An intelligent acquisition and processing system for personalized denture processing parameters

Abstract

The present invention relates to the technical field of denture processing data handling, and particularly relates to an intelligent acquisition and processing system for personalized denture processing parameters. The system includes: a denture data acquisition module for acquiring point cloud data within a patient's oral cavity; a denture data processing module for analyzing the uniform distribution degree of the point cloud data to evaluate the density uniformity of each denture model; constructing a neighborhood adjustment factor by analyzing the distribution of denture model clustering clusters and combining with the density uniformity; comparing the difference between the neighborhood adjustment factor and the average factor of the oral model to construct the neighborhood radius of the oral model; evaluating the bending degree of each point cloud data in the oral model at the fitting surface to determine the abnormal weight coefficient; and a denture model generation module for processing the dentures of patients to be installed. This application aims to eliminate abnormal points in the patient's oral cavity point cloud data and improve the precision and accuracy of personalized denture processing.

Description

Technical Field

[0001] The present invention relates to the technical field of denture processing data processing, and particularly relates to an intelligent acquisition and processing system for personalized denture processing parameters. Background Art

[0002] Since the sizes, shapes, and arrangements of teeth of each person are different, when processing dentures, it is necessary to collect parameters according to the actual situation of each person and customize personalized dentures. Design parameters are key factors directly affecting the denture design results. The setting of these parameters will affect the complexity, accuracy, and aesthetics of the design. The acquisition of design parameters is divided into contact measurement methods and digital measurement methods. The contact measurement method uses impression materials to collect the morphology of oral teeth, and the operation is relatively cumbersome and the efficiency is low. The digital measurement method uses methods such as laser scanning or optical scanning, which can quickly collect data of teeth with complex shapes, and the data acquisition accuracy is relatively high.

[0003] Although the acquisition accuracy of the digital measurement method is relatively high, in actual operation, due to the complexity of the patient's oral environment, the collected data may be deviated or noisy. When using the radius-based outlier detection algorithm to detect outliers in the point cloud data, the values of the neighborhood radius and the outlier discrimination threshold usually take empirical values. Too large or too small values will affect the accuracy of outlier detection, thereby reducing the accuracy and accuracy of personalized denture processing. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide an intelligent acquisition and processing system for personalized denture processing parameters, and the specific technical solutions adopted are as follows:

[0005] The present invention proposes an intelligent acquisition and processing system for personalized denture processing parameters, and the system includes:

[0006] A denture data acquisition module, which is used to obtain the three-dimensional point cloud data in the oral cavity of a large number of patients with dentures installed in their oral cavities, and record it as a denture model, and obtain the three-dimensional point cloud data in the oral cavity of patients waiting to install dentures, and record it as an oral model;

[0007] A denture data processing module, which is used to cluster all the point cloud data in each denture model to obtain multiple clustering clusters and the local density of each point cloud data; comprehensively consider the extreme distribution and dispersion degree of the local density of all the point cloud data in each denture model, and combine the total number of clustering clusters to evaluate the density uniformity of each denture model;

[0008] By measuring the distribution of the distances from each cluster in each denture model to the remaining clusters, the average inter-cluster distance of each denture model is obtained, and in combination with the average distribution of the local densities of all the point cloud data in each denture model and the density uniformity, a neighborhood adjustment factor for each denture model is constructed to determine the average factor;

[0009] For the oral cavity model, according to the construction method of the neighborhood adjustment factor, the neighborhood adjustment factor of the oral cavity model is obtained; the difference between the neighborhood adjustment factor of the oral cavity model and the average factor is compared, and in combination with the distances between all the point cloud data, the neighborhood radius of the oral cavity model is determined;

[0010] A fitting surface is obtained by fitting all the point cloud data in the oral cavity model, and the bending degrees of each point cloud data in the oral cavity model and all the point cloud data within its neighborhood radius at the fitting surface are respectively evaluated to determine the abnormal weight coefficient of each point cloud data in the oral cavity model;

[0011] A denture model generation module is used to determine approximate normal points and approximate abnormal points based on the abnormal weight coefficients of all the point cloud data in the oral cavity model, and respectively analyze the distribution of the point cloud data within the neighborhood radii of each approximate normal point and each approximate abnormal point to determine the abnormal discrimination value of the oral cavity model, and process the denture for the patient whose denture is to be installed.

[0012] Preferably, the evaluation method for the density uniformity of each denture model is as follows:

[0013] The range and variance of the local densities of all the point cloud data in each denture model are respectively calculated, and the reciprocal of the product of the number of all the clusters in each denture model, the range and the variance is used as the density uniformity of each denture model.

[0014] Preferably, the method for obtaining the average inter-cluster distance of each denture model is as follows:

[0015] In each denture model, the distance from the cluster center of each cluster to the cluster centers of all the remaining clusters is calculated, and the minimum value among the distances is denoted as the cluster distance of each cluster;

[0016] The average inter-cluster distance of each denture model is the average level of the cluster distances of all the clusters in each denture model.

[0017] Preferably, the expression of the neighborhood adjustment factor of each denture model is: ; where represents the neighborhood adjustment factor of the i-th denture model; represents the average inter-cluster distance of the i-th denture model; represents the mean of the local densities of all the point cloud data in the i-th denture model; Denotes the density uniformity of the i-th denture model.

[0018] Preferably, the average factor is the mean of the neighborhood adjustment factors of all denture models.

[0019] Preferably, the method for determining the neighborhood radius of the oral cavity model is as follows:

[0020] Calculate the mean of the distances between all point cloud data in the oral cavity model, denoted as the average distance, and take a preset multiple of the average distance as the initial neighborhood radius of the oral cavity model;

[0021] The neighborhood radius of the oral cavity model The expression is: ; In the formula, Denotes the initial neighborhood radius; Denotes the neighborhood adjustment factor of the oral cavity model; Denotes the average factor; tanh( ) represents the hyperbolic tangent function.

[0022] Preferably, the method for determining the abnormal weight coefficient of each point cloud data in the oral cavity model is as follows:

[0023] Calculate the principal curvatures of each point cloud data in the oral cavity model and all point cloud data within its neighborhood radius at the fitting surface. The principal curvatures include the maximum curvature and the minimum curvature. Calculate the sum of the maximum curvature and the minimum curvature of each point cloud data, denoted as the curvature sum value of each point cloud data;

[0024] Calculate the variance of the maximum curvature among the principal curvatures of all point cloud data within the neighborhood radius of each point cloud data in the oral cavity model, denoted as the curvature variance of each point cloud data in the oral cavity model;

[0025] The abnormal weight coefficient of the point cloud data j in the oral cavity model The expression is: ; In the formula, Denotes the curvature sum value of the point cloud data j in the oral cavity model; Denotes the curvature variance of the point cloud data j in the oral cavity model; Denotes the number of all point cloud data in the cluster where the point cloud data j is located among all clusters obtained by clustering all point cloud data in the oral cavity model.

[0026] Preferably, the method for determining the approximate normal points and approximate abnormal points is as follows:

[0027] Take the abnormal weight coefficients of all point cloud data in the oral cavity model as the input of the threshold segmentation algorithm, output the segmentation threshold, and take the point cloud data with abnormal weight coefficients greater than the segmentation threshold as approximate abnormal points, and take the point cloud data with abnormal weight coefficients less than or equal to the segmentation threshold as approximate normal points.

[0028] Preferably, the method for determining the abnormal discrimination value of the oral cavity model is as follows:

[0029] Count the total number of point cloud data within the neighborhood radius of all approximately normal points in the oral cavity model, and take the maximum value among the total numbers as the first maximum value of the oral cavity model;

[0030] Count the total number of point cloud data within the neighborhood radius of all approximately abnormal points in the oral cavity model, and take the maximum value among the total numbers as the second maximum value of the oral cavity model;

[0031] The abnormal discrimination value of the oral cavity model is the average value of the first maximum value and the second maximum value of the oral cavity model.

[0032] Preferably, the processing of the denture for the patient to be installed includes:

[0033] The point cloud data with the total number of point cloud data within the neighborhood radius in the oral cavity model less than the abnormal discrimination threshold is regarded as abnormal points, and is removed from the oral cavity model, and all the point cloud data in the oral cavity model after removing the abnormal points is used as the input of the three-dimensional modeling software, and the denture model to be processed for the patient to be installed is output.

[0034] The present invention has the following beneficial effects:

[0035] By analyzing the uniform distribution of the three-dimensional point cloud data in the oral cavity of the patient with dentures installed, the present application constructs the density uniformity, which can more accurately evaluate the distribution characteristics of the point cloud data in the denture model, thereby providing a basis for subsequent adjustment of the neighborhood radius and reducing the parameter error set manually; further, by clustering all the point cloud data in the denture model to obtain multiple clustering clusters, according to the distribution of the clustering clusters and combining with the density uniformity, the neighborhood adjustment factor is constructed, making the selection of the neighborhood radius more reasonable, thereby reducing the risk of misjudging normal points as abnormal points and reducing the possibility of missing abnormal points; further, by comparing the difference between the neighborhood adjustment factor and the average factor of the patient to be installed with dentures, the neighborhood radius of the oral cavity model is constructed, which helps to improve the accuracy and reliability of the final denture model; further, by analyzing the bending degree of the fitting surface formed by the point cloud data in the oral cavity model, the abnormal weight coefficient of the point cloud data is constructed, which helps to more accurately identify abnormal points and normal points in the point cloud data. By removing abnormal points, the quality of the point cloud data is improved, and further the accuracy of the denture processing for the patient to be installed with dentures is improved. According to the characteristics of the point cloud data in each patient's oral cavity, the present application removes the interference of abnormal points in the point cloud data and accurately generates a personalized denture model suitable for each patient, improving the accuracy and accuracy of personalized denture processing. Description of the Drawings

[0036] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0037] Figure 1 The block diagram of an intelligent acquisition and processing system for personalized denture processing parameters provided by an embodiment of the present application;

[0038] Figure 2 The schematic diagram of the neighborhood radius acquisition process provided by an embodiment of the present application. Detailed implementation manners

[0039] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following, in combination with the drawings and preferred embodiments, will detail the specific implementation manners, structures, features, and effects of an intelligent acquisition and processing system for personalized denture processing parameters proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0040] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0041] The following will specifically describe the specific solution of an intelligent acquisition and processing system for personalized denture processing parameters provided by the present invention in combination with the drawings.

[0042] Please refer to Figure 1 , which shows the block diagram of an intelligent acquisition and processing system for personalized denture processing parameters provided by an embodiment of the present invention. The system includes: a denture data acquisition module 101, a denture data processing module 102, and a denture model generation module 103.

[0043] The denture data acquisition module 101 is used to obtain the three-dimensional point cloud data in the oral cavity of a large number of patients with dentures installed in their oral cavities, and record it as a denture model, and obtain the three-dimensional point cloud data in the oral cavity of patients waiting to install dentures, and record it as an oral cavity model.

[0044] A digital oral scanner is used to scan the oral cavity of a patient. The scanner utilizes the principles of laser reflection and optical imaging to obtain oral point cloud data with color information. For a large number of patients with dentures in their oral cavities, three-dimensional point cloud data of their oral cavities is obtained and recorded as a denture model, and three-dimensional point cloud data of the oral cavity of a patient waiting to have a denture installed is obtained and recorded as an oral cavity model.

[0045] It should be noted that for a large number of patients with dentures in their oral cavities, the collected point cloud data is oral point cloud data of different ages, genders, and denture positions.

[0046] The denture data processing module 102 is used to cluster all the point cloud data in each denture model to obtain multiple clustering clusters and the local density of each point cloud data; synthesize the extreme distribution and dispersion degree of the local density of all the point cloud data in each denture model, and combine the total number of clustering clusters to evaluate the density uniformity of each denture model; obtain the average inter-cluster distance of each denture model by measuring the distribution of the distances from each clustering cluster to the other clustering clusters in each denture model, and combine the average distribution of the local density of all the point cloud data in each denture model and the density uniformity to construct a neighborhood adjustment factor for each denture model to determine the average factor; for the oral cavity model, obtain the neighborhood adjustment factor of the oral cavity model according to the construction method of the neighborhood adjustment factor; compare the difference between the neighborhood adjustment factor of the oral cavity model and the average factor, and combine the distances between all the point cloud data to determine the neighborhood radius of the oral cavity model; fit all the point cloud data in the oral cavity model to obtain a fitting surface, and respectively evaluate the bending degree of each point cloud data in the oral cavity model and all the point cloud data within its neighborhood radius at the fitting surface to determine the abnormal weight coefficient of each point cloud data in the oral cavity model.

[0047] When using an oral scanner to scan the oral cavity of a patient with a denture, since the oral cavity may contain metal restorations, metal restorations are usually made of highly reflective materials, which will produce strong reflections during the laser scanning process, resulting in the scattering or reflection of the scanning beam on the metal surface, thereby increasing the noise points and abnormal points in the point cloud data. In addition, if there is saliva on the tooth surface during the scanning process, the saliva will affect the reflection and absorption of the laser beam, resulting in data loss or increased noise in the scanned point cloud data. Therefore, it is necessary to detect outliers in the point cloud data. When using the radius-based method to detect outliers in the point cloud data, it is determined whether the number of points within the neighborhood of a given radius for each point is less than the set outlier discrimination threshold. If it is less than the outlier discrimination threshold, then the point is determined to be an outlier. The setting of the neighborhood radius and the outlier discrimination threshold is usually manually set, and too large or too small values will affect the accuracy of outlier detection.

[0048] The more evenly the point cloud data is distributed, the smaller the density difference between normal points and abnormal points. Therefore, the neighborhood radius should be appropriately reduced to more accurately represent the neighborhood characteristics of each point cloud data. The oral conditions of different patients are different, and the degree of uniformity of the point cloud data is also different. When there are metal restorations in the patient's oral cavity, due to the strong reflective characteristics of the metal restorations, the point cloud at the location of the metal restorations will be relatively dense, while the reflective characteristics of the normal tooth surface are slightly weaker than those of the metal restorations. Therefore, the point cloud at the location of the normal teeth will be slightly sparser than the point cloud at the location of the metal restorations. When there are dental caries in the patient's oral cavity, the reflective characteristics of the dental caries are slightly weaker than those of the normal teeth. Therefore, the point cloud at the location of the dental caries will be sparser than the point cloud of the normal teeth. For the abnormal points formed by the reflection and scattering of metal restorations and saliva, the metal restorations have strong reflective characteristics, so the abnormal points near the metal restorations are relatively dense; while the area of saliva is usually small, and the number of abnormal points at the location of saliva is usually sparse.

[0049] Therefore, according to the characteristics of different tissues in the oral cavity, different neighborhood radii are set to detect the presence of abnormal points. Specifically:

[0050] S1: Cluster all the point cloud data in each denture model to obtain multiple clusters and the local density of each point cloud data; comprehensively consider the extreme distribution and dispersion degree of the local density of all the point cloud data in each denture model, and combine the total number of clusters to evaluate the density uniformity of each denture model.

[0051] (1) First, in order to distinguish different tissue components in the oral cavity of patients with dentures and more accurately identify the characteristics of the dentures, the three-dimensional coordinates of all the point cloud data in each denture model are used as the input of the density clustering algorithm to obtain multiple clusters and the local density of each point cloud data.

[0052] It should be noted that there are many commonly used density clustering algorithms. In this embodiment, the DPC density clustering algorithm is used to obtain the local density and clusters of each point. The implementer can also use the DBSCAN density clustering algorithm to obtain the local density and clusters of each point. This embodiment does not make special restrictions.

[0053] Among them, the DPC density clustering algorithm is a well-known technology, and its specific clustering principle and the specific process of obtaining the local density will not be elaborated here.

[0054] (2) Second, comprehensively consider the extreme distribution and dispersion degree of the local density of all the point cloud data in each denture model, and combine the total number of clusters to evaluate the density uniformity of each denture model. Specifically:

[0055] Calculate the range and variance of the local density of all the point cloud data in each denture model respectively, and take the reciprocal of the product of the number of all the clustering clusters, the range, and the variance in each denture model as the density uniformity of each denture model.

[0056] As can be understood from the density uniformity of each denture model, the smaller the range of the local density of all the point cloud data in the denture model, the smaller the difference in the density between the point cloud data with the maximum local density and the point cloud data with the minimum local density, which means that the distribution of the point cloud data in the denture model is more uniform; the smaller the variance of the local density of the point cloud data in the denture model, the more uniform the distribution of the point cloud data in the denture model; the fewer the number of clustering clusters, the more uniform the distribution of the point cloud data in the denture model; therefore, if the range of the local density of all the point cloud data in the denture model is smaller, the variance of the local density of the point cloud data in the denture model is smaller, and the number of clustering clusters is fewer, then the density uniformity of the denture model is greater, indicating that the distribution of the point cloud data in the denture model is more uniform.

[0057] On the contrary, if the range of the local density of all the point cloud data in the denture model is larger, it means that the difference in the density between the point cloud data with the maximum local density and the point cloud data with the minimum local density is greater, that is, the distribution of the point cloud data in the denture model is more dispersed; and if the variance of the local density of the point cloud data in the denture model is larger and the number of clustering clusters is more, then the density uniformity of the denture model is smaller, indicating that the distribution of the point cloud data in the denture model is more dispersed.

[0058] S2: By measuring the distribution of the distances from each clustering cluster to the other clustering clusters in each denture model, obtain the average inter-cluster distance of each denture model, and combine the average distribution of the local density of all the point cloud data in each denture model and the density uniformity to construct the neighborhood adjustment factor of each denture model to determine the average factor.

[0059] When the point cloud data is very dense, setting a smaller neighborhood radius can better capture the local features of the points; when the point cloud data is sparser, a larger neighborhood radius needs to be set to avoid mislabeling normal points as abnormal points. When the distance between the clustering clusters is large, it means that the distance between the abnormal point clustering cluster and the normal point clustering cluster is large, so appropriately increasing the neighborhood radius can avoid the problem that the number of points in the abnormal point neighborhood and the number of points in the normal point neighborhood are close due to the aggregation of abnormal points, and can more accurately distinguish normal points and abnormal points.

[0060] Therefore, according to the distribution characteristics of the point cloud data, construct the neighborhood adjustment factor of each denture model to obtain the average factor, specifically:

[0061] (1) In each denture model, calculate the distance from the cluster center of each cluster to the cluster centers of all other clusters, and record the minimum value among the distances as the cluster distance of each cluster;

[0062] The average inter-cluster distance of each denture model is the average level of the cluster distances of all clusters in each denture model.

[0063] It should be noted that there are many methods to measure the distance between points. In this embodiment, the Euclidean distance between the cluster centers of each cluster and the cluster centers of all other clusters is calculated to measure the distance between the cluster centers of different clusters. Implementers can also use other methods to measure the distance between points, such as Manhattan distance or DTW distance. Regarding the selection of the method to measure the distance between points, this embodiment does not make special restrictions.

[0064] Among them, the calculation process of the Euclidean distance is a well-known technology, and its specific calculation steps will not be elaborated here.

[0065] (2) Further, based on the density uniformity, the local density, and the average inter-cluster distance, determine the neighborhood adjustment factor of each denture model, specifically:

[0066] The neighborhood adjustment factor of the i-th denture model The expression is: ; where represents the average inter-cluster distance of the i-th denture model; represents the mean value of the local densities of all point cloud data in the i-th denture model; represents the density uniformity of the i-th denture model.

[0067] From the neighborhood adjustment factor of each denture model, it can be understood that when the average inter-cluster distance of the denture model is large, it means that the distance between different clusters in the denture model is far, so the neighborhood radius can be appropriately increased at this time; when the mean value of the local densities of all point cloud data in the denture model is small, it means that the point cloud space is relatively sparse, so a larger neighborhood radius can be set; when the density uniformity is small, it means that the distribution of each point in the point cloud space is more uneven, so a larger neighborhood radius should be set; that is, when the average inter-cluster distance of the denture model is large, the mean value of the local densities of all point cloud data in the denture model is small, and the density uniformity is small, the neighborhood adjustment factor is larger, indicating that the distribution of the point cloud data in the denture model is more uneven at this time, and a larger neighborhood radius should be set;

[0068] On the contrary, when the average inter-cluster distance of the denture model is small, the mean value of the local densities of all point cloud data in the denture model is large, and the density uniformity is large, the neighborhood adjustment factor is smaller, indicating that the distribution of the point cloud data in the denture model is more uniform at this time, and a smaller neighborhood radius can be set.

[0069] (3) Further, take the mean value of the neighborhood adjustment factors of all denture models as the average factor.

[0070] Thus far, according to the distribution characteristics of the point cloud data in the denture model, the average factor has been obtained, which is used to set the denture design parameters for the patient to receive the denture installation.

[0071] S3: Compare the difference between the neighborhood adjustment factor of the oral cavity model and the average factor, and combine the distances between all point cloud data to determine the neighborhood radius of the oral cavity model.

[0072] (1) First, for the oral cavity model, obtain the neighborhood adjustment factor of the oral cavity model according to the construction method of the neighborhood adjustment factor.

[0073] (2) Second, calculate the mean value of the distances between all point cloud data in the oral cavity model, denoted as the average distance, and take a preset multiple of the average distance as the initial neighborhood radius of the oral cavity model;

[0074] It should be noted that the value of the preset multiple is set artificially. To prevent the value of the preset multiple from being set too large, resulting in an overly large neighborhood radius and misdetecting normal points as abnormal points, and at the same time to prevent the value of the preset multiple from being set too small, resulting in an overly small neighborhood radius and missing abnormal points, so the value of the preset multiple is limited to the range of [0.5, 1]. In this embodiment, the value of the preset multiple is 0.5. The implementer can also set it according to the specific situation, and this embodiment does not make special restrictions.

[0075] (3) Further, according to the initial neighborhood radius, and combining the difference between the neighborhood adjustment factor and the average factor, construct the neighborhood radius of the oral cavity model, specifically:

[0076] The neighborhood radius of the oral cavity model The expression is: ; where represents the initial neighborhood radius; represents the neighborhood adjustment factor of the oral cavity model; represents the average factor; tanh( ) represents the hyperbolic tangent function, which is used to adjust the input data to the range of [-1, 1].

[0077] It can be understood from the neighborhood radius that if the neighborhood adjustment factor of the oral cavity model is greater than the average factor, it indicates that the distribution of the point cloud data in the oral cavity model is more uneven, and the neighborhood radius of the oral cavity model should be increased; conversely, when the neighborhood adjustment factor of the oral cavity model is less than the average factor, it indicates that the distribution of the point cloud data in the oral cavity model is more uniform, and the neighborhood radius of the oral cavity model should be appropriately reduced.

[0078] Preferably, the schematic diagram of the neighborhood radius acquisition process provided in this embodiment is as Figure 2 shown.

[0079] S4: Fit all the point cloud data in the oral model to obtain a fitted surface, and respectively evaluate the bending degree of each point cloud data in the oral model and all the point cloud data within its neighborhood radius at the fitted surface, and determine the abnormal weight coefficient of each point cloud data in the oral model.

[0080] Furthermore, it is necessary to determine the value of the abnormal discrimination threshold. If the abnormal discrimination threshold is too large, some normal points located at the edge position will be misjudged as abnormal points due to the small number of neighboring points; if the abnormal discrimination threshold is too small, some abnormal points with a large number of neighboring points will be missed. Therefore, when setting the abnormal discrimination threshold, the abnormal discrimination threshold should be made as small as possible than the number of neighboring points of most normal points and larger than the number of neighboring points of most abnormal points.

[0081] In the point cloud data of the oral cavity of the patient to be installed with dentures, the abnormal points are usually located near the saliva. The point cloud data at the saliva will cause uneven protrusions on the surface of the teeth, resulting in a large curvature of the abnormal points at the saliva. Since the saliva area is usually small, the number of abnormal points generated by the saliva is usually small. The point cloud data of normal teeth is usually relatively flat, the fitted surface will be relatively smooth, and due to the weak reflectivity of normal teeth, there are fewer abnormal points near them.

[0082] Therefore, by fitting all the point cloud data in the oral model to obtain a fitted surface, and respectively evaluating the bending degree of each point cloud data in the oral model and all the point cloud data within its neighborhood radius at the fitted surface, and determining the abnormal weight coefficient of each point cloud data in the oral model, specifically:

[0083] (1) Fit all the point cloud data in the oral model to obtain a fitted surface. There are many common surface fitting methods. In this embodiment, the least squares method is used to obtain the fitted surface formed by all the point cloud data in the oral model. Implementers can also use other fitting methods such as polynomial function fitting. There is no special limitation on the selection of the fitting method in this embodiment. The least squares method is a well-known technology, and the specific process of fitting the point cloud data will not be elaborated here.

[0084] (2) Further, calculate the respective principal curvatures of each point cloud data in the oral model and all the point cloud data within its neighborhood radius at the fitted surface. The principal curvature includes the maximum curvature value and the minimum curvature value. Calculate the sum value of the maximum curvature value and the minimum curvature value of each point cloud data, and record it as the curvature sum value of each point cloud data;

[0085] Calculate the variance of the maximum curvature value among the principal curvatures of all the point cloud data within the neighborhood radius of each point cloud data in the oral model, and record it as the curvature variance of each point cloud data in the oral model;

[0086] Among them, the calculation process of the principal curvature is a well-known technology, and the specific process will not be elaborated here.

[0087] (3) Further, based on the sum of curvatures, the curvature variance, and the number of all point cloud data in the cluster where the point cloud data is located, construct the abnormal weight coefficient of each point cloud data in the oral cavity model, specifically:

[0088] The abnormal weight coefficient of the point cloud data j in the oral cavity model The expression is: ; In the formula, represents the sum of curvatures of the point cloud data j in the oral cavity model; represents the curvature variance of the point cloud data j in the oral cavity model; represents the number of all point cloud data in the cluster where the point cloud data j is located among all the clusters obtained by clustering all the point cloud data in the oral cavity model.

[0089] It can be understood from the abnormal weight coefficients of each point cloud data in the oral cavity model that when the sum of curvatures is relatively large, it indicates that the bending degree of the surface position where the corresponding point cloud data is located is relatively large, and further indicates that the probability of this point being an abnormal point is very high; when the value of the curvature variance is relatively large, it indicates that the distribution of all point cloud data within the neighborhood radius of the corresponding point cloud data is relatively chaotic, and further indicates that this point cloud data may be an abnormal point; when the number of all point cloud data in the cluster where the point cloud data is located is small, it indicates that this point cloud data may be an abnormal point;

[0090] That is, the larger the sum of curvatures, the larger the curvature variance, and the smaller the number of all point cloud data in the cluster where the point cloud data is located, the larger the abnormal weight coefficient, indicating that the probability of the corresponding point cloud data being an abnormal point is greater; on the contrary, if the sum of curvatures is smaller, the curvature variance is smaller, and the number of all point cloud data in the cluster where the point cloud data is located is larger, the smaller the abnormal weight coefficient, indicating that the probability of the corresponding point cloud data being an abnormal point is smaller.

[0091] The denture model generation module 103 is used to determine approximate normal points and approximate abnormal points based on the abnormal weight coefficients of all point cloud data in the oral cavity model, and respectively analyze the distribution of point cloud data within the neighborhood radius of each approximate normal point and each approximate abnormal point, determine the abnormal discrimination value of the oral cavity model, and process the denture for the patient to be fitted with a denture.

[0092] (1) Use the abnormal weight coefficients of all point cloud data in the oral cavity model as the input of the threshold segmentation algorithm, output the segmentation threshold, regard the point cloud data with an abnormal weight coefficient greater than the segmentation threshold as approximate abnormal points, and regard the point cloud data with an abnormal weight coefficient less than or equal to the segmentation threshold as approximate normal points.

[0093] It should be noted that there are many commonly used threshold segmentation algorithms. In this embodiment, the Otsu threshold segmentation algorithm is used to obtain the segmentation threshold of the abnormal weight coefficient. Implementers can also use other threshold segmentation methods. Regarding the selection of the threshold segmentation algorithm, no special restrictions are imposed in this embodiment.

[0094] Among them, the Otsu threshold segmentation algorithm is a well-known technology in the field of threshold segmentation, and its specific principle will not be elaborated here.

[0095] (2) Further, count the total number of point cloud data within the neighborhood radius of all approximately normal points in the oral cavity model, and take the maximum value among the total numbers as the first maximum value of the oral cavity model;

[0096] Count the total number of point cloud data within the neighborhood radius of all approximately abnormal points in the oral cavity model, and take the maximum value among the total numbers as the second maximum value of the oral cavity model;

[0097] The abnormal discrimination value of the oral cavity model is the average of the first maximum value and the second maximum value of the oral cavity model.

[0098] (3) Further, regard the point cloud data with the total number of point cloud data within the neighborhood radius in the oral cavity model less than the abnormal discrimination threshold as abnormal points, and remove them from the oral cavity model. Then, regard all the point cloud data in the oral cavity model after removing the abnormal points as the input of the 3D modeling software, and output the prosthetic model to be processed for the patient to be installed with dentures.

[0099] It should be noted that the above-mentioned sequence of embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0100] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are to illustrate the differences from other embodiments.

[0101] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent acquisition and processing system for personalized denture processing parameters, characterized in that, The system includes: A denture data acquisition module, which is used to obtain the three-dimensional point cloud data in the oral cavity of a large number of patients with dentures installed in their oral cavities, and record it as a denture model, and obtain the three-dimensional point cloud data in the oral cavity of the patient to be installed with dentures, and record it as an oral cavity model; A denture data processing module, which is used to cluster all the point cloud data in each denture model to obtain multiple clustering clusters and the local density of each point cloud data; comprehensively consider the extreme distribution and dispersion degree of the local density of all the point cloud data in each denture model, and combine the total number of clustering clusters to evaluate the density uniformity of each denture model; By measuring the distribution of the distances from each clustering cluster in each denture model to the other clustering clusters, obtain the average inter-cluster distance of each denture model, and combine the average distribution of the local density of all the point cloud data in each denture model and the density uniformity to construct the neighborhood adjustment factor of each denture model, and take the mean value of the neighborhood adjustment factors of all the denture models as the average factor; For the oral cavity model, according to the construction method of the neighborhood adjustment factor, obtain the neighborhood adjustment factor of the oral cavity model; compare the difference between the neighborhood adjustment factor of the oral cavity model and the average factor, and combine the distances between all the point cloud data to determine the neighborhood radius of the oral cavity model; Fit all the point cloud data in the oral cavity model to obtain a fitting surface, and cluster all the point cloud data; and respectively evaluate the bending degree of each point cloud data in the oral cavity model and all the point cloud data within its neighborhood radius at the fitting surface, and combine the number of point cloud data in the clustering cluster where each point cloud data is located to determine the abnormal weight coefficient of each point cloud data in the oral cavity model; A denture model generation module, which is used to determine approximate normal points and approximate abnormal points based on the abnormal weight coefficients of all the point cloud data in the oral cavity model, and respectively analyze the distribution of the point cloud data within the neighborhood radius of each approximate normal point and each approximate abnormal point, determine the abnormal discrimination value of the oral cavity model, and process the dentures of the patient to be installed with dentures; The method for determining the abnormal discrimination value of the oral cavity model is: Count the total number of point cloud data within the neighborhood radius of all approximate normal points in the oral cavity model, and take the maximum value in the total number as the first maximum value of the oral cavity model; Count the total number of point cloud data within the neighborhood radius of all approximate abnormal points in the oral cavity model, and take the maximum value in the total number as the second maximum value of the oral cavity model; The abnormal discrimination value of the oral cavity model is the mean value of the first maximum value and the second maximum value of the oral cavity model.

2. The intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that, The method for evaluating the density uniformity of each denture model is: Calculate the range and variance of the local density of all the point cloud data in each denture model respectively, and take the reciprocal of the product of the number of all clustering clusters, the range and the variance in each denture model as the density uniformity of each denture model.

3. The intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that, The method for obtaining the average inter-cluster distance of each denture model is: In each denture model, calculate the distance from the clustering center of each clustering cluster to the clustering centers of all the other clustering clusters, and record the minimum value of the distances as the clustering distance of each clustering cluster; The average inter-cluster distance of each denture model is the average level of the clustering distances of all the clustering clusters in each denture model.

4. The intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that The expression of the neighborhood adjustment factor for each denture model is as follows: ; where represents the neighborhood adjustment factor of the i-th denture model; represents the average inter-cluster distance of the i-th denture model; represents the mean of the local densities of all the point cloud data in the i-th denture model; represents the density uniformity of the i-th denture model.

5. The intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that, The method for determining the neighborhood radius of the oral cavity model is as follows: Calculate the average value of the distances between all point cloud data in the oral cavity model, denoted as the average distance, and use a preset multiple of the average distance as the initial neighborhood radius of the oral cavity model; Neighborhood radius of the oral cavity model The expression is as follows: ; where represents the initial neighborhood radius; represents the neighborhood adjustment factor of the oral cavity model; represents the average factor; tanh( ) represents the hyperbolic tangent function.

6. The intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that The method for determining the abnormal weight coefficient of each point cloud data in the oral cavity model is as follows: Calculate the respective principal curvatures of each point cloud data in the oral cavity model and all point cloud data within its neighborhood radius at the fitting surface. The principal curvature includes the maximum curvature value and the minimum curvature value. Calculate the sum value of the maximum curvature value and the minimum curvature value of each point cloud data, denoted as the curvature sum value of each point cloud data; Calculate the variance of the maximum curvature values among the principal curvatures of all point cloud data within the neighborhood radius of each point cloud data in the oral cavity model, denoted as the curvature variance of each point cloud data in the oral cavity model; Abnormal weight coefficient of point cloud data j in the oral cavity model The expression is as follows: ; In the formula, represents the sum of curvatures of point cloud data j in the oral cavity model; represents the curvature variance of point cloud data j in the oral cavity model; represents the number of all point cloud data in the cluster where point cloud data j is located among all the clusters obtained by clustering all the point cloud data in the oral cavity model.

7. An intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that The method for determining the approximate normal points and approximate abnormal points is as follows: Use the abnormal weight coefficients of all point cloud data in the oral cavity model as the input of the threshold segmentation algorithm, output the segmentation threshold. The point cloud data with an abnormal weight coefficient greater than the segmentation threshold is used as the approximate abnormal point, and the point cloud data with an abnormal weight coefficient less than or equal to the segmentation threshold is used as the approximate normal point.

8. The intelligent acquisition and processing system for personalized denture processing parameters according to claim 1, characterized in that, The processing of the denture for the patient to be installed with the denture includes: Use the point cloud data with the total number of point cloud data within the neighborhood radius in the oral cavity model less than the abnormal discrimination threshold as the abnormal points, remove them from the oral cavity model, and use all the point cloud data in the oral cavity model after removing the abnormal points as the input of the 3D modeling software to output the to-be-processed denture model of the patient to be installed with the denture.

Images