WO2026032014A1 - Umbrella-type scanning method and system for missing dentition or defective dentition - Google Patents

Abstract

The present invention relates to the field of intraoral scanning for implant restoration, and in particular to an umbrella-type scanning method and system for missing dentition or defective dentition. The method comprises: providing a plurality of positioning bodies, wherein each positioning body is provided with an identification body at least comprising a top feature, a cutting face geometry and curve feature, and an auxiliary rod identification body; presetting a CAD model corresponding to each identification body; acquiring the current tooth design, and acquiring point cloud data of the plurality of positioning bodies by means of an intraoral scanner; preprocessing the point cloud data to remove background noise and a non-target point cloud, and performing automatic identification and segmentation to extract the feature point cloud of each identification body; and using the feature point clouds and the CAD models for processing to output a registration result. In the present invention, during scanning, feature point clouds are accurately extracted by means of an automatic identification and segmentation technology, four coplanar key points are used to construct an initial transformation matrix, and the transformation matrix is optimized by means of an iterative least square method, thereby improving the accuracy of scanning data and the quality of model reconstruction.

Description

An umbrella scanning method and system for edentulous or missing teeth. Technical Field

This invention relates to the field of intraoral scanning for implant restoration, and more particularly to an umbrella-type scanning method and system for missing or damaged teeth.

Background Technology

Digital intraoral scanning is a key technology for digital impression taking in dental implant restoration. Based on optical scanning principles, this technology uses lasers or other light sources for intraoral projection, enabling real-time depth reconstruction and acquisition of local point cloud data. Point cloud registration algorithms are then used to fuse the data, and reconstruction algorithms generate three-dimensional mesh data, allowing for real-time scanning of the morphology of intraoral tissues. During the digital design phase of implant restoration, a scanning rod is attached to the implant or implant abutment to acquire the three-dimensional morphology of the scanning rod, surrounding dentition, and soft tissues. This data is then used to calculate the precise three-dimensional position of the implant.

In existing technologies, the scanning rod is usually cylindrical and mainly located on the side, with a small scanning area at the top and a lack of sufficient geometric features. This makes it impossible for this conventional strategy to quickly and accurately obtain the three-dimensional shape of the scanning rod. The scanning strategy directly affects the scanning accuracy. Previous studies on natural teeth have shown that the method of starting from the occlusal surface, moving to the palatal side, and then to the buccal side has relatively high accuracy.

The scanning process, starting from the occlusal surface and moving to the palatal and then to the buccal surface, further includes scanning the occlusal surface of the teeth, the inner side of the teeth near the palate (palatal side), and the outer side of the teeth near the cheek (buccal side):

Scanning the occlusal surfaces: First, the occlusal surfaces of the teeth are scanned; these are the upper or lower surfaces where the teeth meet. The scanner moves across the teeth, capturing detailed images of the occlusal surfaces.

Scanning the palatal (lingual) side: Next, the scanner moves to the inside of the teeth, that is, the part near the hard palate (lingual side). The dentist or technician will carefully insert the scanner into the mouth while capturing images of the palatal side.

Scan the buccal side: Then, the scanner moves to the outside of the teeth, that is, the part closer to the buccal side.

During the scan, the dentist or technician ensures that every corner of the teeth is scanned, including the contact points between teeth and the junction of the teeth and gums. After the scan is complete, the device integrates the captured data into a complete 3D model. This model can be used for diagnosis, treatment planning, or the fabrication of restorations.

However, conventional scanning rods have a small occlusal surface area and lack obvious characteristic morphology, making them unable to achieve the aforementioned scanning strategy. Therefore, this invention requires a new scanning strategy involving missing or damaged teeth.

Summary of the Invention

The purpose of this invention is to address the aforementioned technical problems by providing an umbrella-type scanning method and system for edentulous or missing teeth. This objective can be achieved through the following technical solutions:

This invention provides an umbrella scanning method for missing or damaged teeth, comprising the following steps:

Multiple positioning bodies are provided. Each positioning body includes a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. Each positioning body is provided with at least a marker body including a top feature, a cutting surface geometry and curve feature, and an auxiliary rod identification body.

Pre-set the CAD model corresponding to each identifier;

To obtain the design for this tooth: Based on the distance between two adjacent implant abutments, a suitable auxiliary rod length is selected, and the scanning rod body is configured and installed on each implant abutment.

Point cloud data of multiple positioning bodies are acquired using an intraoral scanner;

The point cloud data is preprocessed to remove background noise and non-target point clouds, and the feature point cloud of each marker is automatically identified and segmented.

These feature point clouds and CAD models are processed to output registration results; until the registration termination condition is met, the corresponding point cloud model is output.

Furthermore, by processing these feature point clouds with the CAD model, the output registration results include:

Four coplanar key points are selected from the feature point cloud, and a preliminary change matrix is calculated based on the four coplanar key points for preliminary alignment with the CAD model of the scanning rod body.

The initial transformation matrix is used as input for iterative optimization. In each iteration, the corresponding update selection and translation are performed based on the minimum distance error between the position of the feature point cloud and the CAD model.

Furthermore, it also includes calculating the overlap of the two majority feature point clouds under the current transformation during each iteration. The overlap is the ratio of the number of overlapping points to the total number of points, which is used to quantify the alignment accuracy. When the overlap is lower than a preset threshold, the ICP algorithm parameters are adjusted, including the number of iterations, adjusting the search range, or optimizing the distance metric.

Furthermore, the termination conditions include whether the threshold for the number of iterations has been reached, whether the error has reached a predetermined upper limit threshold, or a combination of the number of iterations and the error threshold.

Furthermore, point cloud data of multiple positioning bodies are acquired via an intraoral scanner, including:

Using an intraoral scanner, the implant is scanned first from the top to obtain three-dimensional position data of multiple positioning elements, providing accurate reference points for subsequent scans;

The scanning begins from the occlusal surface of the oral cavity and is performed on the top of the scanning bar installed on the implant, with locking scanning based on the top features of the scanning bar;

Scan the scanning rod from the buccal and lingual sides of the oral cavity to identify the geometric and curvature features of the cutting surface and obtain point cloud data of the contact between the scanning rod and adjacent teeth, the scanning rod or soft tissue;

Scanning the gingival soft tissue, which has a smaller impact on accuracy, and improving the scanning of soft tissue.

Based on the same inventive concept, the present invention also provides an umbrella-type scanning system for missing or damaged teeth, comprising multiple positioning bodies, a scanner head, and a scanning control device:

Multiple positioning bodies are included, each comprising a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. Each positioning body is equipped with at least a top feature, cutting surface geometry and curve features, and an auxiliary rod identification body. An appropriate auxiliary rod length is selected based on the distance between two adjacent implant composite abutments. The scanning rod body is configured and installed on the composite abutment of each implant.

The scanner head acquires point cloud data of multiple positioning objects through an intraoral scanner;

The scanning control device is configured to: pre-set the CAD model corresponding to each marker;

The point cloud data is preprocessed to remove background noise and non-target point clouds, and the feature point cloud of each marker is automatically identified and segmented.

These feature point clouds and CAD models are processed to output registration results; until the registration termination condition is met, the corresponding point cloud model is output.

Compared with the prior art, the present invention has at least one of the following technical advantages:

This invention provides a digital umbrella scanning method for missing or damaged teeth. It scans multiple positioning devices with specific geometric and curvilinear features mounted on implants. Each positioning device includes a scanning rod body equipped with a lateral auxiliary rod. These positioning devices correspond to a pre-set CAD model, allowing adjustment of the auxiliary rod length based on the distance between implants, thus ensuring precise installation. During scanning, point cloud data collected by an intraoral scanner is preprocessed to remove background noise and non-target point clouds. Feature point clouds are accurately extracted using automatic recognition and segmentation technology. An initial transformation matrix is constructed using four coplanar key points to achieve preliminary alignment between the point cloud data and the CAD model. The transformation matrix is further optimized using iterative least squares, finely adjusting the alignment and effectively reducing errors between the point cloud and the model. This improves the accuracy of the scanned data and the quality of model reconstruction, significantly enhancing the accuracy of dental implant restoration design and the reliability of implementation. It solves the problem that traditional intraoral scanning techniques often suffer from the difficulty in accurately capturing complex internal details due to the simple geometry of the scanning rod (e.g., cylindrical) and the small area of the occlusal surface, lacking sufficient geometric features.

Attached Figure Description

To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below:

Figure 1 is a flowchart of the umbrella scanning method in an embodiment of the present invention;

Figure 2 is a schematic diagram of the installation of the positioning body under tooth loss in an embodiment of the present invention;

Figure 3 is a data processing flowchart of the oral scanner scanning positioning body in an embodiment of the present invention;

Figure 4 is a point cloud matching calculation diagram of the oral scanner scanning positioning body in an embodiment of the present invention.

Detailed Implementation

The invention will now be described with reference to specific embodiments and certain accompanying drawings, but the invention is not limited thereto. Any drawings described are merely illustrative and not limiting. In the drawings, for illustrative purposes, the dimensions of some elements may be enlarged and not drawn to scale. Dimensions and relative dimensions do not necessarily correspond to actual instances of implementing the invention.

Furthermore, the terms first, second, third, etc., used in the specification and claims are used to distinguish similar elements and are not necessarily used to describe a sequence or chronological order. The terms are interchangeable where appropriate, and embodiments of the invention can operate in sequences other than those described or shown herein.

Furthermore, the terms "top," "bottom," "upper," "lower," "side," "side," etc., used in the specification and claims are for descriptive purposes and not necessarily for describing relative positions. Such terms are interchangeable where appropriate, and embodiments of the invention described herein can operate in directions other than those described or shown herein.

The term "comprising" as used in the claims should not be construed as limited to the devices listed herein; it does not exclude other elements or steps. It should be interpreted as specifying the presence of the stated feature, integer, step, or component, but does not exclude the presence or addition of one or more other features, integers, steps, or components, or groups thereof. Therefore, the scope of the expression "device comprising devices A and B" should not be limited to devices consisting solely of components A and B, meaning that, relative to the invention, the only relevant components of the device are A and B.

In the context of this invention, a marker is understood to include any 2D or 3D geometric object capable of defining a predetermined reference point, the coordinates of which can be determined. The types of suitable markers are essentially unlimited. Preferred markers have regular geometric shapes because the predetermined reference point of the marker can be easily defined, or even intuitively. Suitable markers can be, for example, equilateral triangles, squares, rectangles, circles, or regular polygons, where, for example, the center point can be considered the predetermined reference point. Suitable markers can also be line segments, where points can be considered the predetermined reference points. Suitable 3D markers can be, for example, spheres, hemispheres, or cubes, whose center points can then be considered the predetermined reference points. The applicant has found that the size and shape of the markers can vary within a very wide range, and markers of different shapes and/or sizes can be used as part of the same pattern. The applicant has found that the contrast between the marker and its surroundings can be further enhanced by adding strongly contrasting edges around the marker.

In the context of this invention, pre-calibration is understood to mean the identification process, including its results, in which identification techniques are applied to identify, in a clinical setting, a given type of implant or various plant connections, in order to obtain all the information needed to retrieve the implant's position and orientation from a pattern formed by markers as part of the scan localization. The information thus collected forms reference information in the process of associating the visualization pattern with the type of implant or equivalent in the clinical setting presented in the oral cavity.

In the context of this invention, the “vector” of a dental implant or implant root is defined as spatial information that first defines the orientation axis of the dental implant root. The carrier also includes the position of the top of the dental implant root along the orientation axis. Therefore, the top is defined as the side corresponding to the height of the implant root.

The creation process of this invention will be described.

Clinical research by the inventors has revealed that current intraoral scanning technology faces insufficient accuracy in edentulous jaw scanning. Because tooth loss in edentulous patients results in a lack of distinct geometric features in the soft tissue, the availability of necessary geometric features during scanning is limited. Furthermore, the mucosal soft tissues such as the buccal and lingual sides of the mouth may introduce data noise during scanning. As the amount of scanned data increases, the cumulative error gradually increases, further reducing scanning accuracy. Although some studies have attempted to enhance the geometric features of the soft tissue transition phase by adding auxiliary devices to the scanning rod, this method has achieved some success in improving the accuracy of edentulous implant scanning, but it still does not reach the level of accuracy comparable to single-tooth or multi-tooth implant restorations. In addition, the conventional implant scanning strategy usually involves scanning the occlusal surface of the dentition first, followed by the buccal or lingual sides of the dentition. This method has been shown to provide high scanning accuracy in single-tooth or multi-tooth implants. In natural teeth and crown restorations, the soft tissue transition area is the crown contour. For edentulous implant restorations, this transition area is primarily soft tissue.

In modern dental implant and restoration procedures, obtaining accurate intraoral three-dimensional images is crucial for successful edentulous implant restorations. Currently, most intraoral scanners rely on optical technology, using lasers or other light sources to scan within the oral cavity and obtain detailed images of the implant and surrounding soft tissues. However, for edentulous patients, because implants are typically embedded under the bone and reliable geometric landmarks are lacking in the oral cavity, traditional cylindrical scanning probes often fail to provide sufficient accuracy to guide precise implant placement. This is primarily because the shape and size of traditional scanning probes limit their intraoral scanning range and landmark capture capabilities.

In view of the limitations of existing technologies, this invention proposes an umbrella scanning method and system for edentulous or missing teeth. The system consists of multiple positioning bodies, each including a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. The positioning body includes not only top features and cutting surface geometry, but also the auxiliary rod on the side of the scanning rod body, used to enhance geometric feature recognition during the soft tissue transition phase. Combined with automatic feature point cloud matching and iterative optimization using the ICP algorithm, the alignment accuracy between point cloud data and the model is improved. By selecting four coplanar key points to construct a transformation matrix and performing continuous iterations, the optimization process considers the overlap of point clouds and the minimum distance error, further refining the alignment. By setting termination conditions, the final output model is ensured to reach a high accuracy standard. This umbrella scanning method is not only suitable for conventional edentulous restorations, but also particularly suitable for complex edentulous jaw implant restorations.

First Embodiment

In this embodiment, for cases of missing or severely damaged teeth, conventional scanning methods often struggle to capture accurate three-dimensional data due to the lack of stable reference points. This embodiment introduces multiple positioning rods, including at least top features, cutting surface geometry and curve features, and auxiliary rod recognition bodies, to enhance feature recognition capabilities during intraoral scanning and ensure data accuracy. Simultaneously, an automated feature point cloud registration and alignment algorithm is employed to improve the automation level of data processing and the accuracy of the final results. The specific implementation scheme is as follows:

As shown in Figures 1 to 4, this invention provides an umbrella scanning method for missing or damaged teeth. This method uses multiple positioning bodies, each comprising a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. Each positioning body is equipped with at least a marker body including top features, cutting surface geometry and curve features, and an auxiliary rod identification body. Each identification body is relative to a pre-set CAD model, ensuring the accuracy and reproducibility of the scan. During the scanning process, firstly, an appropriate auxiliary rod length is selected based on the distance between two adjacent implant composite abutments. The scanning rod body is configured and installed on each implant composite abutment. An intraoral scanner is used to collect point cloud data of the positioning bodies. Then, the point cloud data is preprocessed to remove background noise and non-target point clouds. Through edge detection and curve fitting, the feature point cloud of each marker body is automatically identified and segmented. These feature point clouds are then processed with the CAD model to output the registration result. The process continues until the registration termination condition is met, at which point cloud models are output.

The process involves processing these feature point clouds with a CAD model. The resulting registration output includes: first, selecting four coplanar key points from the feature point clouds and calculating a preliminary transformation matrix based on these key points for initial alignment with the CAD model of the scanning rod; then, using the preliminary transformation matrix as input for iterative optimization, and updating and translating accordingly based on the minimum distance error between the feature point cloud and the CAD model in each iteration. During each iteration, the overlap between the two majority feature point clouds under the current transformation is calculated. The overlap is the ratio of the number of overlapping points to the total number of points, used to quantify the alignment accuracy. When the overlap falls below a preset threshold, the ICP algorithm parameters are adjusted, including the number of iterations, the search range, or the distance metric. Termination conditions include whether the iteration threshold is reached, whether the error reaches a predetermined upper limit, or a combination of the iteration number and error threshold.

The specific registration process is shown in Figure 4. Four coplanar key points are determined. First, vectors ac and ab are calculated using the cross product formula. Calculate the corresponding normal vector to vector It is perpendicular to the plane formed by points a, b, and c. Verify whether the fourth point d is coplanar using the normal vector by calculating vector ad and the normal vector. Given a point set, considering numerical precision, if the point set is close to zero, then point d also lies on the plane. Next, a preliminary transformation matrix is estimated using four coplanar points. Specifically, this involves finding the geometric center e of the four points and the center e′ of the corresponding point in the target point cloud. The optimal rotation matrix R and translation vector t are calculated using the positions of the four points in the original and target clouds. Specifically, the least squares method is used to determine the rotation matrix R and translation vector t in point cloud registration to minimize the total registration error, i.e., by minimizing ∑||(R* _pi +t) _-p′i || ^² , where _pi and _p′i are the corresponding points in the original and target point clouds.

Once the initial transformation matrix is obtained, optimization is performed iteratively. In each iteration, the nearest corresponding point is found for the current transformed point. Based on the newly found corresponding point pair, the transformation matrix is recalculated to further reduce the average distance between points. The algorithm terminates when the amount of the transformation matrix update is less than a certain threshold, or the preset number of iterations is reached, or the error between point clouds is small enough.

As shown in Figures 2 and 4, the multiple scanning poles are designed with an umbrella-shaped scanning section and corresponding connecting parts. The top of the umbrella-shaped scanning section is equipped with a first top surface, on both sides of which are symmetrically arranged first and second cutting surfaces with unique geometric shapes. These cutting surfaces extend outward from the first top surface, forming clear boundaries with each other and with the first top surface. This constitutes a multi-anisotropic structure at the top, greatly enhancing the scanning pole's recognition and positioning capabilities in intraoral scanning equipment, especially from the top view of the pole. Furthermore, the auxiliary pole in this design has a unique function: eliminating noise and unnecessary data generated during the scanning process. Through its specially designed structure and materials, the auxiliary pole can reduce reflection and scattering in the scanned data, optimizing data quality. During the data processing stage, the configuration of the auxiliary pole allows the algorithm to more effectively separate and eliminate non-target point clouds, improving the clarity and usability of the final image. This integrated elimination function not only simplifies subsequent data processing steps but also provides more accurate and reliable scanning results.

Therefore, by integrating registration algorithms and optimizing them for the complex geometric features of the top of the umbrella-shaped scanning section, it is possible to quickly and accurately identify and locate the unique data points generated by the umbrella structure, effectively processing them. This not only enhances the automation level of the scanning process but also ensures high data reproducibility and precise alignment, greatly supporting complex dental restorations and implant procedures.

Simultaneously, point cloud data of multiple positioning bodies are acquired using an intraoral scanner. As shown in Figure 2, the intraoral scanner first scans the implant from the top to obtain three-dimensional position data of multiple positioning bodies, providing accurate reference points for subsequent scans. Then, the top of the scanning bar installed on the implant is scanned starting from the occlusal surface of the oral cavity, and a locking scan is performed based on the top features of the scanning bar. That is, precise positioning is achieved through the top features of the scanning bar, and the locking scan ensures that the scanning bar is horizontally aligned with the floor of the oral cavity. Next, the scanning bar is scanned on the buccal and lingual sides of the oral cavity to identify the cutting geometry and curve features of the scanning bar, and to obtain point cloud data of the contact between the scanning bar and adjacent teeth, the scanning bar, or soft tissue, thereby improving the overall scanning accuracy. Finally, the scanning of the gingival soft tissue, which has a smaller impact on accuracy, is also included to complete the soft tissue scan and ensure the comprehensiveness of the data.

Second Embodiment

An umbrella-type scanning system for missing or damaged teeth includes multiple positioning elements, a scanner head, and a scanning control device.

Multiple positioning bodies are included, each comprising a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. Each positioning body is equipped with at least a top feature, cutting surface geometry and curve features, and an auxiliary rod identification body. An appropriate auxiliary rod length is selected based on the distance between two adjacent implant composite abutments. The scanning rod body is configured and installed on the composite abutment of each implant.

The scanner head acquires point cloud data of multiple positioning objects through an intraoral scanner;

The scanning control device is configured to: pre-set the CAD model corresponding to each marker;

The point cloud data is preprocessed to remove background noise and non-target point clouds, and the feature point cloud of each marker is automatically identified and segmented.

These feature point clouds and CAD models are processed to output registration results; until the registration termination condition is met, the corresponding point cloud model is output.

Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims (5)

An umbrella scanning method for edentulous or missing teeth, characterized by comprising the following steps: Multiple positioning bodies are provided, each of which includes a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. Each positioning body is provided with an identifier including at least a top feature, a cutting surface geometry and curve feature, and an auxiliary rod identification body. Pre-set the CAD model corresponding to each of the aforementioned identifiers; To obtain the design for this tooth: A suitable auxiliary rod length is selected based on the distance between two adjacent implant abutments. The scanning rod body is then configured and mounted onto each implant abutment. Point cloud data of multiple positioning bodies are acquired using an intraoral scanner; The point cloud data is preprocessed to remove background noise and non-target point clouds, and the feature point cloud of each of the markers is automatically identified and segmented. These feature point clouds are processed with the CAD model to output registration results; until the registration termination condition is met, the corresponding point cloud model is output, wherein, The registration results obtained by processing these feature point clouds with the CAD model further include: Four coplanar key points are selected through the feature point cloud, and a preliminary change matrix is calculated based on the four coplanar key points for preliminary alignment with the CAD model of the scanning rod body. The initial transformation matrix is used as input for iterative optimization. In each iteration, the corresponding update selection and translation are performed based on the minimum distance error between the position of the feature point cloud and the CAD model. The umbrella scanning method as described in claim 1 is characterized in that it further includes, in each iteration, calculating the overlap of two majority feature point clouds under the current transformation, wherein the overlap is the ratio of the number of overlapping points to the total number of points, used to quantify the alignment accuracy; when the overlap is lower than a preset threshold, adjusting the ICP algorithm parameters, wherein the parameters include the number of iterations, adjusting the search range, or optimizing the distance metric. The umbrella scanning method as described in claim 2 is characterized in that the termination condition includes whether the threshold of the number of iterations is reached, whether the error reaches a predetermined upper limit threshold, or a combination of the number of iterations and the error threshold. The umbrella scanning method as described in claim 1, characterized in that acquiring point cloud data of multiple positioning bodies via an intraoral scanner includes: The intraoral scanner is used to scan the implant from the top first to obtain three-dimensional position data of multiple positioning bodies, providing accurate reference points for subsequent scans; The scanning begins from the occlusal surface of the oral cavity and scans the top of the scanning bar installed on the implant, and a locking scan is performed based on the top features of the scanning bar; The scanning bar is scanned from the buccal and lingual sides of the oral cavity to identify the geometry of the cutting surface and the curve features, and to obtain the point cloud data of the contact between the scanning bar and adjacent teeth, the scanning bar or soft tissue; Scanning the gingival soft tissue, which has a smaller impact on accuracy, and improving the scanning of soft tissue. An umbrella-type scanning system for missing or damaged teeth includes multiple positioning elements, a scanner head, and a scanning control device. Multiple positioning bodies are provided, each comprising a scanning rod body and an auxiliary rod disposed on the side of the scanning rod body. Each positioning body is provided with at least a top feature, cutting surface geometry and curve features, and an auxiliary rod identification body. An appropriate auxiliary rod length is selected based on the distance between two adjacent implant composite abutments. The scanning rod body is configured and installed on the composite abutment of each implant. The scanner head acquires point cloud data of multiple positioning objects through an intraoral scanner; The scanning control device is configured to: pre-set the CAD model corresponding to each marker; The point cloud data is preprocessed to remove background noise and non-target point clouds, and the feature point cloud of each marker is automatically identified and segmented. The feature point cloud is processed with the CAD model to output the registration result; until the registration termination condition is met, the corresponding point cloud model is output.