Registration methods, devices, storage media, and electronic equipment for implantation scanning rods

Abstract

This application provides a registration method, apparatus, storage medium, and electronic device for implantation scanning rods, relating to the field of medical devices. Specifically, a host computer acquires depth images of multiple scanning rods and standard models of their respective registration sites. In situations where limited oral space, soft tissue obstruction, and noise interference cause point cloud adhesion, the scanning point cloud is first precisely separated into point clouds corresponding to individual scanning rods based on the recognition results in the depth images. Then, the feature surface point cloud used for registration is identified from this segment, and finally, it is registered with the target registration model that best matches its shape. This avoids mismatches or incorrect registrations caused by point cloud confusion, allowing the spatial position of each scanning rod to be calculated independently and stably.

Description

Technical Field

[0001] This application relates to the field of medical devices, and more specifically, to a registration method, apparatus, storage medium, and electronic device for implantation scanning rods. Background Technology

[0002] For edentulous patients who have lost all their upper or lower jaw teeth, multiple implant scanning rods (hereinafter referred to as scanning rods) need to be inserted at the implant site during implant surgery. The depth, angle and position information of the implant in the jawbone are obtained by a 3D digital impression instrument to ensure that the final tooth is accurately placed.

[0003] like Figure 1 As shown, to accurately acquire the three-dimensional spatial position of implants, a scanning rod is commonly used in clinical practice for positioning. The scanning rod is a specialized instrument designed with a geometric structure easily scanned and identified by oral scanning equipment (hereinafter referred to as an intraoral scanner). It includes a connecting end 11, a registration area 12, and a classification area 13. The connecting end 11 connects to the top of the implant already placed in the patient's mouth. The registration area 12 has a relatively complex geometry designed to facilitate registration, while the classification area 13 is designed to represent the manufacturer's specific geometry. When the dentist scans the inside of the oral cavity with the intraoral scanner, the scanning rod is captured as a reference marker, thereby deducing the precise position and angle of each implant within the oral cavity.

[0004] In current clinical practice for edentulous patients, it is often necessary to install and scan multiple scanning rods at one time to achieve synchronous positioning of multiple implants. This can shorten the time patients need to open their mouths and improve their treatment experience, and also avoid the cumulative error in the relative spatial relationship between multiple implants due to multiple scans.

[0005] However, in practice, it was found that when scanning multiple scanning rods simultaneously, the raw point cloud data acquired by the oral scanning device itself has problems such as high noise, blurred boundaries, and severe local occlusion due to factors such as limited scanning space, soft tissue obstruction, equipment noise, and light reflection. In addition, the fact that multiple scanning rods are close to each other, have similar structures, and have similar surface reflective properties causes the point clouds of different scanning rods to often stick together, the contour boundaries cannot be clearly distinguished, and even the point cloud of one scanning rod is misjudged as part of another scanning rod.

[0006] The aforementioned factors can sometimes cause the registration algorithm to fail to accurately distinguish point cloud clusters of different scanning rods, leading to pose mismatch, mismatch, reduced registration accuracy, or even registration failure, thus failing to meet the precise positioning requirements of multiple scanning rods simultaneously. Summary of the Invention

[0007] In order to overcome at least one deficiency in the prior art, this application provides a registration method, apparatus, storage medium and electronic device for planting scanning rods, which can avoid mismatch or misregistration caused by point cloud confusion, and enable the spatial position of each scanning rod to be calculated independently and stably.

[0008] In a first aspect, this application provides a registration method for an implantation scanning rod, the method comprising: Acquire depth images of multiple scanning rods and registration models of the multiple scanning rods, wherein the multiple scanning rods are installed in the patient's oral cavity, and the registration model of each scanning rod is a standard model of the registration site; Based on the recognition results of the plurality of scanning rods in the depth image, a scanning rod point cloud corresponding to each scanning rod is determined from the scanning point cloud based on the depth image; Feature surface point clouds are identified from the scanning rod point cloud, wherein the feature surface point cloud represents the point cloud generated by the plane used for registration of the corresponding scanning rod; The feature surface point cloud is registered with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud. The target registration model is the registration model that is most similar in shape to the feature surface point cloud.

[0009] Secondly, this application provides a registration device for an implantation scanning rod, the device comprising: The data preparation module is used to acquire depth images of multiple scanning rods and registration models of the multiple scanning rods, wherein the multiple scanning rods are installed in the patient's oral cavity, and the registration model of each scanning rod is a standard model of the registration site; The point cloud separation module is used to determine the scanning rod point cloud corresponding to each scanning rod from the scanning point cloud based on the depth image according to the recognition results of the multiple scanning rods in the depth image; The model registration module is used to identify feature surface point clouds from the scanning rod point cloud, wherein the feature surface point cloud represents the point cloud generated by the plane used for registration of the corresponding scanning rod; The model registration module is further used to register the feature surface point cloud with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud, wherein the target registration model is the registration model most similar in shape to the feature surface point cloud.

[0010] Thirdly, this application provides a storage medium storing a computer program that, when executed by a processor, implements the registration method for the implantation scanning rod.

[0011] Fourthly, this application provides an electronic device, which includes a processor and a memory, the memory storing a computer program, which, when executed by the processor, implements the registration method for the implantation scanning rod.

[0012] Compared with the prior art, this application has the following beneficial effects: The registration method, apparatus, storage medium, and electronic device for implantation scanning rods provided in this application acquire depth images of multiple scanning rods and standard models of their respective registration sites via a host computer. In situations where limited oral space, soft tissue obstruction, and noise interference cause point cloud adhesion, the scanning point cloud is first precisely separated into point clouds corresponding to individual scanning rods based on the recognition results in the depth images. Then, the feature surface point cloud used for registration is identified from this data, and finally, it is registered with the target registration model that best matches its shape. This avoids mismatches or incorrect registrations caused by point cloud confusion, allowing the spatial position of each scanning rod to be calculated independently and stably. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic diagram of the structure of the scanning rod provided in an embodiment of this application; Figure 2 A schematic flowchart illustrating the registration method for the implantation scanning rod provided in an embodiment of this application; Figure 3 A detailed schematic diagram of step S2 provided in the embodiments of this application; Figure 4 A detailed schematic diagram of step S3 provided in the embodiments of this application; Figure 5 A schematic diagram of the registration device for the implantation scanning rod provided in an embodiment of this application; Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of this application (hereinafter referred to as "the embodiments") clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0016] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0017] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0018] In the description of this application, it should be noted that the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0019] Based on the above statement, as introduced in the background technology, due to the limited scanning space, soft tissue occlusion, equipment noise and light reflection, the raw point cloud data acquired by the intraoral scanning device itself has problems such as high noise, blurred boundaries and severe local occlusion. In addition, multiple scanning rods are close to each other, have similar structures and similar surface material reflective properties, which causes the point clouds of different scanning rods to often stick together and the contour boundaries cannot be clearly distinguished. Consequently, the registration algorithm cannot accurately distinguish the point cloud clusters of different scanning rods, ultimately affecting the registration accuracy.

[0020] It should be noted that the defects in the solutions in the prior art are the result of practice and careful research. Therefore, the discovery process of the above problems and the solutions proposed by the embodiments of this application in the following text should be regarded as contributions to this application in the process of invention and creation, and should not be understood as technical content known to those skilled in the art.

[0021] Based on the discovery of the above-mentioned technical problems, this embodiment provides a registration method for implantation scanning rods. For example... Figure 2 As shown, the method includes: S1, acquire depth images of multiple scanning rods and registration models of multiple scanning rods.

[0022] Multiple scanning rods are installed inside the patient's oral cavity, and the registration model of each scanning rod is the standard model of the registration site 12.

[0023] S2, based on the recognition results of multiple scanning rods in the depth image, determine the scanning rod point cloud corresponding to each scanning rod from the scanning point cloud based on the depth image.

[0024] S3, identify feature surface point clouds from the scanning rod point cloud.

[0025] Among them, the feature surface point cloud represents the point cloud generated by the plane used for registration by the corresponding scanning rod.

[0026] S4. Register the feature surface point cloud with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud.

[0027] Among them, the target registration model is the registration model that is most similar to the shape of the feature surface point cloud.

[0028] This embodiment can be understood as follows: the host computer acquires the depth images of multiple scanning rods and the standard models of their respective registration parts 12. In cases where the oral cavity space is limited, soft tissue occlusion and noise interference cause the point clouds to stick together, the scanning point cloud is first accurately separated into the point cloud corresponding to a single scanning rod based on the recognition results in the depth image. Then, the feature surface point cloud used for registration is identified from it. Finally, it is registered with the target registration model that best matches the shape.

[0029] In this way, mismatches or incorrect matching caused by point cloud confusion are avoided, and the spatial position of each scanning rod can be calculated independently and stably.

[0030] It should be noted that the host computer implementing this method, as an electronic device, can be, but is not limited to, a mobile terminal, tablet computer, laptop computer, desktop computer, server, customized embedded device, etc. The server can be a single server or a group of servers. The server group can be centralized or distributed (e.g., the servers can be a distributed system). In some embodiments, the server can be local or remote relative to the user terminal. In some embodiments, the server can be implemented on a cloud platform; by way of example only, the cloud platform can include private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, inter-cloud, multi-cloud, etc., or any combination thereof. In some embodiments, the server can be implemented on an electronic device having one or more components.

[0031] In practical applications, the host computer communicates with the scanning device, receives depth images from the scanning device, and reads the registration models of multiple scanning rods to complete subsequent registration calculations.

[0032] To make the solution provided in this embodiment clearer, the following is combined with... Figure 2 Each step of the method is described in detail. However, it should be understood that the operations in the flowchart may not be implemented in sequence, and steps without logical contextual relationships may be reversed in order or performed simultaneously. Furthermore, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowchart, or remove one or more operations from the flowchart. See also... Figure 2 The following is a detailed explanation of step S1 in the diagram: S1, acquire depth images of multiple scanning rods and registration models of multiple scanning rods.

[0033] Multiple scanning rods are installed inside the patient's oral cavity, and the registration model of each scanning rod is the standard model of the registration site 12.

[0034] In practical applications, this depth image can be acquired in real time by an intraoral scanning device placed inside the oral cavity. It can reflect the color and spatial information inside the oral cavity. Therefore, its pixel value includes not only the color information inside the oral cavity, but also the distance from each point in the oral cavity to the intraoral scanning device. The registration model is not the complete shape of the entire scanning rod, but only includes the standard three-dimensional geometric model of the registration part 12 used for spatial positioning. The registration part 12 is usually designed as a structure with complex geometry and prominent features.

[0035] For example, see [link to previous article] Figure 1The registration part 12 is a protrusion or a specific curved surface on the head of the scanning rod, so as to provide a stable and reproducible matching basis in subsequent registration.

[0036] Based on the above explanation of the depth image and registration model in step S1, the following will elaborate on step S2: S2, based on the recognition results of multiple scanning rods in the depth image, determine the scanning rod point cloud corresponding to each scanning rod from the scanning point cloud based on the depth image.

[0037] It should also be noted that since each pixel in the depth image includes its distance from the intraoral scanning device, point cloud reconstruction can be performed by combining the camera intrinsic parameters of the intraoral scanning device, thus obtaining the scanned point cloud of the oral cavity. However, when the host computer processes edentulous jaw scan data, it faces the raw, unprocessed scanned point cloud, which is reconstructed from the entire oral cavity scene and includes mixed information from the gingiva, mucosa, salivary reflective areas, and multiple scanning rods. Since the scanning rods are small, closely spaced, and their surface materials are prone to specular reflection, their point clouds occupy only a small local area within the overall point cloud.

[0038] In this situation, if the host computer directly performs scanning rod identification on the entire scanned point cloud, it cannot eliminate interference caused by soft tissue deformation and occlusion, resulting in a lack of clear directionality in the identification process and difficulty in stably locking the effective point set of each scanning rod. To solve the above problems, such as Figure 3 As shown, this embodiment provides the following optional implementation methods for step S2: S2-1, Based on the recognition results of multiple scanning rods in the depth image, determine the depth information corresponding to the multiple scanning rods from the depth image.

[0039] As an optional implementation, the host computer can construct mask images of multiple scanning rods based on the recognition results of multiple scanning rods in the depth image; and use the mask images to crop the depth image to obtain the depth information corresponding to the multiple scanning rods.

[0040] In practical applications, an intraoral scanning device inserted into the oral cavity can be activated to perform real-time scanning of the patient's oral cavity, which has multiple scanning poles installed, to acquire a depth image containing the scanning poles and surrounding soft and hard tissues, denoted as org_img. Subsequently, a deep learning-based image recognition model, such as UNet or Mask R-CNN, is called to perform pixel-level localization and contour extraction on the multiple scanning poles in org_img, generating a mask image that only covers the region of each scanning pole, denoted as mask_img. Then, a pixel-by-pixel logical AND operation is performed on this mask image and the synchronously acquired depth map, so that only the valid depth values ​​of the region identified by mask_img are retained in the depth map, while the rest are set to invalid or zero values, thereby obtaining depth information specific to multiple scanning poles.

[0041] S2-2, cluster the reconstructed point cloud based on depth information to obtain the scan rod point cloud corresponding to each scan rod.

[0042] In practical applications, the host computer can process the point cloud reconstructed based on depth information using a region growing segmentation method. This method, based on the geometric continuity and normal similarity between adjacent points in the point cloud, gradually expands the connected region starting from the initial seed point, thereby dividing the originally mixed point cloud into several spatially separated and structurally self-consistent point cloud clusters. Each separated independent point cloud cluster corresponds to a complete 3D point set of a scanning rod, referred to in this embodiment as the scanning rod point cloud. Therefore, based on this clustering method, the host computer can sequentially traverse each scanning rod point cloud.

[0043] It should also be noted that, in addition to the region growth segmentation method, the host computer can also use other clustering methods that conform to point cloud processing conventions and are suitable for oral clinical scenarios, including but not limited to, density-based spatial clustering of applications with noise (DBSCAN) based on Euclidean distance, or the method of using octree spatial partitioning combined with connected component analysis.

[0044] This can be understood as follows: the host computer first uses the recognition results of multiple scanning rods in the depth image to accurately delineate the depth information region corresponding to each scanning rod in the depth image, and then reconstructs the point cloud based only on this region, so that the subsequent point cloud processing is only for the scanning rod body, thereby excluding the point clouds corresponding to soft tissue deformation, occlusion defects and other interfering structures in the oral cavity, so that the clustering objects are only the point clouds corresponding to each scanning rod.

[0045] Based on the above introduction to the point cloud of the scanning rod, we will continue with... Figure 2 Step S3 will be explained in detail below: S3, identify feature surface point clouds from the scanning rod point cloud.

[0046] Among them, the feature surface point cloud represents the point cloud generated by the plane used for registration by the corresponding scanning rod.

[0047] It should also be noted that the oral cavity of edentulous patients is small, and after multiple scanning poles are installed, soft tissue obstruction, scanning light reflection, and equipment noise often cause the point cloud data to stick together and have blurred boundaries, making it difficult to clearly distinguish the planes to which each belongs. In this process, if the host computer directly divides the mixed point cloud according to a preset fixed plane, it is easy to incorrectly merge the geometric surfaces of different scanning poles, or to split the registration area that should be uniform on the same scanning pole into different subsets, making it impossible to accurately identify the feature surfaces actually used for registration. To solve the above problems, such as... Figure 4 As shown, this embodiment provides the following optional implementation methods for step S3: S3-1, classify the scanning rod point cloud according to the plane to which it belongs, and obtain multiple point cloud subsets.

[0048] S3-2, obtain the disorder level of each point cloud subset.

[0049] S3-3, determine the feature surface point cloud based on the disorder level of each point cloud subset.

[0050] Among them, the feature surface point cloud is the subset of point clouds with the highest degree of disorder.

[0051] In this embodiment, the host computer does not use a fixed plane that is manually preset or empirically defined. Instead, it divides the point cloud into multiple subsets based on the orientation of the normal vectors of each point in the scanning rod point cloud, and then calculates the disorder level of each subset. Since the registration part 12 of the scanning rod is designed to have a complex geometric shape and rich surface orientations, the normal vector distribution dispersion of its corresponding point cloud subset is higher than that of other parts. Therefore, the host computer can select the point cloud subset with the highest disorder level as the feature surface point cloud, so that the part represented by the feature surface point cloud is the same part as the actual registration part of the scanning rod.

[0052] for Figure 4 In step S3-1, it should also be noted that the point cloud data acquired by edentulous jaw scanning does not carry information about the surface to which it belongs. If the host computer does not establish a spatial reference graphic that matches the actual geometry of the scanning rod, it cannot determine which surface of the scanning rod a certain point belongs to. In this case, points in the point cloud that embody the complex geometric features of the registration area 12 are instead categorized into different subsets due to their dispersed orientation, making it impossible to accurately identify the feature surface used for registration. To solve the above problem, this embodiment provides the following optional implementation method for step S3-1: S3-1-1, Construct the directed bounding box of the point cloud of the package scanning rod.

[0053] An oriented bounding box (OBB) includes multiple reference faces.

[0054] In practical applications, when processing the point cloud of the scanning rod, the host computer does not directly use a fixed-oriented cube to wrap it, but first calculates the geometric centroid of the point cloud in three-dimensional space. Then, based on the distribution deviation of all points relative to the centroid, a normalized covariance matrix is ​​constructed. Then, eigenvalue decomposition is performed on the matrix to obtain the eigenvector matrix. Its column vectors reflect the three main directions in which the point cloud extends most significantly, forming a local coordinate system adapted to the posture of the scanning rod.

[0055] when This means that the coordinate system is left-handed, and the host computer needs to invert the eigenvectors corresponding to the smallest eigenvalues ​​to make the corrected matrix... satisfy To form a standard right-handed orthogonal coordinate system, and with As a rotation matrix; for each global coordinate point Transform to this local coordinate system using the following expression:

[0056] In the local coordinate system, the minimum value of the projection range is obtained along the X, Y, and Z axes respectively. With the maximum value This aligns the local axis with the bounding box, and thus the center of the bounding box is obtained. The expression is:

[0057] Then, the host computer obtains the center of the directed bounding box in global coordinates according to the following expression. And all parameters of the directed bounding box, expressed as:

[0058] In this way, the constructed oriented bounding box can truly fit the natural shape and placement posture of the scanning rod point cloud.

[0059] Based on the above description of step S3-1-1 and the bounding box therein, step S3-1 further includes: S3-1-2, classify the scanning rod point cloud according to its orientation to obtain multiple point cloud subsets.

[0060] In each point cloud subset, each point faces the same reference plane.

[0061] When classifying the scanning rod point cloud by orientation, as an optional implementation, for each scanning point in the scanning rod point cloud, the host computer can obtain the dot product of the normal vector between the scanning point and each reference surface; based on the dot product of the normal vector between the scanning point and each reference surface, the target reference surface in which the scanning point is oriented is determined, wherein the dot product of the normal vector between the target reference surface and the scanning point is the largest; and the scanning point is classified into the point cloud subset corresponding to the target reference surface.

[0062] In practical applications, after constructing the directed bounding box, the host computer can first determine its rotation matrix. Six reference normal vectors , , , , , Transformed into six facet normal vectors that fit the actual orientation of the current scan rod point cloud. And for each scan point in the scan rod point cloud; Calculate its own normal vector The dot product operation is performed on each of the six surface normal vectors. Since the dot product value reflects the cosine relationship between the angles between the two vectors, the larger the value, the closer the normal vector of the point is to the normal vector of the corresponding reference surface. Therefore, the host computer assigns each scan point to the point cloud subset corresponding to the reference surface with the largest dot product of its normal vector.

[0063] This can be understood as follows: the host computer first constructs a oriented bounding box that fits the main direction of the scanning rod point cloud, so that its six reference surfaces can truly reflect the orientation of the scanning rod in the front, back, left, right, up, and down directions when it is actually placed in the mouth; then, it classifies the points according to the alignment relationship between the normal of each point and these reference surfaces, so that each subset of the point cloud contains scanning points whose normal vectors are oriented towards the same reference surface.

[0064] for Figure 3 In step S3-2, when obtaining the disorder level of each point cloud subset, this embodiment can quantify the uncertainty of the angle distribution based on the Shannon entropy definition in information theory. That is, the higher the entropy value, the more undulating the surface geometry, which means that the surface is a feature surface designed for registration.

[0065] Based on this concept, for each point cloud subset, the host computer can obtain the polar angle and azimuth angle of each scan point; according to the preset polar angle spacing and azimuth angle spacing, the polar angle and azimuth angle of each scan point in the point cloud subset are statistically analyzed to obtain the polar angle statistical results and azimuth angle statistical results of the point cloud subset; based on the polar angle statistical results and azimuth angle statistical results of the point cloud subset, the Shannon entropy of the polar angle and azimuth angle in the point cloud subset is calculated and used as the disorder level of the point cloud subset.

[0066] In practical applications, the host computer can process each subset of the point cloud. For each scan point, first extract its unit normal vector. Then, based on the spherical coordinate transformation relationship, calculate the polar angle corresponding to the normal vector. and azimuth The expression is: The range of values ​​is

[0067] The range of values ​​is

[0068] Based on this, the host computer will select all points. Coordinates are statistically analyzed and entered into a two-dimensional histogram array according to a preset level of precision. The polar direction is divided into The polar distance, the azimuth direction is divided into The azimuth spacing is obtained at a scale of The counting matrix.

[0069] This can be understood as the host computer first converting the surface orientation of all scanned points in each subset of the point cloud into two angles: the tilt angle when viewed from directly above (polar angle) and the azimuth angle around the center. These angles are then grouped and statistically analyzed using a fine grid. For a scale of... The counting matrix needs to be divided into 10 levels of tilt and 20 levels of rotation direction to form a 200-cell counting matrix.

[0070] Based on this counting matrix, calculate the number of points in each cell. Total number of points in this subset The interval probability is denoted as The expression is:

[0071] In the formula, Indicates the first One tilt gear, Indicates the first One direction gear.

[0072] The host computer further substitutes the above interval probabilities into the Shannon entropy formula to obtain the degree of disorder. The expression is:

[0073] The degree of disorder directly reflects the discreteness of the normal distribution of the surface represented by the subset, thus objectively reflecting its geometric complexity.

[0074] It should be understood that the registration part 12 of the scanning rod used in clinical practice has a geometric structure with high curvature variation and multi-directional normal distribution. The unit normal vector of the point cloud in this part has a more dispersed polar angle and azimuth angle distribution in the spherical coordinate system, and the calculated Shannon entropy is higher than the point cloud subset corresponding to other parts of the scanning rod.

[0075] Therefore, after the host computer completes the Shannon entropy calculation for each subset of point clouds, it compares the entropy values ​​corresponding to each subset, selects the subset of point clouds with the largest value, and identifies it as the feature surface point cloud.

[0076] Based on the above embodiments regarding the feature surface point cloud, the following will continue to discuss... Figure 2 The S4 step is explained in detail below: S4. Register the feature surface point cloud with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud.

[0077] In this embodiment, the host computer can perform global registration between the identified feature surface point cloud and the pre-stored standard registration model. Currently, a large number of related registration algorithms are widely used in clinical practice, and their essence is to find the optimal alignment relationship between their spatial positions.

[0078] Specifically, the host computer first extracts corresponding geometric feature points from the feature surface point cloud and the standard registration model, such as edge turning points, curvature extrema points, or high-density sampling points. Then, it calculates a rigid body transformation matrix through an iterative optimization algorithm, which can make the two sets of feature points coincide as much as possible in space. Based on this matrix, the true spatial position of the scanning rod in the patient's oral cavity, i.e., its position and pose, can be uniquely determined.

[0079] It should also be noted that, in clinical practice, the scanning rods of different manufacturers are designed in the classification part 13 to represent the unique geometric structure of their respective manufacturers, and to satisfy a specific positional relationship with the registration part 12. Based on this, this embodiment also identifies which target manufacturer the scanning rod corresponding to the scanning rod point cloud comes from.

[0080] Therefore, as an optional implementation, the multiple scanning rods in this embodiment can come from multiple manufacturers, and each scanning rod includes a classification part 13 representing its manufacturer. Based on this, the host computer acquires classification models from multiple manufacturers, where each manufacturer's classification model is a standard model corresponding to the classification part 13; according to the position of the feature surface point cloud, a classification point cloud is determined from the scanning rod point cloud, wherein the registration part 12 represented by the feature surface point cloud and the classification part 13 represented by the classification point cloud satisfy a preset positional relationship; the classification point cloud is compared with the classification models of the multiple scanning rods to determine the target manufacturer of the scanning rod corresponding to the scanning rod point cloud.

[0081] This can be understood as follows: after the host computer completes the feature surface point cloud recognition, it maps the geometric relationship between the registration part 12 and the classification part 13 in the standard model, such as the spatial offset of the classification part 13 relative to the registration part 12, onto the spatial coordinate system of the current scanning rod point cloud, thereby accurately calculating the spatial range of the classification part 13; and based on this, extracting point cloud data containing only this region from the scanning rod point cloud.

[0082] Based on this, the host computer calls the established category parameter library and performs similarity matching with the classification model of each manufacturer to determine the target manufacturer with the highest matching degree.

[0083] Based on the same inventive concept as the registration method for the implantation scanning rod provided in this embodiment, this embodiment also provides a registration device for the implantation scanning rod. This device includes at least one software functional module that can be stored in a memory or embedded in an electronic device. The processor in the electronic device executes the executable module stored in the memory. For example, the software functional module and computer program included in this device. Please refer to... Figure 5 Functionally, the device may include: The data preparation module 21 is used to acquire depth images of multiple scanning rods and registration models of multiple scanning rods. The multiple scanning rods are installed in the patient's oral cavity, and the registration model of each scanning rod is a standard model of the registration site 12. The point cloud separation module 22 is used to determine the scanning rod point cloud corresponding to each scanning rod from the scanning point cloud based on the depth image according to the recognition results of multiple scanning rods in the depth image; The model registration module 23 is used to identify feature surface point clouds from the scanning rod point cloud, wherein the feature surface point cloud represents the point cloud generated by the plane used for registration of the corresponding scanning rod; The model registration module 23 is also used to register the feature surface point cloud with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud. The target registration model is the registration model that is most similar to the shape of the feature surface point cloud.

[0084] Optionally, the point cloud separation module 22 determines the scanning rod point cloud corresponding to each scanning rod from the depth image-based scanning point cloud based on the recognition results of multiple scanning rods in the depth image, including: Based on the recognition results of multiple scanning rods in the depth image, the depth information corresponding to the multiple scanning rods is determined from the depth image; The reconstructed point cloud based on depth information is clustered to obtain the scan rod point cloud corresponding to each scan rod.

[0085] Optionally, the point cloud separation module 22 determines the depth information corresponding to the multiple scanning rods from the depth image based on the recognition results of the multiple scanning rods in the depth image, including: Based on the recognition results of multiple scan bars in the depth image, construct mask images for multiple scan bars; The depth image is cropped using a mask image to obtain depth information corresponding to multiple scan bars.

[0086] Optionally, the model registration module 23 identifies the feature surface point cloud from the scan rod point cloud in the following ways: The point cloud of the scanning rod is classified according to the plane to which it belongs, resulting in multiple subsets of point cloud; Obtain the disorder level of each subset of the point cloud; Based on the degree of disorder of each point cloud subset, the feature surface point cloud is determined, where the feature surface point cloud is the point cloud subset with the highest degree of disorder.

[0087] Optionally, the model registration module 23 classifies the scan rod point cloud according to its plane to obtain multiple point cloud subsets, including: Construct a directed bounding box for the point cloud of the package scanning pole, wherein the directed bounding box includes multiple reference surfaces; The scanning rod point cloud is classified according to its orientation, resulting in multiple point cloud subsets, where each point in each point cloud subset faces the same reference plane.

[0088] Optionally, the model registration module 23 classifies the scan rod point cloud according to orientation to obtain multiple point cloud subsets, including: For each scan point in the scan rod point cloud, obtain the dot product of the normal vector between the scan point and each reference surface; The target reference plane facing the scan point is determined by the dot product of the normal vectors between the scan point and each reference plane, where the dot product of the normal vectors between the target reference plane and the scan point is the largest. The scan points are categorized into a subset of the point cloud corresponding to the target reference surface.

[0089] Optionally, the model registration module 23 obtains the degree of disorder for each subset of the point cloud in the following ways: For each subset of point clouds, obtain the polar angle and azimuth angle of each scan point within it; Based on the preset polar angle spacing and azimuth angle spacing, the polar angle and azimuth angle of each scan point in the point cloud subset are statistically analyzed to obtain the polar angle statistical results and azimuth angle statistical results of the point cloud subset. Based on the polar angle and azimuth angle statistics of the point cloud subset, the Shannon entropy of the polar angle and azimuth angle in the point cloud subset is calculated and used as the disorder of the point cloud subset.

[0090] Optionally, multiple scanning bars come from multiple manufacturers, and each scanning bar includes a classification part 13 representing its manufacturer. The model registration module 23 is also used for: Obtain classification models from multiple manufacturers, where each manufacturer's classification model is the standard model for the corresponding classification part 13; Based on the position of the feature surface point cloud, the classification point cloud is determined from the scanning rod point cloud, wherein the registration part 12 represented by the feature surface point cloud and the classification part 13 represented by the classification point cloud satisfy a preset positional relationship. By comparing the classification point cloud with the classification models of multiple scanning rods, the target manufacturer of the scanning rod corresponding to the scanning rod point cloud is determined.

[0091] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0092] It should also be understood that if the above embodiments are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0093] Therefore, this embodiment also provides a storage medium, which is a computer-readable storage medium. The storage medium stores a computer program, which, when executed by a processor, implements the registration method for the implantation scanning rod provided in this embodiment. The storage medium can be any medium capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0094] Please refer to Figure 6 This embodiment provides an electronic device, which may include a processor 32 and a memory 31. The memory 31 stores a computer program, and the processor reads and executes the computer program in the memory 31 corresponding to the above-described embodiments to implement the registration method for the implantation scanning rod provided in this embodiment.

[0095] See also Figure 6 The electronic device also includes a communication unit 33. The memory 31, processor 32 and communication unit 33 are electrically connected to each other directly or indirectly through system bus 34 to realize data transmission or interaction.

[0096] The memory 31 can be an information recording device based on any electronic, magnetic, optical, or other physical principles, used to record execution instructions, data, etc. In some embodiments, the memory 31 can be, but is not limited to, volatile memory, non-volatile memory, memory drive, etc.

[0097] In some embodiments, the volatile memory may be random access memory (RAM); in some embodiments, the non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory, etc.; in some embodiments, the storage drive may be a disk drive, solid-state drive, any type of storage disk (such as optical disc, DVD, etc.), or similar storage media, or a combination thereof.

[0098] The communication unit 33 is used to send and receive data over a network. In some embodiments, the network may include a wired network, a wireless network, a fiber optic network, a telecommunications network, an intranet, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth network, a ZigBee network, or a near field communication (NFC) network, or any combination thereof. In some embodiments, the network may include one or more network access points. For example, the network may include wired or wireless network access points, such as base stations and / or network switching nodes, through which one or more components of the service request processing system can connect to the network to exchange data and / or information.

[0099] The processor 32 may be an integrated circuit chip with signal processing capabilities, and may include one or more processing cores (e.g., a single-core processor or a multi-core processor). By way of example only, the processor described above may include a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), an Application Specific Instruction-set Processor (ASIP), a Graphics Processing Unit (GPU), a Physics Processing Unit (PPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), a Programmable Logic Device (PLD), a controller, a microcontroller unit, a Reduced Instruction Set Computing (RISC) computer, or a microprocessor, or any combination thereof.

[0100] Understandable. Figure 6 The structure shown is for illustrative purposes only. Electronic devices may also have more advanced features. Figure 6 Showing more or fewer components, or having with Figure 6 The different configurations shown. Figure 6 The components shown can be implemented using hardware, software, or a combination thereof.

[0101] It should be understood that the apparatus and methods disclosed in the above embodiments can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0102] The above descriptions are merely various embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A registration method for an implantation scanning rod, characterized in that, The method includes: Acquire depth images of multiple scanning rods and registration models of the multiple scanning rods, wherein the multiple scanning rods are installed in the patient's oral cavity, and the registration model of each scanning rod is a standard model of the registration site; Based on the recognition results of the plurality of scanning rods in the depth image, a scanning rod point cloud corresponding to each scanning rod is determined from the scanning point cloud based on the depth image; Feature surface point clouds are identified from the scanning rod point cloud, wherein the feature surface point cloud represents the point cloud generated by the plane used for registration of the corresponding scanning rod; The feature surface point cloud is registered with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud. The target registration model is the registration model that is most similar in shape to the feature surface point cloud.

2. The registration method for the implantation scanning rod according to claim 1, characterized in that, Based on the recognition results of the plurality of scanning rods in the depth image, the scanning rod point cloud corresponding to each scanning rod is determined from the scanning point cloud based on the depth image, including: Based on the recognition results of the plurality of scanning rods in the depth image, depth information corresponding to the plurality of scanning rods is determined from the depth image; The reconstructed point cloud based on the depth information is clustered to obtain the scan rod point cloud corresponding to each scan rod.

3. The registration method for the implantation scanning rod according to claim 2, characterized in that, Based on the recognition results of the plurality of scanning rods in the depth image, the depth information corresponding to the plurality of scanning rods is determined from the depth image, including: Based on the recognition results of the multiple scanning rods in the depth image, a mask image of the multiple scanning rods is constructed; The depth image is cropped using the mask image to obtain depth information corresponding to the plurality of scanning rods.

4. The registration method for the implantation scanning rod according to claim 1, characterized in that, Identifying feature surface point clouds from the scanning rod point cloud includes: The point cloud of the scanning rod is classified according to the plane to which it belongs, resulting in multiple point cloud subsets; Obtain the disorder level of each subset of the point cloud; Based on the degree of disorder of each subset of point clouds, a feature surface point cloud is determined, wherein the feature surface point cloud is the subset of point clouds with the highest degree of disorder.

5. The registration method for the implantation scanning rod according to claim 4, characterized in that, The point cloud of the scanning rod is classified according to its plane, resulting in multiple point cloud subsets, including: Construct a directed bounding box that encloses the point cloud of the scanning rod, wherein the directed bounding box includes multiple reference surfaces; The scanning rod point cloud is classified according to its orientation to obtain multiple point cloud subsets, wherein each point in each point cloud subset faces the same reference plane.

6. The registration method for the implantation scanning rod according to claim 5, characterized in that, The scanning rod point cloud is classified according to orientation to obtain the multiple point cloud subsets, including: For each scan point in the scan rod point cloud, obtain the dot product of the normal vector between the scan point and each of the reference surfaces; The target reference surface toward which the scan point faces is determined based on the dot product of the normal vectors between the scan point and each of the reference surfaces, wherein the dot product of the normal vectors between the target reference surface and the scan point is the largest. The scan points are categorized into a subset of the point cloud corresponding to the target reference surface.

7. The registration method for the implantation scanning rod according to claim 4, characterized in that, Obtain the disorder level of each subset of the point cloud, including: For each subset of the point cloud, obtain the polar angle and azimuth angle of each scan point therein; Based on the preset polar angle spacing and azimuth angle spacing, the polar angle and azimuth angle of each scanning point in the point cloud subset are statistically analyzed to obtain the polar angle statistical results and azimuth angle statistical results of the point cloud subset. Based on the polar angle statistics and azimuth statistics of the point cloud subset, the Shannon entropy of the polar angle and azimuth angle in the point cloud subset is calculated and used as the disorder level of the point cloud subset.

8. The registration method for the implantation scanning rod according to any one of claims 1-7, characterized in that, The plurality of scanning bars come from multiple manufacturers, and each scanning bar includes a classification part indicating its manufacturer. The method further includes: Obtain the classification models of the multiple manufacturers, wherein the classification model of each manufacturer is a standard model for the corresponding classification part; Based on the position of the feature surface point cloud, a classification point cloud is determined from the scanning rod point cloud, wherein the registration part represented by the feature surface point cloud and the classification part represented by the classification point cloud satisfy a preset positional relationship. The classification point cloud is compared with the classification models of the multiple scanning rods to determine the target manufacturer of the scanning rod corresponding to the scanning rod point cloud.

9. A registration device for an implantation scanning rod, characterized in that, The device includes: The data preparation module is used to acquire depth images of multiple scanning rods and registration models of the multiple scanning rods, wherein the multiple scanning rods are installed in the patient's oral cavity, and the registration model of each scanning rod is a standard model of the registration site; The point cloud separation module is used to determine the scanning rod point cloud corresponding to each scanning rod from the scanning point cloud based on the depth image according to the recognition results of the multiple scanning rods in the depth image; The model registration module is used to identify feature surface point clouds from the scanning rod point cloud, wherein the feature surface point cloud represents the point cloud generated by the plane used for registration of the corresponding scanning rod; The model registration module is further used to register the feature surface point cloud with the target registration model to obtain the spatial position of the scanning rod corresponding to the scanning rod point cloud, wherein the target registration model is the registration model most similar in shape to the feature surface point cloud.

10. A storage medium, characterized in that, The storage medium stores a computer program that, when executed by a processor, implements the registration method for the implantation scanning rod according to any one of claims 1-7.

11. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing a computer program that, when executed by the processor, implements the registration method of the implantation scanning rod according to any one of claims 1-7.

Images