The angle of the normal direction is corrected according to the axis of the coordinate system to make the corrected direction match the implantation operation direction. The corrected direction parameter is determined as the implantation normal attitude angle information.

[0112] According to the structural adaptation requirements in the implantation operation specifications, the surface morphology of the target point area is analyzed for flatness and area. The continuous surface area of ​​the target point area is measured, and continuous surfaces within the target point area whose flatness meets the implantation operation specifications are selected as effective working surfaces. Specifically, flatness is quantified by the surface curvature change rate, the height difference between adjacent sampling points, and the proportion of geometric discontinuities. Only continuous sub-regions with constant curvature change, height differences within a preset controllable range, and no geometric discontinuities are retained. These sub-regions are then divided and merged into connected candidate surfaces through region growth. After filtering out small areas, the candidate surface with the largest area and closest to the target implantation point is selected as the effective working surface. The boundary range and geometric parameters of the effective working surface are determined to form the effective working surface parameter information corresponding to the target point area. The effective working surface parameter information is used to clarify the contact range and contact posture between the operation execution end and the operation object during the implantation operation.

[0113] The target spatial coordinates, implantation normal attitude angle information, and effective working surface parameter information corresponding to the target point area are associated and integrated to form a set of pose parameters corresponding to the standard implantation point. The consistency of the pose parameter set is checked, the degree of conformity between each parameter and the implantation operation specification is compared, logical conflicts between parameters are eliminated, and it is ensured that the pose parameter set fully meets the requirements of the implantation operation specification.

[0114] In the 3D solid model, based on the set of pose parameters corresponding to the standard implantation point, the spatial point corresponding to the standard implantation point is accurately located, and a unique positioning mark is drawn at the spatial point. The unique positioning mark is a unique 3D mark graphic that can be significantly distinguished from other areas in the 3D solid model.

[0115] The pose parameters of the standard implantation point are bound and stored with the drawn exclusive positioning markers to establish a one-to-one correspondence between the parameters and the markers. This facilitates the quick retrieval of the complete pose parameter information of the standard implantation point through the exclusive positioning markers during subsequent implantation path planning, intraoperative operation guidance, and point verification, thus completing the positioning marking of the standard implantation point in the three-dimensional solid model.

[0116] Step S2: Based on the preset implantation operation specifications, construct an implantation point quality quantitative evaluation model. Substitute the pose parameter set corresponding to the standard implantation point obtained in Step S1 into the implantation point quality quantitative evaluation model to complete the multi-dimensional quantitative quality evaluation of the standard implantation point and obtain the comprehensive quality score of the standard implantation point. Compare the comprehensive quality score with the preset qualified threshold to determine whether the standard implantation point meets the compliance requirements. If the standard implantation point meets the compliance requirements, generate a compliance pass instruction and enter the subsequent implantation path planning process. If the standard implantation point does not meet the compliance requirements, generate a compliance fail instruction and enter the alternative implantation point planning and anomaly warning process.

[0117] In this embodiment, the implantation point quality quantitative evaluation model is a mathematical model established based on the implantation operation specifications, used for multi-index quantitative weighted calculation of standard implantation points. It includes a multi-dimensional evaluation index system corresponding one-to-one with the implantation operation specifications, weight coefficients corresponding to each evaluation index, quantitative assignment rules, and qualified thresholds. The multi-dimensional evaluation index system is a set of quantitative evaluation indicators corresponding one-to-one with the structural adaptation requirements, pose accuracy requirements, and positioning reference structure selection requirements in the implantation operation specifications. For example, when the work object is human temporal bone tissue used for cochlear implantation surgery, the multi-dimensional evaluation index system includes at least one of the following: safe distance between the standard implantation point and key anatomical structures, integrity of bone structure in the implantation area, patency of the implantation channel, and accessibility of the implantation operation space. For example, when the work object is a precision mechanical part, the multi-dimensional evaluation index system includes at least one of the following: material strength, wall thickness uniformity, processing stress concentration, assembly coaxiality matching, and processing path accessibility in the area where the standard implantation point is located.

[0118] Step S2's core design addresses the technical problems in existing implantation site assessment methods, including reliance on human experience for quality assessment, lack of unified quantitative standards, single assessment dimensions, vague compliance judgment rules, and missing differentiated processing procedures. Current technologies largely rely on operators' subjective experience to judge the usability of implantation sites, lacking a standardized quantitative assessment system. This fails to comprehensively cover multiple dimensions such as structural fit, safety obstacle avoidance, and operational feasibility. Compliance judgments lack clear boundaries and do not establish corresponding differentiated processing procedures for different judgment results. This easily leads to unqualified implantation sites entering subsequent operational stages, causing operational errors and structural damage, or qualified implantation sites being misjudged, resulting in increased operational costs and reduced efficiency. Especially in cochlear implantation surgery, subjective assessment bias can easily lead to medical risks such as facial nerve damage and residual hearing loss. In industrial settings, defects can easily lead to production losses such as scrapped parts and assembly failures. Therefore, this step constructs a unified quantitative assessment model for implantation point quality based on the implantation operation specifications preset in step S1. It establishes a multi-dimensional assessment system that corresponds one-to-one with the implantation operation requirements. Through standardized weighted calculations, it achieves objective quantitative assessment of implantation point quality. Combined with clear qualification thresholds, it completes compliance judgments and sets corresponding differentiated processing procedures for compliant and non-compliant results. This ensures the objectivity, comprehensiveness, and repeatability of implantation point quality assessment, while also achieving precise diversion and control of the implantation operation process, preventing unqualified implantation points from entering subsequent operation stages. Simultaneously, it adapts to the assessment needs of different types of operation objects, providing core support for the safety, stability, and operational accuracy of the entire implantation point optimization method, effectively avoiding operational risks and cost losses caused by implantation point quality defects.

[0119] Please see Figure 2 Preferably, the specific steps for determining whether a standard implantation site meets compliance requirements include:

[0120] Based on the implantation operation specifications preset in step S1, a quantitative evaluation model for implantation point quality is constructed, clarifying the multi-dimensional evaluation index system required for implantation point quality evaluation. This multi-dimensional evaluation index system forms a one-to-one correspondence with the structural adaptation requirements, pose accuracy requirements, and positioning reference structure selection requirements in the implantation operation specifications. For example, the evaluation indicators corresponding to the structural adaptation requirements are: the degree of constant surface curvature of the implantation point area, the compliance of the height difference between adjacent sampling points, regional geometric continuity, the regularity of spatial sub-unit arrangement, and the degree of non-overlapping interference of spatial sub-units. The evaluation indicators corresponding to the pose accuracy requirements are: the degree of spatial coordinate analysis deviation, the repeatability stability of coordinate information extraction, the accuracy of implantation normal attitude angle calculation, and the degree of matching between the pose analysis results and the actual spatial position. The evaluation indicators corresponding to the positioning reference structure selection requirements are: the repeatability stability of the spatial position recognition of the positioning reference structure, the reliability of the geometric contour extraction of the positioning reference structure, the distinguishability between the positioning reference structure and the surrounding area, and the degree of spatial correlation stability between the positioning reference structure and the standard implantation point. The implantation site quality quantitative assessment model, based on the implantation operation specifications preset in step S1, is a mathematical operation model that realizes multi-dimensional and quantitative assessment of the quality of standard implantation sites. Its core components include a multi-dimensional assessment index system, quantitative assignment rules for each assessment index, weight coefficients, and qualification thresholds. Each component corresponds one-to-one with the quantitative constraints of the implantation operation specifications, ensuring that the assessment results can truly reflect the degree of fit between the standard implantation site and the operation specifications.

[0121] For each evaluation indicator in the multi-dimensional evaluation indicator system, a corresponding quantitative assignment rule is set. The quantitative assignment rule is used to convert the actual parameters of the evaluation indicator into standardized quantitative scores, and at the same time, a corresponding weight coefficient is set for each evaluation indicator. The quantitative assignment rule is a standardized parameter-to-score conversion rule, which is used to convert the actual detection parameters corresponding to each evaluation indicator into standardized quantitative scores from 0 to 100. The conversion logic strictly follows the quantitative constraints in the implantation operation specification: for positive evaluation indicators, when the actual parameters fully comply with the implantation operation specification constraints, a score of 100 is assigned; when the actual parameters exceed the specification constraint range, a score of 0 is assigned; when the actual parameters are within the specification constraint range but have not reached the optimal state, the score is linearly allocated according to the deviation ratio between the actual parameters and the optimal parameters in the specification. For example, the quantitative scoring rules for structural adaptation evaluation indicators are as follows: Regarding the constancy of surface curvature in the implantation site area, a score of 100 is assigned when the surface curvature remains completely constant; a score of 0 is assigned when the curvature change exceeds the controllable range specified in the standard; and a score is deducted based on the proportion of the curvature change deviation to the controllable range specified in the standard when the curvature change falls between these two ranges. Regarding the compliance of height difference between adjacent sampling points, a score of 100 is assigned when the height difference is completely within the preset controllable range specified in the standard; a score of 0 is assigned when the height difference exceeds the range; and a score is deducted based on the proportion of the height difference deviation to the controllable range specified in the standard when the height difference falls between these two ranges. The following are the deduction criteria: For the geometric continuity of a region, 100 points are awarded when there are no geometric discontinuities within the region, and 0 points are awarded when discontinuities exist. For the regularity of the spatial sub-unit layout, 100 points are awarded when the relative coordinates of the sub-units completely follow the standard regularity layout rules, and 0 points are awarded when the layout deviation exceeds the allowable range of the standard. If the deviation is between the two, points are deducted proportionally based on the layout deviation. For the degree of non-overlapping interference of spatial sub-units, 100 points are awarded when there is no coordinate overlap or spatial path interference between the sub-units, and 0 points are awarded when there is overlap or interference.

[0122] The weighting coefficients are standardized quantitative coefficients, ranging from 0 to 1. The sum of the weighting coefficients of all evaluation indicators is 1. They are used to clarify the proportion of influence of each evaluation indicator in the comprehensive quality evaluation. Their setting logic is consistent with the implantation operation specifications preset in step S1. They are determined according to the influence level of each evaluation indicator on the implantation operation: evaluation indicators directly related to the safety and core accuracy of the implantation operation are assigned higher weighting coefficients. For example, the analytic hierarchy process is used to compare and score each evaluation indicator pairwise, construct a judgment matrix and perform a consistency test. After the test is passed, the feature vector of each indicator is calculated, and the feature vector is normalized and used as the weighting coefficient of each evaluation indicator to ensure the scientific and reasonable weighting allocation.

[0123] Based on the implantation operation specifications and actual needs, a pass threshold corresponding to the comprehensive quality score is set. This pass threshold is a standardized score ranging from 0 to 100 points, used to determine whether the standard implantation point meets the compliance requirements of the implantation operation. It is set based on the minimum compliance standards of each quantitative constraint in the implantation operation specifications, combined with industry operation accuracy requirements and the safety threshold of the work object. The work object safety threshold refers to the maximum permissible critical value of various operation parameters and structural deviations set to avoid damage to the work object. It is preset based on the work object's tolerance limit, industry safety standards, and historical safety data statistics. Furthermore, the pass threshold must be consistent with the quantitative assignment rules and the setting logic of the weighting coefficients. For example, the pass threshold is set to 70 points. That is, when the comprehensive quality score of the standard implantation point is ≥70 points, it is determined to meet the compliance requirements; when the comprehensive quality score is <70 points, it is determined to not meet the compliance requirements.

[0124] The set of pose parameters and the 3D solid model corresponding to the standard implantation point obtained in step S1 are retrieved. The actual detection parameters corresponding to each indicator of the multi-dimensional evaluation index system are extracted to complete the collection and organization of the basic evaluation parameters. Next, according to the quantitative assignment rules corresponding to each evaluation index, the actual detection parameters are converted into corresponding standardized quantitative scores. Then, the standardized quantitative score of each evaluation index is multiplied by the corresponding weight coefficient to obtain the weighted score of each index. Finally, the weighted scores of all evaluation indicators are summed to obtain the comprehensive quality score corresponding to the standard implantation point, thus completing the multi-dimensional quantitative quality evaluation of the standard implantation point.

[0125] The calculated comprehensive quality score is compared with the preset qualification threshold. If the comprehensive quality score is greater than or equal to the qualification threshold, the standard implantation point is deemed to meet the compliance requirements, a compliance approval instruction is generated, and the subsequent implantation path planning process is triggered.

[0126] If the overall quality score is less than the passing threshold, the standard implantation point is deemed non-compliant, a compliance failure instruction is generated, and the subsequent alternative implantation point planning and anomaly warning process is triggered. For example, based on a 3D solid model, preset implantation operation constraints, and an evaluation index system, the system searches within a preset spatial search range, centered on the standard implantation point, for a set of alternative implantation points that meet structural constraints, safety constraints, and operational requirements. Each alternative implantation point in the set is then subjected to a quality quantification evaluation, sorted from highest to lowest overall quality score, and the highest-scoring optimal alternative implantation point, along with its corresponding 3D coordinates and attitude parameters, is output. Simultaneously, anomaly warning information is generated, marking the specific defects and their severity in the 3D solid model. If no alternative implantation point conforms to the implantation operation specifications within the spatial search range, an implantation operation prohibition warning is generated, terminating the current implantation planning process.

[0127] Step S3: If the standard implantation point is deemed compliant, the target spatial coordinates corresponding to the standard implantation point located in Step S1 are taken as the endpoint of the implantation path, and the initial entry point corresponding to the implantation operation is taken as the starting point of the implantation path. Simultaneously, a set of auxiliary operation points matching the operation type to which the standard implantation point belongs is retrieved, and according to the process space mapping rules of the auxiliary operation point set and the standard space rectangular coordinate system corresponding to the 3D solid model, the spatial coordinate matching and labeling of each auxiliary operation point are completed in the 3D solid model. Combining the process execution order of the auxiliary operation points and the structural constraints of the 3D solid model, the segmented planning and global optimization of the implantation path are completed, with the starting point of the path as the beginning and each labeled auxiliary operation point as a necessary intermediate node, to generate the implantation path.

[0128] In this embodiment, the initial entry point refers to the starting spatial position of the implantation execution end entering the target area to be implanted. Its spatial coordinates are determined comprehensively based on the boundary of the target area to be implanted in the three-dimensional solid model and the entry requirements of the implantation operation specification. The auxiliary operation point set refers to the standardized spatial node set that is bound to the standard execution procedure of the implantation operation and is pre-fixed under the corresponding operation type and operation scenario. It is a general predetermined node for the same type of implantation operation. For example, when the operation object is human temporal bone tissue used for cochlear implantation surgery, the corresponding auxiliary operation point set includes bone surface positioning node, drilling start node, drilling end node, implant placement node, and fixation execution node. Each node has built-in spatial mapping rules and connection requirements between the preceding and following procedures corresponding to the surgical procedure. For example, when the operation object is precision mechanical parts used for fastener implantation and assembly, the corresponding auxiliary operation point set includes tooling alignment node, pre-implantation start node, pressing end node, and locking execution node. Each node has built-in spatial mapping rules and connection requirements between the assembly procedure and the procedure.

[0129] Step S3's core design addresses the technical problems in existing implantation path planning and area control, namely, the disconnect between path nodes and predetermined work procedures, the lack of standardized basis for auxiliary operation point labeling, and the lack of linkage between path planning logic and previous implantation point positioning results. Existing technologies largely rely on manual experience for path node labeling and path planning, failing to match standardized auxiliary operation points already established for similar operations. The path planning lacks sufficient matching degree with standard implantation procedures and previously verified implantation point parameters, easily leading to problems such as poor process integration, path space interference, and substandard operational accuracy. Simultaneously, the division of work areas lacks clear quantitative basis and hierarchical control rules, hindering the implementation of implantation operations. Standardized safety management throughout the entire process can easily lead to medical risks such as damage to critical anatomical structures in medical implantation scenarios, and problems such as component collisions and assembly failures in industrial precision assembly scenarios. Based on this, this step uses the standard implantation point that has passed the compliance verification in step S2 as the core benchmark, directly retrieves the set of fixed auxiliary operation points under the corresponding operation type, and completes node annotation based on a unified coordinate system and operation specifications to avoid the subjective bias of manual annotation. Then, the path planning is completed with the path start and end points and standardized auxiliary operation points as necessary nodes to ensure a high degree of matching between the path and the operation procedures and implantation point parameters, providing complete digital support for the accurate and safe execution of implantation operations.

[0130] Please see Figure 3 Preferably, the specific steps for generating the implantation path include:

[0131] Using the target spatial coordinates corresponding to the standard implantation point located in step S1 as the endpoint of the implantation path and the initial entry point corresponding to the implantation operation as the starting point of the implantation path, the three-dimensional solid model, standard spatial rectangular coordinate system, preset implantation operation specifications, pose parameter set corresponding to the standard implantation point verified in step S2, and auxiliary operation point set matching the operation type of the standard implantation point are retrieved simultaneously to complete the preparation of the preliminary basic data for path planning.

[0132] Using the standard Cartesian coordinate system corresponding to the 3D solid model as a reference, and combining the process space mapping rules of the auxiliary operation point set, each auxiliary operation point in the auxiliary operation point set is mapped to the 3D solid model to obtain the target space coordinates of each auxiliary operation point, thus completing the accurate calibration of the 3D space coordinates of each auxiliary operation point. Among them, the process space mapping rule refers to a set of quantitative mapping criteria that are pre-fixed in the corresponding set of auxiliary operation points based on the standardized full-process execution requirements of the implantation operation type to which the operation object belongs. This set of criteria is used to transform abstract process execution nodes into quantifiable three-dimensional spatial positioning and attitude constraints. The rule uses the standard spatial rectangular coordinate system of the standard implantation point as a unified spatial reference, the standard implantation point as the core process anchor point, and the positioning reference structure as a local positioning correction reference. It can not only achieve standardized and repeatable mapping of auxiliary operation points in the three-dimensional solid models of different operation objects under the same type of implantation operation, but also adapt to the individual structural differences of different operation objects through preset parameter floating constraints. Its core components include: preset relative position parameters between each auxiliary operation point and the standard implantation point, the associated positioning reference structure that binds each auxiliary operation point one by one, and the process execution order corresponding to each auxiliary operation point.

[0133] Specifically, the steps for obtaining the target spatial coordinates of each auxiliary operation point include:

[0134] From the process space mapping rules built into the auxiliary operation point set, the preset relative position parameters between each auxiliary operation point and the standard implantation point are read, including relative coordinate offset, relative angle constraint, and relative distance range. The relative position parameters are general standardized rules for implantation operations of the same type and do not change with the structural differences of individual operation objects. Among them, the relative coordinate offset is the preset offset range of the auxiliary operation point relative to the standard implantation point in the X, Y, and Z axes of the standard space rectangular coordinate system; the relative angle constraint is the allowable range of the attitude angle of the auxiliary operation point relative to the implantation normal of the standard implantation point; and the relative distance range is the straight-line distance constraint range between the auxiliary operation point and the standard implantation point.

[0135] Using the target spatial coordinates of the standard implantation point as the process anchor point and its implantation normal attitude angle as the attitude reference, the initial three-dimensional spatial coordinates of each auxiliary operation point in the current three-dimensional solid model are calculated in the standard spatial rectangular coordinate system according to the relative position parameters corresponding to each auxiliary operation point. During the calculation process, it is ensured that the initial coordinates are within the relative distance range and the initial pose meets the relative angle constraint, thus completing the initial mapping of the auxiliary operation points. Simultaneously, the corresponding process content, accuracy requirements, and associated positioning reference structure information are bound to each auxiliary operation point.

[0136] The structural features of the associated positioning reference structure for each auxiliary operation point are obtained. Using the initial three-dimensional spatial coordinates of each auxiliary operation point as the center of a sphere, a basic search radius is set based on its structural features. For example, a spherical search area is delineated using twice the overall contour size of the core geometric positioning features of the associated positioning reference structure as the basic search radius. Specifically, structural features refer to the set of quantitative geometric attributes and spatial morphological parameters possessed by the associated positioning reference structure, which can be used for precise spatial positioning, high repeatability recognition, and pose deviation verification. These are the core positioning reference for correcting the initial three-dimensional spatial coordinates of the auxiliary operation points through spatial search. All parameters possess uniqueness, stability, and reproducibility in the three-dimensional solid model and do not change during the operation execution process. For example, structural features include the core geometric positioning features and morphological features of the associated positioning reference structure. Stability characteristics and spatial correlation constraint characteristics; for example, when the work object is human temporal bone tissue used for cochlear implantation surgery, the associated positioning reference structure bound to the drilling start node is a conical ridge, and its corresponding structural features include: the three-dimensional coordinates of the geometric center of the conical ridge, the coordinates of the ridge contour boundary, the surface normal vector of the conical ridge, the fixed relative position with the circular window niche, the minimum safe distance with the facial nerve recess, and the axial pointing parameter of the ridge matching the drilling process; when the work object is a precision mechanical part used for fastener implantation assembly, the associated positioning reference structure bound to the pre-implantation start node is an assembly reference hole, and its corresponding structural features include: the three-dimensional coordinates of the center of the reference hole, the coordinates of the hole wall contour boundary, the hole center axis vector, the inner diameter of the reference hole, the perpendicularity parameter with the assembly positioning surface, and the hole coaxiality parameter matching the pre-implantation process.

[0137] Because different work objects exhibit individual structural differences, local morphological deviations, and surface topological features in their 3D solid models, it is impossible to directly use fixed initial mapping coordinates to meet the requirements for precise positioning and process execution. Therefore, it is necessary to use an associated positioning reference structure as a reference, and through spherical region search and structural feature similarity matching, to achieve adaptive correction of the spatial coordinates of auxiliary operation points, thereby adapting to the personalized structural features of the work objects and improving positioning accuracy and process execution reliability. Within the spherical search area, the actual structural features of each auxiliary operation point are extracted point by point, and the similarity between the actual structural features and the structural features of the corresponding associated positioning reference structure is calculated. For example, the similarity calculation uses the core geometric positioning features and morphological stability features in the structural features as the core comparison indicators, and uses the Euclidean distance method to quantify the degree of fit between the two in geometric parameters and spatial morphology. The similarity value ranges from 0 to 1. The closer the value is to 1, the higher the degree of fit between the local geometric features of the candidate point and the structural features of the associated positioning reference structure, and the more it meets the process execution positioning requirements of the auxiliary operation point.

[0138] Based on the similarity score, the candidate point with the highest similarity value within the spherical search area is selected, and its three-dimensional spatial coordinates are used as the target spatial coordinates of the auxiliary operation point. Simultaneously, it is verified whether the candidate correction parameters meet the preset relative coordinate offset range, relative angle constraints, and relative distance range in the process space mapping rules, as well as the requirements for safety spacing and structural adaptation in the embedded operation specifications. If the verification passes, the initial three-dimensional spatial coordinates of the auxiliary operation point are updated to the target spatial coordinates, completing the positioning correction of a single node. If the verification fails, the basic search radius of the spherical search area is gradually reduced by 5%, down to a minimum of 50% of the basic search radius. If correction parameters that meet all constraints are still not obtained, a manual review process is immediately triggered, and the operator manually adjusts the coordinates of the auxiliary operation point.

[0139] Based on the execution sequence of each auxiliary operation point, the starting point of the implantation path, each auxiliary operation point, and the ending point of the implantation path are sorted to form a node sequence for implantation path planning. The process requirements, pose accuracy requirements, and connection rules of each node in the sequence are clarified to ensure that the spatial direction of the node sequence is completely matched with the execution logic of the implantation operation, providing fixed node constraints for subsequent path planning.

[0140] Based on the standard spatial rectangular coordinate system, a cuboid space covering all auxiliary operation points, the endpoint of the implantation path, and the preset operation range is constructed as the implantation operation global space. The preset operation range is determined comprehensively based on the maximum motion stroke of the implantation operation execution end and the overall structural dimensions of the object being operated on.

[0141] Based on the pose accuracy requirements in the implantation operation specifications, the entire implantation operation space is discretized into a three-dimensional, equal-size grid. For example, the side length of a single grid cell is set to be no greater than half of the maximum allowable deviation of pose accuracy in the implantation operation specifications. This balances the accuracy of path planning with computational efficiency. The higher the pose accuracy requirement, the closer the grid side length is to one-quarter of the maximum allowable deviation. The entire implantation operation space is decomposed into several grid cells with spatial coordinate codes, providing a unified spatial carrier for the subsequent quantitative marking of various constraint elements.

[0142] Based on the target spatial coordinates of each auxiliary operation point and the standard implantation point, and combined with the side length of a single grid cell and the average spacing between adjacent nodes in the node sequence, a calculation kernel width adapted to the rasterized spatial resolution is determined. For example: First, the straight-line distances of all adjacent nodes in the node sequence are calculated, and their average value is taken as the average node spacing. Combined with the side length of a single grid cell, the base value of the calculation kernel width is set as the weighted average of the average node spacing and the grid side length. The weights are allocated according to the pose accuracy requirements of the implantation operation; the higher the pose accuracy requirement, the greater the weight of the grid side length. After determining the calculation kernel width, the geometric center of each grid cell in the entire implantation operation space is used as the calculation base point. A Gaussian kernel function is used to iterate through and calculate the density contribution value of all nodes in the node sequence to the grid cell. The standard deviation of the Gaussian kernel function is consistent with the calculation kernel width. The density contribution value of a node to the grid cell decreases in a Gaussian distribution as the distance between the node and the geometric center of the grid increases; the closer the distance, the greater the contribution value. When the distance exceeds the range corresponding to the calculation kernel width, the contribution value is ignored to avoid interference from irrelevant nodes in the grid density calculation. After summing the density contribution values ​​of all nodes to the raster cell, the spatial distribution density value of the nodes corresponding to the raster cell is obtained. This completes the traversal calculation of the density values ​​of all raster cells in the entire domain, providing an accurate quantitative basis for subsequent spatial clustering and partitioning operations.

[0143] Based on a preset quantile threshold for the density value of the entire grid cells, all grid cells are clustered and divided into operational spaces, clearly defining the set of grid cells and the continuous spatial boundaries of each operational space. The operational space refers to a hierarchical spatial region used for standardized path planning, defined around the sequence of implanted operation nodes and adapting to the movement range of the implanted operation execution end, based on the spatial distribution density of nodes. This includes the core operational area, transitional operational area, and peripheral operational area. Furthermore, the minimum safe avoidance distance for each operational space grid cell is assigned based on a positive correlation of node spatial distribution density; that is, the higher the node spatial distribution density value of a grid cell, the greater its minimum safe avoidance distance. This ensures that the safety control level of the core operational area is higher than that of the transitional and peripheral operational areas, providing a quantitative basis for the safety constraints of subsequent path planning. The quantile grading threshold refers to the critical value used to classify the spatial distribution density values ​​of nodes. It is set jointly based on the safety control level requirements in the implantation operation specifications, the importance of the operation area, and the structural risk level of the 3D entity model. For example, the 75th and 50th quantiles are set as grading thresholds. Grid cells with a density value ≥ 75th quantile are classified as core operation areas, grid cells with a density value ≤ 75th quantile are classified as transitional operation areas, and grid cells with a density value < 50th quantile are classified as peripheral operation areas. If the operation is a high-risk precision medical implantation scenario, the grading threshold can be increased to the 80th and 60th quantiles to further narrow the core operation area and improve safety. Control precision; minimum safe avoidance distance refers to the minimum safe distance threshold that the implanted work execution end must maintain between itself and the geometric structure within the three-dimensional solid model during movement; for example, the minimum safe avoidance distance in the core work area is set to 0.5mm to 1.0mm, the minimum safe avoidance distance in the transition work area is set to 0.2mm to 0.5mm, and the minimum safe avoidance distance in the outer work area is set to 0.1mm to 0.2mm. For precision medical implantation scenarios such as human temporal bone tissue, the minimum safe avoidance distance in the core work area can be increased to 1.0mm to 1.5mm. For precision mechanical parts assembly scenarios, the minimum safe avoidance distance in the core work area can be set to 0.05mm to 0.1mm.

[0144] Based on the geometric structure data of the 3D solid model, the 3D isosurface boundaries of the object's solid structure, surrounding tooling fixtures, and prohibited touch areas are extracted as the obstacle space contour. Two adjacent nodes in the node sequence are divided into an independent path planning unit, with the preceding node within the unit as the sub-path start point and the following node as the sub-path end point. Using the obstacle space contour of the 3D solid model as spatial constraints, combined with the minimum safe avoidance distance of each grid unit, path planning is performed to generate sub-paths corresponding to each path planning unit. For example, for scenarios with no complex obstacles in the work space and high requirements for path planning speed, a fast expanding random tree algorithm is used for path planning. For precision implantation and assembly scenarios with dense obstacles in the work space and high requirements for path optimality, the A* algorithm is used for path planning. Simultaneously, the heuristic function of the A* algorithm is optimized, using the straight-line distance between nodes combined with obstacle avoidance costs as heuristic values ​​to improve the accuracy of path planning.

[0145] Connect the sub-paths of all path planning units in the order of node sequence, eliminate coordinate and attitude breakpoints at the connection of adjacent sub-paths, ensure that the spliced ​​path is continuous and uninterrupted, and smoothly transition the pose parameters of each node on the path to obtain the implanted path, and map it into the standard space rectangular coordinate system of the three-dimensional solid model.

[0146] Step S4: Based on the implantation path and node sequence generated in Step S3, update the current target node in real time; continuously collect real-time optical 3D data of the current actual working area through the light source emission and reception structure of the guidance auxiliary device; perform dynamic registration between the real-time optical 3D data and the 3D solid model to complete the standard spatial rectangular coordinate system and deviation calibration between the actual working space and the 3D solid model; based on the registration result, determine whether the effective acquisition space range of the current actual working area completely covers the current target node, so as to perform targeted visual guidance for the target node; specifically, if it is not covered, calculate the required movement direction and movement distance of the guidance auxiliary device and generate the corresponding displacement guidance command; if it is covered, perform targeted visual guidance through the light source emission component of the guidance auxiliary device to locate the target node until the target operating device reaches the standard implantation point, completing the closed-loop guidance of the entire implantation operation.

[0147] In this embodiment, the guidance and assistance device refers to a dedicated device mounted on the implantation operation execution mechanism to provide real-time optical positioning and path guidance for the implantation operation. It can be linked with the 3D solid model and the implantation path in real time. It consists of a support and fixing structure that can be adapted to the implantation operation execution mechanism, a light source emission and receiving structure that integrates structured light projection and multimodal image acquisition, a data transmission structure that realizes coordinate system calibration and bidirectional interaction of full data, and an external processing structure that performs 3D data calculations, logic control, and command generation. Among them, the support and fixing structure provides a stable installation benchmark for the device and can adjust the installation posture according to the shape and degree of freedom of the execution mechanism to adapt to the installation requirements of different implantation scenarios. The light source emission and receiving structure includes a light source emission component and a light source receiving component. The former is used to project structured light patterns and targeted visual guidance marks, and the latter is used to acquire optical images modulated by the surface morphology of the operation area to provide a data source for real-time 3D modeling and dynamic registration. The data transmission structure is used to realize real-time data transmission and spatial coordinate system calibration between various modules of the device, between the device and the host computer, and between the device and the target operating equipment. The external processing structure provides core calculation and control support for the entire process guidance. The target node refers to the spatial node in the implantation path node sequence that is defined according to the execution order of the standard implantation operation procedure and is currently required to complete the positioning and corresponding operation procedure. It is bound one by one to the functional feature points in the three-dimensional solid model, including the initial entry point, each auxiliary operation point, and the standard implantation point. The target node is updated in real time according to the completion verification result of the current procedure.

[0148] Preferably, the specific steps for performing dynamic registration between real-time optical 3D data and a 3D solid model include:

[0149] Preprocessing is performed on the optical 3D data of the current actual working area collected in real time by the guidance and assistance device. Point cloud denoising, downsampling, outlier removal and normal vector calculation are performed in sequence to remove discrete noise points, redundant data and outliers that deviate from the real structure in the original data. Real-time surface features such as surface texture, curvature distribution and geometric contour are extracted from the preprocessed point cloud data. Through adaptive local spatial meshing, surface fitting and smooth reconstruction, a real-time local surface model is constructed to provide a standardized real-time data foundation for subsequent registration.

[0150] By extracting real-time surface features from the 3D solid model, a standard surface model of the 3D solid model is constructed. Point-by-point geometric analysis is performed on the standard surface model to extract edge feature points and curvature extrema points. The 3D coordinates, local geometric attributes, and spatial relationships of each feature point are clarified to form a registration datum feature set. At the same time, coordinate normalization processing is performed on the standard surface model to construct its axis-aligned bounding box and mark the geometric center. Principal component analysis is used to determine the principal inertial axis of the standard surface model with the geometric center as the origin, locking the rigid mapping relationship between the standard surface model and the standard spatial rectangular coordinate system to ensure the spatial consistency of the registration datum. An axis-aligned bounding box refers to the smallest hexahedral spatial bounding structure that completely encloses all geometric elements of a standard surface model, with each side parallel to the corresponding coordinate axis, based on the X, Y, and Z axes of a standard Cartesian coordinate system. Its size is determined based on the difference between the extreme coordinates of the standard surface model in the X, Y, and Z axes. Specifically, the maximum and minimum values ​​of all sampling points of the standard surface model in the X, Y, and Z axes are taken, and the difference between the maximum and minimum values ​​of the three axes is used as the side length of the bounding box for the corresponding axis. This results in the smallest axis-aligned bounding box that can completely enclose the standard surface model without redundancy. In one embodiment, the side lengths of each axis of the axis-aligned bounding box can be extended outward by 10% to 20% proportionally according to the proportion of the overlapping expansion area of ​​the local spatial region to be registered, to ensure complete coverage of the feature matching range.

[0151] The curvature values ​​of each sampling point of the real-time local surface model are calculated, real-time curvature feature points are extracted, and curvature similarity matching is performed between the real-time curvature feature points and the curvature extreme points in the registration reference feature set. The Euclidean distance method is used to quantify the degree of fit between the curvature features and local geometry of the two, and effective curvature feature points with curvature similarity greater than a preset matching threshold are selected. The matching threshold is a critical judgment value used to determine whether the real-time curvature feature points and the curvature extreme points in the registration reference feature set have feature similarity and can be used as effective matching feature points. It is jointly set by the pose accuracy requirements in the implantation operation specifications, the safety level of the operation scenario, and the structural complexity of the operation object. Its value ranges from 0 to 1. The closer the value is to 1, the higher the similarity requirement for feature matching and the stricter the control of registration accuracy. For example, for precision medical implantation scenarios such as human temporal bone tissue, the matching threshold is set to 0.85 to 0.95. For precision mechanical parts assembly scenarios, the matching threshold is set to 0.75 to 0.90. The higher the pose accuracy requirement and the higher the operation safety risk level, the higher the matching threshold is set.

[0152] A spherical extended verification range is defined with the effective curvature feature points as the center and half the length of the spatial diagonal of the standard surface model axis-aligned bounding box as the radius. The matching density of the effective curvature feature points within the extended verification range is statistically analyzed, and the continuous region with the highest matching density is selected as the potential registration region in the real-time local surface model. This solves the problem of global and local scale differences between real-time local acquisition data and the global 3D solid model, reduces the calculation range of subsequent registration, and improves registration efficiency and accuracy.

[0153] Within the selected potential registration area, using the matched effective curvature feature points as a reference, and based on the dual constraints of curvature feature similarity and spatial distance, edge feature points of the mapped standard surface model are searched. Specifically, within the potential registration area, for each edge feature point of the standard surface model, a local search sub-region is delineated with the edge feature point as the center and a preset spatial search threshold as the radius. This preset spatial search threshold is a critical value of the spatial radius that limits the feature matching range. Based on scene accuracy requirements, feature point distribution density, and the preset registration area scale, candidate points with a curvature feature similarity greater than the preset similarity threshold with the edge feature points of the standard surface model are selected within the local search sub-region. The candidate point with the highest similarity is selected as the mapping point of the edge feature point. All mapped edge points are connected in the edge order of the standard surface model to form a closed real-time operation area that matches the contour of the standard surface model. At the same time, the functional feature points and positioning reference structural feature points bound to the current target node in the registration reference feature set are mapped to the corresponding positions in the real-time operation area within the potential registration area. The optimal rigid transformation matrix is ​​solved based on the spatial coordinate deviation of the feature point pairs, and the translation vector and rotation parameters are locked to achieve accurate spatial alignment between the real-time local surface model and the standard surface model. The preset similarity threshold is a critical value used to determine whether candidate points of a real-time local surface model and edge feature points of a standard surface model have sufficient geometric similarity and can be used as effective mapping edge points. It is set based on the pose accuracy requirements in the implantation operation specifications, the safety level of the operation scenario, and the complexity of the edge structure of the operation object. The value ranges from 0 to 1. The closer the value is to 1, the higher the requirements for the curvature characteristics and local geometric similarity of the edge feature points, and the stricter the control of the edge mapping accuracy. For example, for precision medical implantation scenarios such as human temporal bone tissue, the preset similarity threshold is set to 0.85 to 0.95. For precision mechanical parts assembly scenarios, the preset similarity threshold is set to 0.75 to 0.90. The higher the pose accuracy requirements, the higher the operation safety risk level, and the more complex the edge structure, the higher the preset similarity threshold is set.

[0154] Please see Figure 4 Preferably, the specific steps for targeted visual guidance of target nodes include:

[0155] Within a standard Cartesian coordinate system, the target spatial coordinates of the current target node are used as the geometric center. Combined with a preset operational tolerance range, the effective coverage judgment area for the target node is defined. This effective coverage judgment area is a spherical spatial region centered on the target node's target spatial coordinates and with the operational tolerance range as its radius, providing a quantitative benchmark for coverage judgment. Specifically, the operational tolerance range refers to the maximum allowable deviation of the target node in its spatial position to ensure the feasibility and accuracy of the corresponding process. This is configured by combining the pose accuracy requirements in the implantation operation specifications, the execution requirements of the corresponding process, and the design structural tolerances of the work object. For example, for precision medical implantation scenarios such as human temporal bone tissue, the operational tolerance range is set to 0.3mm to 0.5mm; for precision mechanical parts assembly scenarios, the operational tolerance range is set to 0.02mm to 0.05mm. The higher the pose accuracy requirements and the stricter the process execution requirements, the smaller the operational tolerance range is set.

[0156] The closed real-time operation area matching the current actual operation area is compared with the effective coverage determination area of ​​the target node. If the closed real-time operation area completely contains the effective coverage determination area of ​​the target node, the target node is determined to be covered; otherwise, the target node is determined to be uncovered.

[0157] If the target node is determined not to be covered, the geometric centroid of the actual working area that is closed in real time and matches the contour of the standard surface model is calculated. The three-dimensional spatial position deviation between this deviation and the geometric centroid of the target node's effective coverage determination area is calculated. Based on the three-dimensional spatial position deviation, the required three-dimensional movement direction and linear movement distance of the guidance and assistance device are calculated, and standardized displacement guidance commands are generated. Through the light source emission component of the guidance and assistance device, the adjustment direction, movement amplitude, and attitude adjustment amount of the device are intuitively indicated by the differentiated beam color, directional light spot, flashing frequency, and beam intensity, driving the automatic actuator or guiding the operator to adjust the spatial pose of the guidance and assistance device. After each pose adjustment is completed, real-time optical three-dimensional data acquisition, dynamic registration, and coverage verification are re-executed until it is determined that the actual working area completely covers the target node.

[0158] If the target node is determined to be covered, the targeted positioning guidance operation of the guidance and assistance device is initiated. The three-dimensional coordinates of the functional feature points bound to the target node in the standard spatial rectangular coordinate system and the target spatial coordinates of the target node are converted into the physical projection coordinates corresponding to the actual working space. Combined with the pre-calibrated internal and external parameters and spatial projection mapping relationship of the light source emitting and receiving components of the guidance and assistance device, the precise coordinate mapping between the target node in the digital model and the physical surface of the actual working area is completed. The target node on the surface of the actual working area is located, and a targeted positioning closed light spot matching the working boundary of the target node is projected onto the surface of the actual working area. The core point of the target node is marked with a cross light spot. At the same time, an axial guide line and a working range outline matching the implantation process are projected to intuitively indicate the operating direction and contact range of the target operating equipment, so as to realize the precise visual positioning and operation guidance of the target node.

[0159] This embodiment introduces an implantation point optimization system based on optical positioning and visual modeling, including a model building module, a compliance judgment module, an intelligent planning module, and a visualization guidance module;

[0160] The model building module is used to perform full-view optical 3D scanning of the target area of ​​the object to be implanted using an adapted non-contact optical scanning device, and to collect multimodal optical 3D data of the target area. The collected multimodal optical 3D data is then processed sequentially with point cloud denoising, point cloud registration, outlier removal, mesh generation, and surface reconstruction to generate a 3D solid model of the object's physical structure, providing a unified digital reference carrier for the entire implantation point planning and operation guidance process. Based on preset implantation operation specifications, a positioning reference structure matching the implantation operation is extracted from the 3D solid model. Using the extracted positioning reference structure as the origin, a standard spatial rectangular coordinate system dedicated to the implantation operation is established, clarifying the axial rules and scale accuracy of the coordinate system. Based on preset implantation operation parameter requirements, the 3D spatial coordinates, implantation normal attitude angle, and effective working surface parameters corresponding to the standard implantation point are calculated in the standard spatial rectangular coordinate system, completing the binding and storage of the positioning mark of the standard implantation point and the corresponding pose parameter set in the 3D solid model.

[0161] The compliance determination module is used to construct a quantitative evaluation model for implantation point quality based on preset implantation operation specifications. It clarifies the multi-dimensional evaluation index system corresponding to each implantation operation specification, the quantitative assignment rules for each evaluation index, weight coefficients, and qualification thresholds. The module substitutes the pose parameter set corresponding to the standard implantation point output by the 3D data acquisition and model building module into the quantitative evaluation model to complete the multi-dimensional quantitative quality evaluation of the standard implantation point and calculate its comprehensive quality score. The comprehensive quality score is compared with the preset qualification threshold to determine whether the standard implantation point meets the compliance requirements of the implantation operation. If the standard implantation point meets the compliance requirements, a compliance pass instruction is generated and the subsequent implantation path planning process is triggered. If the standard implantation point does not meet the compliance requirements, a compliance fail instruction is generated, and the search for alternative implantation points, quality quantitative evaluation, and output of the optimal alternative implantation point are executed simultaneously. Corresponding abnormal warning information is generated simultaneously. If there are no compliant alternative implantation points, an implantation operation prohibition warning is generated and the current implantation planning process is terminated.

[0162] The intelligent planning module, upon receiving a compliance pass instruction from the implantation point quality quantification and compliance judgment module, uses the target spatial coordinates corresponding to the standard implantation point located in the 3D data acquisition and model construction module as the endpoint of the implantation path and the initial entry point corresponding to the implantation operation as the starting point. Simultaneously, it retrieves a set of auxiliary operation points matching the operation type of the standard implantation point. Using the standard Cartesian coordinate system corresponding to the 3D solid model as a reference, and combining the process space mapping rules of the auxiliary operation point set, it completes the spatial coordinate matching, adaptive correction, and precise annotation of each auxiliary operation point in the 3D solid model. Based on the process execution order of each auxiliary operation point, it plans the implantation path... The starting point, auxiliary operation points, and implantation path endpoint are sorted to form a node sequence for implantation path planning. A full-domain implantation operation space covering the entire operation range is constructed. The full-domain implantation operation space is subjected to three-dimensional equal-size rasterization discretization processing. Based on the node sequence, the node spatial distribution density value of the full-domain raster unit is calculated, and the operation space is clustered and positively correlated with the corresponding minimum safe avoidance distance is assigned. Combining the obstacle space contour constraints of the three-dimensional solid model and the minimum safe avoidance distance of each raster unit, the segmented planning and global optimization of the implantation path are completed with the node sequence as the necessary nodes. A smooth, continuous, collision-free implantation path is generated and mapped to the standard spatial rectangular coordinate system of the three-dimensional solid model.

[0163] The visualization guidance module is used to iteratively update the current target node in real time according to the execution order of the standard implantation procedure, based on the implantation path and node sequence output by the intelligent planning module. Through the light source emission and reception structure of the adapted guidance auxiliary device, it continuously collects real-time optical 3D data of the current actual working area, performs preprocessing on the real-time optical 3D data, and constructs a real-time local surface model. It extracts the standard surface model of the corresponding area to be registered from the 3D solid model, completes feature matching and dynamic registration between the real-time local surface model and the standard surface model, solves for the optimal rigid transformation matrix, and completes coordinate system unification and deviation calibration between the actual working space and the standard Cartesian coordinate system of the 3D solid model. Based on the registration results, it determines in real time the effective acquisition space range of the current actual working area. If the target node is not fully covered, the system calculates the required three-dimensional movement direction, linear movement distance, and three-axis attitude adjustment parameters for the guidance and assistance device, generates standardized displacement guidance commands, and drives or guides the guidance and assistance device to complete the pose adjustment until the target node is fully covered. If the target node is covered, the system activates the targeted positioning guidance mode of the guidance and assistance device, completes the precise coordinate mapping between the target node in the digital model and the physical surface of the actual work area, projects targeted visual guidance marks onto the surface of the actual work area through the light source emission component of the guidance and assistance device, achieves precise positioning and operation guidance of the target node, and completes the full process guidance according to the node sequence until the target operating equipment reaches the standard implantation point, completing the closed-loop guidance and effect verification of the implantation operation.

[0164] Working principle and its effects:

[0165] This invention, based on optical positioning and 3D visual modeling technology, constructs a closed-loop management system covering the entire process of precision implantation operations, from site planning, quality control, path generation to execution guidance. Its core working principle is to use standardized implantation operation specifications as a unified standard, and replace the traditional operation mode that relies on manual experience with full-process digital modeling and quantitative analysis. It achieves full-link benchmark unification of implantation point positioning, path planning and on-site operation with a unified spatial coordinate system, fundamentally solving the core pain points of insufficient positioning accuracy, lack of control standards and disconnect between planning and execution in traditional precision implantation operations. It provides full-process traceability, controllability and high precision technical support for precision implantation operations.

[0166] In the implantation point planning stage, this invention constructs a high-precision 3D solid model of the work object through optical 3D scanning, extracts the positioning benchmark structure according to preset specifications to establish a unified standard spatial rectangular coordinate system, calculates the full-dimensional pose parameters of the standard implantation point and completes standardized positioning marking, establishing a precise digital benchmark for the entire process, avoiding subjective deviations in the positioning benchmark, and improving the positioning accuracy and consistency of the implantation point. In the implantation point quality control stage, a multi-dimensional quantitative evaluation model is constructed to objectively determine the compliance of the implantation point using standardized rules, replacing traditional manual experience judgment, screening out unqualified points in advance, avoiding implantation risks, and establishing a unified quality control standard. In the implantation path planning stage, standardized auxiliary operation points are matched to compliant implantation points and coordinate calibration is completed. Graded safety avoidance distances are allocated through full-domain spatial rasterization, and implantation path planning is optimized by combining obstacle contours to ensure path adaptability and operational safety. In the operation execution guidance stage, target nodes are updated in real time based on the planned path. Real-time coordinate calibration is completed through real-time optical data acquisition and dynamic model registration, and targeted visual guidance is achieved by combining coverage judgment, breaking down the barriers between planning and execution, reducing human operation deviations, and improving operational accuracy and efficiency.

[0167] In summary, this invention, through its end-to-end digital, standardized, and closed-loop technical design, not only significantly improves the positioning accuracy, execution efficiency, and standardized management level of precision implantation operations, but also flexibly adapts to various high-requirement precision implantation scenarios such as medical implantation, precision mechanical assembly, and microelectromechanical system integration, possessing strong engineering application value and prospects for large-scale promotion.

[0168] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An implantation point optimization method based on optical positioning and visual modeling, characterized in that, include: A full-view optical 3D scan is performed on the target area of ​​the object to be implanted, and a 3D solid model of the object's physical structure is generated. Based on the preset implantation operation specifications, the positioning reference structure matching the implantation operation is extracted from the three-dimensional solid model, a standard spatial rectangular coordinate system for the implantation operation is established, a set of pose parameters corresponding to the standard implantation point is generated, and the standard implantation point of the three-dimensional solid model is located and marked. A quantitative evaluation model for implantation point quality is constructed. Based on the set of pose parameters corresponding to the standard implantation point, the comprehensive quality score of the standard implantation point is calculated to determine whether the standard implantation point meets the compliance requirements. If the standard implantation point meets the compliance requirements, the set of auxiliary operation points that match the operation type of the standard implantation point is retrieved and mapped and marked in the three-dimensional solid model. The target spatial coordinates of each auxiliary operation point are obtained to construct the full domain space of the implantation operation and divide it into grid units. By calculating the node spatial distribution density value of each grid unit, the minimum safe avoidance distance is allocated. Combined with the obstacle space contour of the three-dimensional solid model, the implantation path is generated. Based on the implantation path, the target node is updated in real time, real-time optical 3D data of the actual working area is collected, dynamic registration is performed with the 3D entity model, it is determined whether the effective acquisition space range of the current actual working area completely covers the current target node, and targeted visualization guidance is performed on the target node.

2. The implantation point optimization method based on optical positioning and visual modeling as described in claim 1, characterized in that, The steps for establishing a standard spatial rectangular coordinate system for the implantation operation include: Multimodal optical 3D data of the target area to be implanted is collected, and spatial dimension is uniformly processed. Global registration is performed according to the spatial position correspondence, and the data is fused into a unified data set under a unified spatial coordinate system. Anomalies are identified and removed from the unified dataset, and spatial grids are generated to construct an initial grid structure that matches the surface morphology of the work object. The initial mesh structure is surface-fitted and smoothly reconstructed to generate a three-dimensional solid model of the object's solid structure. The pre-defined implantation operation specifications include structural adaptation requirements, pose accuracy requirements, safety obstacle avoidance requirements, and positioning reference structure selection requirements for the implantation operation. Based on the preset implantation operation specifications, the global positioning reference structure is extracted from the three-dimensional solid model, and structural regions that meet the requirements for the selection of positioning reference structures in the implantation operation specifications are selected. Geometric feature analysis is performed on the selected structural regions to extract the geometric center, contour boundary and normal direction of the structural regions, establish a standard spatial rectangular coordinate system for implantation, and axial calibration of the standard spatial rectangular coordinate system is performed.

3. The implantation point optimization method based on optical positioning and visual modeling as described in claim 2, characterized in that, The steps for locating and marking the standard implantation points of the three-dimensional solid model include: Based on the structural adaptation requirements and pose accuracy requirements in the preset implantation operation specifications, the target area to be implanted in the three-dimensional solid model is partitioned and analyzed in a standard spatial rectangular coordinate system to identify the target point area. Collect spatial coordinate information of each point within the target location area, filter out the core point set of the target location area through coordinate clustering analysis, and calculate the average coordinate of the core point set as the target spatial coordinates corresponding to the target location area; Based on the surface morphology of the target location area, the normal direction of the core point set of the target location area is calculated, which is used to solve for the implantation normal attitude angle information corresponding to the target location area; Measure the continuous surface area of ​​the target location area, select the continuous surface within the target location area whose flatness meets the implantation operation specifications as the effective working surface; and determine the boundary range and geometric parameters of the effective working surface to form the effective working surface parameter information corresponding to the target location area. The target spatial coordinates, implantation normal attitude angle information and effective working surface parameter information corresponding to the target point area are associated and integrated to form a set of pose parameters corresponding to the standard implantation point. In the 3D solid model, based on the set of pose parameters corresponding to the standard implantation point, the spatial point corresponding to the standard implantation point is located, and a unique positioning mark is drawn at that spatial point.

4. The implantation point optimization method based on optical positioning and visual modeling as described in claim 1, characterized in that, The steps for determining whether the implantation site meets the compliance requirements include: Based on the pre-defined implantation operation specifications, a multi-dimensional evaluation index system for implantation site quality assessment is determined, and a quantitative evaluation model for implantation site quality is constructed; corresponding quantitative assignment rules and weight coefficients are set for each evaluation index in the multi-dimensional evaluation index system. According to the quantitative assignment rules corresponding to each evaluation indicator, the collected pose parameter set is quantitatively converted to obtain the standardized quantitative score corresponding to each evaluation indicator; the standardized quantitative score corresponding to each evaluation indicator is weighted and calculated with the corresponding weight coefficient, and the comprehensive quality score corresponding to the standard implantation point is obtained by weighted summation. If the overall quality score is greater than or equal to the passing threshold, the standard implantation point is deemed to meet compliance requirements, a compliance pass instruction is generated, and the subsequent implantation path planning process is triggered. If the overall quality score is less than the passing threshold, the standard implantation point is deemed to not meet compliance requirements, a compliance fail instruction is generated, and the subsequent alternative implantation point planning and anomaly warning process is triggered.

5. The implantation point optimization method based on optical positioning and visual modeling as described in claim 1, characterized in that, The steps for generating the implantation path include: The target spatial coordinates corresponding to the standard implantation point are taken as the end point of the implantation path, and the initial entry point corresponding to the implantation operation is taken as the starting point of the implantation path; and the set of auxiliary operation points that match the operation type to which the standard implantation point belongs is retrieved. Using the standard Cartesian coordinate system corresponding to the 3D solid model as a reference, and combining the process space mapping rules of the auxiliary operation point set, each auxiliary operation point in the auxiliary operation point set is mapped to the 3D solid model to obtain the target space coordinates of each auxiliary operation point. Based on the process execution order of each auxiliary operation point, the starting point of the implantation path, each auxiliary operation point, and the ending point of the implantation path are sorted to form a node sequence for implantation path planning. Based on the geometric structure data of the 3D solid model, the 3D isosurface boundary of the work object is extracted as the obstacle space contour; The two adjacent nodes in the node sequence are divided into an independent path planning unit. The obstacle space contour of the 3D solid model is used as the spatial constraint. The path planning is performed in combination with the minimum safe avoidance distance of each grid unit to generate the sub-path corresponding to each path planning unit. The sub-paths of all path planning units are connected in the order of the node sequence to obtain the implanted path.

6. The implantation point optimization method based on optical positioning and visual modeling as described in claim 5, characterized in that, The steps of constructing the entire space for the implantation operation and dividing it into grid cells, and allocating the minimum safe avoidance distance by calculating the node spatial distribution density value of each grid cell, include: A cuboid space covering the starting point of the implantation path, each auxiliary operation point, the ending point of the implantation path, and the preset operation range is constructed as the implantation operation global space; the implantation operation global space is discretized into three-dimensional equal-size rasterization, and the implantation operation global space is decomposed into several raster units; Based on the target spatial coordinates of each auxiliary operation point and the standard implantation point, combined with the side length of a single grid cell and the average spacing between adjacent nodes in the node sequence, the calculation kernel width is determined. The geometric center of each grid cell in the entire implantation operation space is used as the calculation base point. The Gaussian kernel function is used to traverse and calculate the density contribution value of all nodes in the node sequence to the grid cell, and the node spatial distribution density value of each grid cell is calculated. Based on the preset quantile threshold of the global grid cell density value, all grid cells are divided into operation space clusters, and the minimum safe avoidance distance of each operation space grid cell is assigned based on the positive correlation of the node spatial distribution density.

7. The implantation point optimization method based on optical positioning and visual modeling as described in claim 5, characterized in that, The steps for obtaining the target space coordinates of each auxiliary operation point include: Read the preset relative position parameters between each auxiliary operation point and the standard implantation point from the process space mapping rules built into the auxiliary operation point set, including relative coordinate offset, relative angle constraint, and relative distance range; Based on the target spatial coordinates and implantation normal attitude angle of the standard implantation point, the initial three-dimensional spatial coordinates of each auxiliary operation point in the current three-dimensional solid model are calculated in the standard spatial rectangular coordinate system according to the relative position parameters of each auxiliary operation point. Obtain the structural features of the positioning reference structure associated with each auxiliary operation point, and use the initial three-dimensional spatial coordinates of each auxiliary operation point as the center of the sphere, combined with its structural features, to set the basic search radius and delineate the spherical search area. Within the spherical search area, the actual structural features of each auxiliary operation point are extracted point by point, and the similarity between the actual structural features and the structural features of the corresponding associated positioning reference structure is calculated. This is used to filter candidate points within the spherical search area, and the three-dimensional spatial coordinates of the candidate points are used as the target spatial coordinates of the auxiliary operation point.

8. The implantation point optimization method based on optical positioning and visual modeling as described in claim 1, characterized in that, The step of acquiring real-time optical 3D data of the actual working area and performing dynamic registration with the 3D solid model includes: Preprocess the optical 3D data of the current actual working area acquired in real time, extract real-time surface features, and construct a real-time local surface model; Extract the standard surface model of the 3D solid model, perform point-by-point geometric analysis, extract the edge feature points and curvature extreme points of the standard surface model to form a registration reference feature set; and perform coordinate normalization processing on the standard surface model to construct its axis-aligned bounding box and mark the geometric center. Calculate the curvature value of each sampling point of the real-time local surface model, extract real-time curvature feature points, perform curvature similarity matching between the real-time curvature feature points and the curvature extreme points in the registration reference feature set, and screen effective curvature feature points. A spherical extended verification range is defined with the effective curvature feature points as the center and half the length of the spatial diagonal of the standard surface model axis-aligned bounding box as the radius. The matching density of the effective curvature feature points within the extended verification range is statistically analyzed, and potential registration regions are selected. Within the selected potential registration area, using the matched effective curvature feature points as a reference, the edge feature points of the mapped standard surface model are searched, and the mapped edge points are connected in the edge order of the standard surface model to form a closed real-time operation area that matches the contour of the standard surface model.

9. The implantation point optimization method based on optical positioning and visual modeling as described in claim 8, characterized in that, The steps for targeted visualization guidance of the target node include: Within the standard Cartesian coordinate system, the effective coverage area of ​​the target node is defined by taking the target spatial coordinates of the current target node as the geometric center and combining the preset operation tolerance range. The closed real-time operation area matching the current actual operation area is compared with the effective coverage determination area of ​​the target node. If the closed real-time operation area completely contains the effective coverage determination area of ​​the target node, it is determined that the target node has been covered; otherwise, it is determined that the target node has not been covered. If the target node is determined to be uncovered, the geometric centroid of the actual working area closed real-time operation area that matches the contour of the standard surface model is calculated, and the three-dimensional spatial position deviation between the geometric centroid of the target node effectively covered judgment area is calculated, and a standardized displacement guidance command is generated. If the target node is determined to be covered, the target node on the surface of the actual work area is located, and the target node is visualized and the work is guided.

10. An implantation point optimization system based on optical positioning and visual modeling, used to implement the implantation point optimization method based on optical positioning and visual modeling as described in any one of claims 1 to 9, characterized in that, It includes a model building module, a compliance assessment module, an intelligent planning module, and a visualization guidance module; The model building module is used to generate a three-dimensional solid model of the physical structure of the work object, extract the positioning reference structure matching the implantation work, establish a standard spatial rectangular coordinate system for the implantation work, generate a set of pose parameters corresponding to the standard implantation point, and locate and mark the standard implantation point of the three-dimensional solid model. The compliance determination module is used to construct a quantitative evaluation model for implantation site quality to determine whether standard implantation sites meet compliance requirements. The intelligent planning module is used to retrieve a set of auxiliary operation points that match the operation type of the standard implantation point, map and label them into the three-dimensional solid model, and generate an implantation path by combining the obstacle space contour of the three-dimensional solid model. The visualization guidance module is used to collect real-time optical 3D data of the actual work area, perform dynamic registration with the 3D entity model, determine whether the effective collection space range of the current actual work area completely covers the current target node, and provide targeted visualization guidance for the target node.

Images