Orthodontic treatment monitoring method, device and equipment and storage medium

Abstract

The invention relates to an orthodontic treatment monitoring method, device and equipment and a storage medium. The orthodontic treatment monitoring method comprises the following steps: acquiring first tooth jaw models in a preset space, and a symbol distance function of each first tooth model in the first tooth jaw models; acquiring a plurality of intraoral images of the user and an initial color field network of the second dental model; determining initial transformation parameters corresponding to each tooth model pair in the first tooth jaw model and the second tooth jaw model, wherein the tooth model pair comprises a first tooth model and a second tooth model corresponding to the same tooth; according to the symbol distance function, determining an initial camera parameter, an initial transformation parameter and a color loss function corresponding to the initial color field network; and if the initial camera parameter, the initial transformation parameter and the color loss function corresponding to the initial color field network meet a first preset condition, determining the initial transformation parameter as a target transformation parameter. According to the embodiment of the invention, a user can conveniently carry out orthodontic treatment monitoring.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to an orthodontic treatment monitoring method, device, equipment and storage medium. Background Art

[0002] Orthodontics is the process of straightening teeth, correcting malocclusions and deformities, in order to achieve beautiful teeth.

[0003] At present, orthodontic treatment monitoring mainly relies on patients to go to the clinic for regular check-ups, and the doctor evaluates the orthodontic effect of the current treatment stage through naked eye observation, or the doctor uses a mouth scanner to scan the patient's teeth, and then compares the scanned teeth with the expected teeth to evaluate the orthodontic effect of the current treatment stage. However, some patients are unable to go to the clinic for regular check-ups due to busy work, time and space constraints, etc., resulting in the inability to adjust the dental braces in time, which in turn leads to longer dental treatment cycles, poor treatment effects, and poor patient treatment experience. Therefore, there is an urgent need for a more convenient orthodontic treatment monitoring method. Summary of the invention

[0004] In order to solve the above technical problems, the present disclosure provides an orthodontic treatment monitoring method, device, equipment and storage medium.

[0005] In a first aspect, the present disclosure provides an orthodontic treatment monitoring method, the method comprising:

[0006] Acquire a first dental jaw model located in a preset space, and a signed distance function of each first tooth model in the first dental jaw model, wherein the signed distance function is used to characterize a first distance from each first coordinate point in the preset space to the first tooth model;

[0007] Acquire multiple intraoral images of the user and an initial color field network of a second dental model, wherein the second dental model corresponds to the multiple intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model;

[0008] Determine initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, wherein the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth;

[0009] Determine the color loss function corresponding to the initial camera parameters, the initial transformation parameters and the initial color field network according to the signed distance function, wherein the initial camera parameters correspond to the multiple intraoral images, and the color loss function is used to characterize the total color deviation between the first tooth model and the second tooth model in each pair of tooth models after the first dental jaw model is similarly transformed according to the initial transformation parameters;

[0010] If the initial camera parameters, the initial transformation parameters, and the color loss function corresponding to the initial color field network meet the first preset condition, the initial transformation parameters are determined as the target transformation parameters.

[0011] In a second aspect, the present disclosure provides an orthodontic treatment monitoring device, the device comprising:

[0012] A first acquisition module, used for acquiring a first dental jaw model in a preset space and a signed distance function of each first tooth model in the first dental jaw model, wherein the signed distance function is used for representing a first distance from each first coordinate point in the preset space to the first tooth model;

[0013] A second acquisition module is used to acquire a plurality of intraoral images of the user and a color field network of a second dental model, wherein the second dental model corresponds to the plurality of intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model,

[0014] A first determination module is used to determine initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, where the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth;

[0015] a second determination module, for determining, according to the signed distance function, initial camera parameters, initial transformation parameters, and a color loss function corresponding to an initial color field network, wherein the initial camera parameters correspond to a plurality of intraoral images, and the color loss function is used to characterize a total color deviation between a first tooth model and a second tooth model in each pair of tooth models after the first dental jaw model is similarly transformed according to the initial transformation parameters;

[0016] The third determination module is used to determine the initial transformation parameters as target transformation parameters if the color loss function corresponding to the initial camera parameters, the initial transformation parameters, and the initial color field network meets the first preset condition.

[0017] In a third aspect, an embodiment of the present disclosure further provides an electronic device, the device comprising:

[0018] one or more processors;

[0019] a storage device for storing one or more programs,

[0020] When one or more programs are executed by one or more processors, the one or more processors implement the orthodontic treatment monitoring method provided in the first aspect.

[0021] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the orthodontic treatment monitoring method provided in the first aspect.

[0022] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages:

[0023] An orthodontic treatment monitoring method, device, equipment and storage medium of the disclosed embodiments can obtain a first jaw model located in a preset space, and a signed distance function of each first tooth model in the first jaw model, wherein the signed distance function is used to characterize the first distance from each first coordinate point in the preset space to the first tooth model; obtain multiple intra-oral images of the user, and an initial color field network of a second jaw model, wherein the second jaw model corresponds to the multiple intra-oral images and includes the second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second jaw model; determine each pair of teeth in the first jaw model and the second jaw model. The initial transformation parameters corresponding to the model pair, the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth; according to the signed distance function, the initial camera parameters, the initial transformation parameters and the color loss function corresponding to the initial color field network are determined, wherein the initial camera parameters correspond to multiple intraoral images, and the color loss function is used to characterize the total color deviation between the first tooth model and the second tooth model in each pair of tooth models after the first tooth model is similarly transformed according to the initial transformation parameters; if the color loss function corresponding to the initial camera parameters, the initial transformation parameters and the initial color field network meet the first preset condition, the initial transformation parameters are determined as the target transformation parameters. . It can be seen that by adopting the above technical solution, it is not necessary to use an oral scanner for tooth scanning, and only multiple intraoral images are needed to obtain the target transformation parameters corresponding to each pair of tooth models in the second tooth model and the first tooth model (i.e., the tooth model corresponding to the expected orthodontic effect) corresponding to the multiple intraoral images, and the doctor can know the orthodontic treatment effect according to the target transformation parameters, so that the user can monitor the orthodontic treatment more conveniently. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0026] Figure 1 A schematic diagram of a flow chart of an orthodontic treatment monitoring method provided by an embodiment of the present disclosure is shown;

[0027] Figure 2 A schematic diagram showing a flow chart of another orthodontic treatment monitoring method provided by an embodiment of the present disclosure;

[0028] Figure 3 A schematic diagram of the structure of a neural network provided by an embodiment of the present disclosure is shown;

[0029] Figure 4 A schematic structural diagram of an orthodontic treatment monitoring device provided by an embodiment of the present disclosure is shown;

[0030] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0031] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0032] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0033] In order to solve the above problems, the embodiments of the present disclosure provide an orthodontic treatment monitoring method, device, equipment and storage medium.

[0034] Combine the following Figures 1 to 3 The orthodontic treatment monitoring method provided by the embodiment of the present disclosure is described. In the embodiment of the present disclosure, the orthodontic treatment monitoring method can be executed by an electronic device. The electronic device can include a device with communication function such as a tablet computer, a desktop computer, a laptop computer, etc., and can also include a device simulated by a virtual machine or a simulator.

[0035] Figure 1 A schematic flow chart of an orthodontic treatment monitoring method provided by an embodiment of the present disclosure is shown.

[0036] like Figure 1 As shown, the orthodontic treatment monitoring method may include the following steps.

[0037] S110, obtaining a first jaw model in a preset space and a signed distance function of each first tooth model in the first jaw model, wherein the signed distance function is used to characterize a first distance from each first coordinate point in the preset space to the first tooth model.

[0038] Specifically, the preset space may include a sphere with a radius of a preset radius, and the preset radius may be 1, but is not limited thereto.

[0039] Specifically, the first dental model is a dental model corresponding to a desired dental orthodontic effect, and the first dental model includes a first tooth model corresponding to at least one tooth.

[0040] The first dental model may include a dental model corresponding to a first-stage orthodontic effect, a dental model corresponding to a second-stage orthodontic effect, or a dental model corresponding to a final orthodontic effect, but is not limited thereto.

[0041] The specific representation forms of the first jaw model and the first tooth model may include a three-dimensional point cloud model, a three-dimensional mesh model, etc., which are not limited here.

[0042] Exemplarily, obtaining the first dental model may include: reading the first dental model from a storage device; or: performing mesh segmentation on the models corresponding to the teeth and gums to segment the tooth models corresponding to each tooth, and using DeepSDF technology to complete the segmented tooth models into a watertight mesh, thereby obtaining the first dental model. However, the present invention is not limited thereto.

[0043] Specifically, the signed distance function is used to describe the geometric shape of the corresponding tooth model. Therefore, the first tooth model can be displayed according to the signed distance function of each first tooth model in the first tooth model.

[0044] Exemplarily, obtaining the signed distance function of each first tooth model in the first dental jaw model may include: reading the signed distance function of each first tooth model in the first dental jaw model from a storage device, but is not limited thereto.

[0045] S120. Acquire multiple intraoral images of the user and an initial color field network of the second dental model, wherein the second dental model corresponds to the multiple intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model.

[0046] Specifically, the intraoral image is an image obtained by capturing the inside of the oral cavity using an imaging device. The image capturing device may include a mobile phone, a computer, a camera, etc., but is not limited thereto.

[0047] Exemplarily, acquiring multiple intraoral images of the user may include: receiving multiple intraoral images of the user sent by other electronic devices, or reading multiple intraoral images of the user from a storage device, etc., but is not limited thereto.

[0048] Specifically, the second dental model is a dental model corresponding to (or behind) multiple intraoral images, and the second dental model includes at least one second tooth model. It should be understood by those skilled in the art that the first dental model is a dental model corresponding to the desired dental orthodontic effect, and the second dental model is a dental model corresponding to multiple intraoral images. Since the first dental model and the second dental model both correspond to the user's teeth, the first dental model can be transformed into a second dental model by optimizing the posture of each first tooth model in the first dental model (that is, by performing a similarity transformation on each first tooth model), and then the difference between each first tooth model and the first tooth model corresponding thereto is obtained. Based on this, the embodiment of the present disclosure needs to determine the target transformation parameters for transforming the first dental model into the second dental model.

[0049] Specifically, the color field network is used to describe the rendering color of the dental model.

[0050] Specifically, the initial color field network is the initial color field network of the second dental model, and the network parameters in the initial color field network can be obtained by random settings, but are not limited to this. In the embodiment of the present disclosure, in the process of optimizing the posture of each first tooth model in the first dental model so that the first dental model becomes the second dental model, the real color field network (i.e., the target color field network) of the second dental model will also be slowly determined, for example, by optimizing the initial color field network so that it slowly approaches the target color field network.

[0051] S130, determining initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, where the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth.

[0052] Specifically, for each pair of tooth models, the geometry of the first tooth model and the second tooth model included therein are consistent, and the difference between the two is a similarity transformation. In other words, the first tooth model and the second tooth model can be aligned by performing a similarity transformation on the first tooth model. Among them, the transformation parameters based on the similarity transformation include a scaling ratio (scale), a translation matrix (rotation), and a rotation matrix (translation). Since the scaling ratio (scale) between each pair of tooth models is the scaling ratio between the world coordinate systems where the first dental model and the second dental model are located, the scaling ratios of each pair of tooth models are the same. In this way, each pair of tooth models in the first dental model and the second dental model includes a common scaling ratio (scale), and their corresponding translation matrices (rotation), and rotation matrices (translation).

[0053] Specifically, the initial transformation parameters are initial transformation parameters corresponding to the tooth model pair.

[0054] In some embodiments, S120 may include: randomly generating initial transformation parameters corresponding to each pair of tooth models.

[0055] In other embodiments, S120 may include: performing three-dimensional reconstruction based on multiple intraoral images to obtain a third dental model; transforming the third dental model to a preset space through a similarity transformation; and performing coarse alignment processing on the first dental model and the transformed third dental model to obtain initial transformation parameters corresponding to each pair of tooth models.

[0056] Specifically, the third dental jaw model is a dental jaw model corresponding to the dental jaw of the user obtained by three-dimensional reconstruction based on a plurality of intraoral images.

[0057] The specific representation form of the third dental model may include a three-dimensional point cloud model, a three-dimensional mesh model, etc., which is not limited here.

[0058] Specifically, the initial camera parameters are rough camera parameters corresponding to the acquisition of multiple intraoral images. In the disclosed embodiment, in the process of optimizing the posture of each first tooth model in the first jaw model so that the first jaw model becomes the second jaw model, the real camera parameters (i.e., target camera parameters) corresponding to the acquisition of multiple intraoral images will also be slowly determined, for example, by optimizing the initial camera parameters to gradually approach the target camera parameters.

[0059] Among them, the camera parameters may include camera intrinsic parameters and camera extrinsic parameters. The camera intrinsic parameters may include the camera focal length, the projection coordinates of the camera center on the pixel plane, etc. The camera extrinsic parameters may include the camera pose. The camera pose may include, for example, a camera rotation matrix and a camera translation matrix. The camera rotation matrix and the camera translation matrix together describe how to transform a point from the world coordinate system to the camera coordinate system.

[0060] For example, the Colmap algorithm is used to estimate the camera parameters (i.e., initial camera parameters) and sparse point cloud (i.e., the third dental model) when acquiring multiple intraoral images. If the image acquisition device is a pinhole camera, the camera parameters include three camera intrinsic parameters, i.e., the camera focal length f, and the projection coordinates of the camera center on the pixel plane (c u ,c v ), each intraoral image corresponds to 6 camera extrinsics (corresponding to 6 degrees of freedom).

[0061] Specifically, the first jaw model and the third jaw model are roughly registered to obtain transformation parameters for overall alignment of the first jaw model and the third jaw model, and the transformation parameters are used as initial transformation parameters corresponding to each tooth model pair.

[0062] Exemplarily, the first jaw model and the third jaw model may be coarsely registered using a feature-based coarse registration method, a coarse registration method based on a random sample consensus (RANSAC) framework, etc., but the present invention is not limited thereto.

[0063] It can be understood that since both the second jaw model and the third jaw model correspond to multiple intraoral images, when the first jaw model and the third jaw model after similarity transformation are roughly aligned as a whole, the first jaw model and the second jaw model are also roughly aligned as a whole, that is, the first tooth model and the second tooth model in the tooth model pair are at least roughly aligned. Therefore, the initial transformation parameters obtained by the rough registration process are closer to the target transformation parameters, which is conducive to obtaining the target transformation parameters corresponding to the tooth model pair as soon as possible and improving the orthodontic treatment monitoring efficiency. In addition, by performing a rough registration process on the first jaw model and the third jaw model after similarity transformation as a whole to obtain the initial transformation parameters, the initial transformation parameters corresponding to each pair of tooth models can be quickly obtained.

[0064] In some other embodiments, S120 may include: performing three-dimensional reconstruction based on multiple intraoral images to obtain a third dental jaw model, wherein the third dental jaw model includes at least one third tooth model; transforming the third dental jaw model to a preset space through a similarity transformation; performing coarse alignment processing on each pair of tooth models in the first dental jaw model and the third dental jaw model after the similarity transformation to obtain initial transformation parameters corresponding to each pair of tooth models in the first dental jaw model and the second dental jaw model.

[0065] S140. Determine the color loss function corresponding to the initial camera parameters, the initial transformation parameters and the initial color field network according to the signed distance function, wherein the initial camera parameters correspond to a plurality of intraoral images, and the color loss function is used to characterize the total color deviation between the first tooth model and the second tooth model in each pair of tooth models after the first dental model is similarly transformed according to the initial transformation parameters.

[0066] Specifically, the total color deviation between the first tooth model and the second tooth model in the tooth model pair may include a deviation in color value.

[0067] S150: If the initial camera parameters, the initial transformation parameters, and the color loss function corresponding to the initial color field network meet a first preset condition, the initial transformation parameters are determined as target transformation parameters.

[0068] Specifically, the first preset condition may include: the color loss function is less than a first preset threshold, wherein the specific value of the first preset threshold may be set by those skilled in the art according to actual conditions, but is not limited thereto.

[0069] Specifically, if the color loss function corresponding to the initial camera parameters, initial transformation parameters and initial color field network meets the first preset condition, it indicates that after the first dental model is similarly transformed according to the initial transformation parameters, that is, after the first dental model in each pair of dental models is similarly transformed according to its corresponding initial transformation parameters, the alignment accuracy with the second dental model is high (or the requirements have been met), and the difference between the initial camera parameters and the actual camera parameters (that is, the actual camera parameters corresponding to the image acquisition of multiple intraoral images) is small (or the requirements have been met), and the difference between the color of the second dental model predicted by the initial color field network and the real color field network of the second dental model is small (or the requirements have been met), at this time, the initial transformation parameters can be determined as the target transformation parameters. Of course, the initial camera parameters can also be determined as the target camera parameters, and the initial color field network can also be determined as the target color field network.

[0070] Those skilled in the art should understand that, since the first tooth model in the first jaw model is aligned with the second jaw model after a similarity transformation according to its corresponding target transformation parameters, in other words, the jaw model obtained after a similarity transformation of the first tooth model in the first jaw model according to its corresponding target transformation parameters can be regarded as the second jaw model. Therefore, the tooth position difference between the first tooth model and the second tooth model can be determined according to the target transformation parameters to achieve orthodontic treatment monitoring.

[0071] It is understandable that the technical solution provided by the embodiment of the present disclosure eliminates the need for users to go to the clinic in person to have the doctor observe with the naked eye or use an oral scanner to scan their teeth. Instead, they only need to take multi-view images of their own teeth (i.e., multiple intraoral images) and then upload them. In this way, users do not need to frequently go to the clinic for follow-up visits, which saves users’ time and frees them from time and space constraints. In addition, since doctors do not need to judge with the naked eye or operate an oral scanner to frequently scan users’ teeth, doctors’ time can also be saved. Moreover, compared to doctors’ visual judgment, the technical solution provided by the embodiment of the present disclosure can quantitatively obtain the target transformation parameters corresponding to each pair of tooth models, avoiding the problem of misjudgment caused by doctors’ lack of experience and other reasons, and improving the accuracy of orthodontic treatment monitoring.

[0072] According to the disclosed embodiment, there is no need to use an oral scanner for teeth scanning. Only a plurality of intraoral images are needed to obtain the target transformation parameters corresponding to each pair of tooth models in the second jaw model and the first jaw model (i.e., the jaw model corresponding to the expected orthodontic effect) corresponding to the plurality of intraoral images. The doctor can obtain the orthodontic treatment effect based on the target transformation parameters, thereby enabling users to monitor orthodontic treatment more conveniently.

[0073] In another embodiment of the present disclosure, the method further comprises:

[0074] S160: If the color loss function corresponding to the initial camera parameters, the initial transformation parameters, and the initial color field network does not meet the first preset condition, the initial camera parameters, the initial transformation parameters, and the initial color field network are optimized until the corresponding color loss function meets the first preset condition.

[0075] S170: Determine the optimized initial transformation parameters as target transformation parameters.

[0076] Specifically, if the color loss function corresponding to the initial camera parameters, initial transformation parameters and initial color field network does not meet the first preset condition, it indicates that after the first tooth model in at least one pair of tooth models is similarly transformed according to its corresponding initial transformation parameters, the alignment accuracy with the second tooth model is low (or does not meet the requirements), and the gap between the initial camera parameters and the actual camera parameters is large (or does not meet the requirements), and the gap between the color of the second dental model predicted by the initial color field network and the actual color field network of the second dental model is large (or does not meet the requirements). At this time, the initial camera parameters, initial transformation parameters and initial color field network can be optimized, and then the color loss function corresponding to the optimized initial camera parameters, optimized initial transformation parameters and optimized initial color field network can be determined. If the color loss function corresponding to the optimized initial camera parameters, the optimized initial transformation parameters, and the optimized initial color field network meets the first preset condition, the optimized initial transformation parameters are determined as the target transformation parameters. If the color loss function corresponding to the optimized initial camera parameters, the optimized initial transformation parameters, and the optimized initial color field network does not meet the first preset condition, the initial camera parameters, the initial transformation parameters, and the initial color field network are continuously optimized until the corresponding color loss function meets the first preset condition, and the final optimized initial transformation parameters are determined as the target transformation parameters. Of course, the final optimized initial camera parameters can also be determined as the target camera parameters, and the final optimized initial color field network can also be determined as the target camera parameters.

[0077] It can be understood that by continuously optimizing the initial camera parameters, initial transformation parameters and initial color field network, the alignment accuracy of the first tooth model and the second tooth model can be made higher and higher after the tooth model pair is similarly transformed according to the corresponding optimized initial transformation parameters, thereby obtaining the target transformation parameters corresponding to each pair of tooth models. In this way, the method for obtaining the target transformation parameters can be simple and convenient.

[0078] Figure 2 The flowchart of another orthodontic treatment monitoring method provided by the embodiment of the present disclosure is shown. The embodiment of the present disclosure is optimized on the basis of the above embodiment, and the embodiment of the present disclosure can be combined with various optional solutions in one or more of the above embodiments.

[0079] like Figure 2 As shown, the orthodontic treatment monitoring method may include the following steps.

[0080] S210, obtaining a first jaw model in a preset space, and a signed distance function of each first tooth model in the first jaw model, wherein the signed distance function is used to characterize a first distance from each first coordinate point in the preset space to the first tooth model.

[0081] Specifically, there are many specific ways to determine the signed distance function of each first tooth model in the first dental jaw model. Typical examples are described below, but they do not constitute a limitation to the present disclosure.

[0082] In some embodiments, the specific process of determining the signed distance function of each first tooth model in the first dental jaw model includes:

[0083] Sampling a first coordinate point in the preset space to obtain a plurality of sampling coordinate points;

[0084] For each sampling coordinate point, the coordinate of the sampling coordinate point is input into the first encoder to obtain the corresponding first encoding vector, the first encoding vector is input into the backbone network to obtain the corresponding feature vector, the feature vector is input into the initial shallow network corresponding to each first tooth model, and the predicted distance from the sampling coordinate point to each first tooth model is obtained;

[0085] According to the predicted distance and the actual distance corresponding to the sampling coordinate points, the distance loss function corresponding to the initial shallow network is determined;

[0086] If the distance loss function corresponding to the initial shallow network satisfies the second preset condition, determining the signed distance function of the first tooth model according to the initial shallow network corresponding to the first tooth model;

[0087] If the distance loss function corresponding to the initial shallow network does not meet the second preset condition, the initial shallow network is optimized until the distance loss function corresponding to the optimized initial shallow network meets the second preset condition, and the signed distance function of the first tooth model is determined based on the optimized initial shallow network corresponding to the first tooth model.

[0088] Specifically, the first coordinate point is any coordinate point in the preset space. Any sampling method known to those skilled in the art can be used to obtain the sampling coordinate point.

[0089] Specifically, the initial shallow network is a randomly initialized shallow network used to characterize the signed distance function of the first tooth model.

[0090] Specifically, the specific structures of the first encoder (ie, the first encoder), the backbone network, the initial shallow network, and the second encoder (ie, the second encoder) can be set by those skilled in the art according to actual conditions and are not limited here.

[0091] Specifically, the predicted distance is the distance from the sampling coordinate point predicted by the initial shallow network to the first tooth model.

[0092] Specifically, the true distance is the actual distance from the sampling coordinate point to the first tooth model.

[0093] Specifically, the distance loss function is used to characterize the deviation between the predicted distance and the actual distance from the sampling coordinate point to each of the first tooth models, that is, the deviation between the initial shallow network and the shallow network of the signed distance function actually used to characterize the first tooth model.

[0094] Specifically, there are many ways to determine the distance loss function, which are described below with reference to typical examples, but are not limited to these.

[0095] In one example, the process of determining the distance loss function includes: for each sampling coordinate point, calculating the distance difference between each pair of predicted distances and true distances corresponding to the sampling coordinate point, and obtaining multiple distance differences corresponding to the sampling coordinate point, wherein each pair of predicted distances and true distances is the predicted distance and true distance from the sampling coordinate point to the same first tooth model; determining the distance loss function according to the multiple distance differences corresponding to each sampling coordinate point, for example, adding the squares (or absolute values) of the multiple distance differences corresponding to each sampling coordinate point, etc., to obtain the distance loss function.

[0096] Specifically, the second preset condition may include: the distance loss function is less than a second preset threshold, wherein the specific value of the second preset threshold can be set by those skilled in the art according to actual conditions, but is not limited thereto.

[0097] Specifically, if the distance loss function corresponding to the initial shallow network satisfies the second preset condition, indicating that the deviation between the predicted distance and the actual distance from the sampling coordinate point to each of the first dental models is small (or has met the requirements), at this time, the initial shallow network can be regarded as a real shallow network used to characterize the signed distance function of the first dental model. Therefore, the signed distance function of the first dental model can be determined based on the initial shallow network corresponding to the first dental model.

[0098] Specifically, if the distance loss function corresponding to the initial shallow network does not meet the second preset condition, it indicates that the deviation between the predicted distance and the actual distance from the sampling coordinate point to each of the first tooth models is large (or does not meet the requirements). At this time, the parameters in the initial shallow network can be optimized, and then the distance loss function corresponding to the optimized initial shallow network is determined. If the distance loss function corresponding to the optimized initial shallow network meets the second preset condition, the signed distance function of the first tooth model is determined according to the optimized initial shallow network corresponding to the first tooth model. If the distance loss function corresponding to the optimized initial shallow network still does not meet the second preset condition, the parameters in the initial shallow network are continuously optimized until the corresponding distance loss function meets the second preset condition, and the signed distance function of the first tooth model is determined according to the final optimized initial shallow network.

[0099] For example, Figure 3 FIG. 1 shows a schematic diagram of a neural network structure provided by an embodiment of the present disclosure. Figure 3 , the neural network starts from the sampling coordinate point x, first passes through the first position encoder 310 to obtain the first encoding vector, and then enters the backbone network 320 to obtain the feature vector. There are K branches starting from the feature vector, where K is the number of first tooth models in the first jaw model. The Sign Distance Function (SDF) of each first tooth model is represented by a shallow network 330, which are the first shallow network, the second shallow network, ..., and the Kth shallow network (i.e., the K initial shallow networks corresponding to the K first tooth models). The feature vectors output by the backbone network 320 are respectively input into the K shallow networks 330 to obtain the distance from the sampling coordinate point x to each first tooth model (i.e., the predicted distance). When the DeepSDF method is used to complete the training of the above neural network, the first shallow network, the second shallow network..., and the Kth shallow network can be trained. According to the trained first shallow network, the second shallow network..., and the Kth shallow network, the SDF of each first tooth model can be determined.

[0100] It can be understood that, compared to each initial shallow network corresponding to a first encoder and a backbone network, setting multiple initial shallow networks to share the same first encoder and the same backbone network is beneficial to reducing the size of the neural network, reducing the number of parameters in the neural network, and thus improving the training speed.

[0101] S220. Acquire multiple intraoral images of the user and an initial color field network of the second dental model, wherein the second dental model corresponds to the multiple intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model.

[0102] Specifically, S220 is similar to S120 and will not be described in detail here.

[0103] S230, determining initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, wherein the second jaw model corresponds to a plurality of intra-oral images and includes a second tooth model, and the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth.

[0104] Specifically, S230 is similar to S130 and will not be described in detail here.

[0105] S240. For each pixel point in the multiple intraoral images, determine the second coordinate point and the ray corresponding to the pixel point in the world coordinate system according to the initial camera parameters, detect whether the ray passes through a first tooth model in the first dental model according to the signed distance function, and obtain a detection result; based on the detection result, determine the second distance from the second coordinate point corresponding to the pixel point to the first dental model after similarity transformation according to the initial transformation parameters, and determine the first color value of the pixel point based on the second distance and the initial color field network.

[0106] Specifically, the second coordinate point is a coordinate point of a pixel point in a world coordinate system determined according to initial camera parameters and normalized to a coordinate point in a preset space through a similarity transformation.

[0107] Specifically, the ray mentioned here is a ray with the second coordinate point as its endpoint.

[0108] Specifically, the first color value is a color value of a pixel predicted according to an initial color field network of the first dental model, a signed distance function of the first tooth model in the first dental model, initial camera parameters, and initial transformation parameters. The color value may include a grayscale value or an RGB value, but is not limited thereto.

[0109] Optionally, detecting whether the ray passes through a first dental model in the first dental model to obtain a detection result includes: collecting multiple sampling points on the ray; for each first dental model, determining multiple first distances from the multiple sampling points to the first dental model according to the signed distance function of the first dental model, and determining whether the ray passes through the first dental model according to the multiple first distances.

[0110] Specifically, multiple sampling points may be randomly collected on the ray, or multiple sampling points may be uniformly collected, and there is no limitation to this.

[0111] Specifically, for each first dental model, the signed distance function of the first dental model is queried according to the coordinates of the multiple sampling points collected on the ray, so as to determine the distances (i.e., the first distances) from the multiple sampling points to the first dental model, and obtain multiple first distances. Whether the ray passes through the first dental model is determined based on the multiple first distances. For example, if the change trend of the multiple first distances corresponding to the multiple sampling points arranged in sequence along the extension direction of the ray is: from a positive number to a negative number, and then to a positive number, it is determined that the ray passes through the first dental model, otherwise it is determined that the ray does not pass through the first dental model.

[0112] Optionally, based on the detection result, determine the second distance from the second coordinate point corresponding to the pixel point to the first dental model after similarity transformation according to the initial transformation parameters, including: if the detection result is that it passes through a first dental model in the first dental model, then the first distance from the second coordinate point corresponding to the pixel point to the first dental model passed by the ray corresponding to the pixel point is taken as the second distance; if the detection result is that it does not pass through the first dental model, then the minimum value of the first distances from the second coordinate point corresponding to the pixel point to each first dental model is taken as the second distance.

[0113] Specifically, according to the second distance and the initial color field network, a neural implicit reconstruction method can be used to determine the first color value of the pixel.

[0114] S250, determining a color loss function according to the first color value and the second color value of each pixel point in the plurality of intra-oral images.

[0115] Specifically, the second color value is the real color value of the pixel in the intraoral image to which it belongs.

[0116] In some embodiments, S250 includes: determining, for each pixel in the multiple intraoral images, a color difference between a first color value and a second color value of the pixel; and determining color loss based on the color difference of each pixel in the multiple intraoral images.

[0117] Exemplarily, determining the color loss based on the color difference of each pixel in the plurality of intraoral images may include: adding the squares (or absolute values, etc.) of the color difference of each pixel to obtain the color loss, but is not limited thereto.

[0118] To explain in detail the process of determining the color loss function provided by the embodiment of the present disclosure, a detailed description will be given below based on a specific example.

[0119] N semantically segmented intraoral images are denoted as Among them, I iis the i-th intraoral image after semantic segmentation, the pixel coordinates of a pixel in the intraoral image after semantic segmentation are marked (u, v), and the corresponding RGB value (i.e., the second color value) is recorded as C (u,v) Since each intraoral image is taken by the same image acquisition device, the camera internal parameters corresponding to the N intraoral images are the same, denoted as (f, c u ,c v ). Each intraoral image has an independent camera pose, and the camera extrinsic parameter of the i-th intraoral image is denoted by p i ∈R 6 (6 degrees of freedom), the overall camera parameters are recorded as Φ = (f, c u ,c v ,p 1 ,p 2 ,...,p N )∈R 6N+3 .

[0120] The first dental model includes K first tooth models, and the K first tooth models have a common scaling ratio s and independent rigid body transformation parameters (i.e., translation matrix and rotation matrix) o i ∈R 6 (6 degrees of freedom), the overall note is Ψ=(s,o 1 ,o 2 ,...,o K )∈R 6K+1 .

[0121] The initial color field network of the first dental model is denoted as f θ (x, ν), which is a function of the spatial three-dimensional coordinate x and direction ν, outputs the RGB value of volume rendering, and θ is the network parameter.

[0122] The process of calculating the color loss function corresponding to the initial camera parameters and the initial transformation parameters is as follows: perform the following operation for each pixel point in the N intraoral images, and calculate the second coordinate point and ray of the pixel point in the world coordinate system through the initial camera parameters corresponding to the intraoral image to which it belongs, collect multiple sampling points on the ray, and for each first dental model, determine multiple first distances from the multiple sampling points to the first dental model according to the signed distance function of the first dental model. If the multiple first distances change from a positive number to a negative number and then to a positive number, it is determined that the ray passes through the first dental model, otherwise it is determined that the ray does not pass through the first dental model. If the ray passes through the i-th first dental model, take the first distance f from the second coordinate point to the i-th first dental model. i is f after fusion (i.e., the second distance); if it does not pass through any first tooth model, f=min i＝1,2,...,K f i, that is, the minimum value of the first distance from the second coordinate point to each first tooth model is the fused f (i.e., the second distance). In this way, the ray corresponding to each pixel corresponds to a first tooth model. According to the neural implicit 3D reconstruction method, the color rendered by each pixel (i.e., the first color value) can be obtained. Since the first color value is related to the camera parameters, transformation parameters, and network parameters of the initial color field network, the first color value is recorded as

[0123] Compare the first color value and the second color value of each pixel in N intra-mouth images, and the color loss is:

[0124]

[0125] Among them, L color (θ, Φ, Ψ) is the color loss, is the first color value of the pixel with pixel coordinates (u, v), C (u,v) The second color value of the pixel with pixel coordinates (u, v).

[0126] S260: If the color loss function corresponding to the initial camera parameters and the initial transformation parameters meets a first preset condition, the initial transformation parameters are determined as target transformation parameters.

[0127] Specifically, S260 is similar to S150 and will not be described in detail here.

[0128] Those skilled in the art should understand that “determining the color loss function corresponding to the optimized initial camera parameters, the optimized initial transformation parameters, and the optimized initial color field network” is similar to S240-S250, except that the optimized initial camera parameters, the optimized initial transformation parameters, and the optimized initial color field network are involved in the calculation of the color loss function, which will not be repeated here.

[0129] The disclosed embodiments utilize neural implicit 3D reconstruction to jointly optimize camera parameters and transformation parameters, and can simply and quickly extract target transformation parameters from multiple intraoral images taken by the user and the first dental model, thereby obtaining tooth posture differences that accurately meet orthodontic requirements.

[0130] Figure 4 A schematic structural diagram of an orthodontic treatment monitoring device provided in an embodiment of the present disclosure is shown.

[0131] like Figure 4 As shown, the orthodontic treatment monitoring device 400 may include:

[0132] A first acquisition module 410 is used to acquire a first dental jaw model located in a preset space, and a signed distance function of each first tooth model in the first dental jaw model, wherein the signed distance function is used to represent a first distance from each first coordinate point in the preset space to the first tooth model;

[0133] A second acquisition module 420 is used to acquire a plurality of intraoral images of the user and an initial color field network of a second dental model, wherein the second dental model corresponds to the plurality of intraoral images and includes a second tooth model, and the color field network is used to characterize a color value corresponding to each pair of second coordinate points and ray directions in the second dental model;

[0134] A first determination module 430 is used to determine initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, where the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth;

[0135] A second determination module 440 is used to determine the color loss function corresponding to the initial camera parameters, the initial transformation parameters and the initial color field network according to the signed distance function, wherein the initial camera parameters correspond to the multiple intraoral images, and the color loss function is used to characterize the total color deviation between the first tooth model and the second tooth model in each pair of tooth models after the first dental jaw model is similarly transformed according to the initial transformation parameters;

[0136] The third determination module 450 is configured to determine the initial transformation parameters as target transformation parameters if the color loss function corresponding to the initial camera parameters, the initial transformation parameters, and the initial color field network satisfies a first preset condition.

[0137] In another embodiment of the present disclosure, the device further includes:

[0138] An optimization module, configured to optimize the initial camera parameters, the initial transformation parameters, and the initial color field network until the corresponding color loss function satisfies the first preset condition if the color loss function corresponding to the initial camera parameters and the initial transformation parameters does not satisfy the first preset condition;

[0139] The fourth determining module is used to determine the optimized initial transformation parameters as target transformation parameters.

[0140] In yet another embodiment of the present disclosure, the first determining module 430 includes:

[0141] A reconstruction submodule, used for performing three-dimensional reconstruction according to the multiple intraoral images to obtain a third dental model;

[0142] A transformation submodule, used for transforming the third dental jaw model into a preset space through similarity transformation;

[0143] The coarse registration submodule is used to perform coarse registration processing on the first dental jaw model and the transformed third dental jaw model to obtain initial transformation parameters corresponding to each pair of tooth models.

[0144] In another embodiment of the present disclosure, the device further comprises:

[0145] A sampling module, used for sampling a first coordinate point in a preset space to obtain a plurality of sampling coordinate points;

[0146] A prediction module, for each sampling coordinate point, inputting the coordinate of the sampling coordinate point into a first encoder to obtain a corresponding first encoding vector, inputting the first encoding vector into a backbone network to obtain a corresponding feature vector, and inputting the feature vector into an initial shallow network corresponding to each first tooth model to obtain a predicted distance from the sampling coordinate point to each first tooth model;

[0147] A fifth determination module is used to determine the distance loss function corresponding to the initial shallow network according to the predicted distance and the actual distance corresponding to the sampling coordinate point;

[0148] a sixth determination module, configured to determine the signed distance function of the first tooth model according to the initial shallow network corresponding to the first tooth model if the distance loss function corresponding to the initial shallow network satisfies the second preset condition;

[0149] The seventh determination module is used to optimize the initial shallow network if the distance loss function corresponding to the initial shallow network does not meet the second preset condition, until the distance loss function corresponding to the optimized initial shallow network meets the second preset condition, and determine the signed distance function of the first tooth model according to the optimized initial shallow network corresponding to the first tooth model.

[0150] In yet another embodiment of the present disclosure, the second determining module 440 includes:

[0151] A first determination submodule is used to determine, for each pixel point in the multiple intraoral images, a second coordinate point in the world coordinate system and a ray corresponding to the pixel point according to the initial camera parameters, detect whether the ray passes through a first tooth model in the first dental model according to the signed distance function, and obtain a detection result; based on the detection result, determine a second distance from the second coordinate point corresponding to the pixel point to the first dental model after similarity transformation according to the initial transformation parameters, and determine a first color value of the pixel point based on the second distance and the color field network;

[0152] The second determination submodule is used to determine the color loss function according to the first color value and the second color value of each pixel point in the multiple intra-oral images.

[0153] In another embodiment of the present disclosure, the second determining submodule includes a detection unit for detecting whether the ray passes through a first jaw model in the first jaw model to obtain a detection result, and the first detection unit is specifically configured to:

[0154] Collect multiple sampling points on the ray;

[0155] For each first tooth model, a plurality of first distances from a plurality of sampling points to the first tooth model are determined according to the signed distance function of the first tooth model, and whether the ray passes through the first tooth model is determined according to the plurality of first distances.

[0156] In another embodiment of the present disclosure, the second determination submodule includes a detection unit for determining, based on the detection result, a second distance from a second coordinate point corresponding to the pixel point to the first dental model after similarity transformation according to the initial transformation parameters, and the detection unit is specifically used to:

[0157] If the detection result is a first tooth model passing through the first dental model, a first distance from the second coordinate point corresponding to the pixel point to the first tooth model passed by the ray corresponding to the pixel point is taken as the second distance;

[0158] If the detection result is that it does not pass through the first dental model, the minimum value of the first distances from the second coordinate point corresponding to the pixel point to each first dental model is used as the second distance.

[0159] It should be noted that Figure 4 The orthodontic treatment monitoring device 400 shown can perform Figure 1 to Figure 2 The various steps in the method embodiment shown in the figure are implemented Figure 1 to Figure 2 The various processes and effects in the method embodiment shown are not described in detail here.

[0160] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present disclosure is shown.

[0161] like Figure 5 As shown, the electronic device may include a processor 501 and a memory 502 storing computer program instructions.

[0162] Specifically, the processor 501 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0163] The memory 502 may include a large capacity memory for information or instructions. By way of example and not limitation, the memory 502 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a tape, or a universal serial bus (USB) drive or a combination of two or more of these. Where appropriate, the memory 502 may include a removable or non-removable (or fixed) medium. Where appropriate, the memory 502 may be inside or outside the integrated gateway device. In a particular embodiment, the memory 502 is a non-volatile solid-state memory. In a particular embodiment, the memory 502 includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (Electrically Erasable Programmable ROM, EEPROM), an electrically rewritable ROM (EAROM) or a flash memory, or a combination of two or more of these.

[0164] The processor 501 reads and executes the computer program instructions stored in the memory 502 to perform the steps of the orthodontic treatment monitoring method provided in the embodiment of the present disclosure.

[0165] In one example, the electronic device may further include a transceiver 503 and a bus 504. Figure 5 As shown, the processor 501, the memory 502 and the transceiver 503 are connected via a bus 504 and communicate with each other.

[0166] The bus 504 includes hardware, software, or both. For example, but not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, the bus 504 may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

[0167] The following is an embodiment of a computer-readable storage medium provided in an embodiment of the present disclosure. The computer-readable storage medium and the orthodontic treatment monitoring method of the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the computer-readable storage medium, reference can be made to the above-mentioned embodiment of the orthodontic treatment monitoring method.

[0168] This embodiment provides a storage medium containing computer executable instructions, which, when executed by a computer processor, are used to perform an orthodontic treatment monitoring method, the method comprising:

[0169] Acquire a first dental jaw model located in a preset space, and a signed distance function of each first tooth model in the first dental jaw model, wherein the signed distance function is used to characterize a first distance from each first coordinate point in the preset space to the first tooth model;

[0170] Acquire multiple intraoral images of the user and an initial color field network of a second dental model, wherein the second dental model corresponds to the multiple intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model;

[0171] Determine initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, wherein the tooth model pair includes a first tooth model and a second tooth model corresponding to the same tooth;

[0172] Determine the color loss function corresponding to the initial camera parameters, the initial transformation parameters and the initial color field network according to the signed distance function, wherein the initial camera parameters correspond to the multiple intraoral images, and the color loss function is used to characterize the total color deviation between the first tooth model and the second tooth model in each pair of tooth models after the first dental jaw model is similarly transformed according to the initial transformation parameters;

[0173] If the initial camera parameters, the initial transformation parameters, and the color loss function corresponding to the initial color field network meet the first preset condition, the initial transformation parameters are determined as the target transformation parameters.

[0174] Of course, the storage medium containing computer executable instructions provided in the embodiments of the present disclosure is not limited to the above method operations, and the computer executable instructions can also execute related operations in the orthodontic treatment monitoring method provided in any embodiment of the present disclosure.

[0175] Through the above description of the implementation methods, technicians in the relevant field can clearly understand that the present disclosure can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented through hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present disclosure can essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer cloud platform (which can be a personal computer, server, or network cloud platform, etc.) to execute the orthodontic treatment monitoring method provided by each embodiment of the present disclosure.

[0176] Note that the above are only preferred embodiments of the present disclosure and the technical principles used. Those skilled in the art will understand that the present disclosure is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present disclosure. Therefore, although the present disclosure is described in more detail through the above embodiments, the present disclosure is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present disclosure, and the scope of the present disclosure is determined by the scope of the appended claims.

Claims

1. A method for monitoring orthodontic treatment, characterized in that: include: Acquire a first dental jaw model located in a preset space, and a signed distance function of each first tooth model in the first dental jaw model, wherein the signed distance function is used to characterize a first distance from each first coordinate point in the preset space to the first tooth model; Acquire multiple intraoral images of the user and an initial color field network of a second dental model, wherein the second dental model corresponds to the multiple intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model; Determine initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, wherein the tooth model pair includes the first tooth model and the second tooth model corresponding to the same tooth; Determine, according to the signed distance function, initial camera parameters, the initial transformation parameters, and a color loss function corresponding to the initial color field network, wherein the initial camera parameters correspond to the multiple intraoral images, and the color loss function is used to characterize the total color deviation between the first tooth model and the second tooth model in each pair of the tooth models after the first dental jaw model is similarly transformed according to the initial transformation parameters; If the initial camera parameters, the initial transformation parameters, and the color loss function corresponding to the initial color field network meet a first preset condition, the initial transformation parameters are determined as target transformation parameters.

2. The method according to claim 1, characterized in that Also includes: If the color loss function corresponding to the initial camera parameters and the initial transformation parameters does not satisfy the first preset condition, optimizing the initial camera parameters, the initial transformation parameters and the initial color field network until the corresponding color loss function satisfies the first preset condition; The optimized initial transformation parameters are determined as the target transformation parameters.

3. The method according to claim 1 or 2, characterized in that: The determining of the initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model comprises: Performing three-dimensional reconstruction according to the multiple intraoral images to obtain a third dental jaw model; Transforming the third dental model into the preset space by similarity transformation; The first dental model and the transformed third dental model are roughly aligned to obtain the initial transformation parameters corresponding to each pair of the tooth models.

4. The method according to claim 1 or 2, characterized in that: The process of determining the signed distance function of each first tooth model in the first dental jaw model comprises: Sampling the first coordinate points in the preset space to obtain a plurality of sampling coordinate points; For each of the sampling coordinate points, the coordinates of the sampling coordinate point are input into a first encoder to obtain a corresponding first encoding vector, the first encoding vector is input into a backbone network to obtain a corresponding feature vector, the feature vector is input into an initial shallow network corresponding to each of the first tooth models, and a predicted distance from the sampling coordinate point to each of the first tooth models is obtained; Determine the distance loss function corresponding to the initial shallow network according to the predicted distance and the actual distance corresponding to the sampling coordinate point; If the distance loss function corresponding to the initial shallow network satisfies a second preset condition, determining the signed distance function of the first tooth model according to the initial shallow network corresponding to the first tooth model; If the distance loss function corresponding to the initial shallow network does not meet the second preset condition, the initial shallow network is optimized until the distance loss function corresponding to the optimized initial shallow network meets the second preset condition, and the signed distance function of the first tooth model is determined based on the optimized initial shallow network corresponding to the first tooth model.

5. The method according to claim 1 or 2, characterized in that: The step of determining the initial camera parameters, the initial transformation parameters, and the color loss function corresponding to the color field network according to the signed distance function includes: For each pixel point in the multiple intra-oral images, determine a second coordinate point and a ray corresponding to the pixel point in the world coordinate system according to the initial camera parameters, detect whether the ray passes through a first tooth model in the first dental model according to the signed distance function, and obtain a detection result, determine a second distance from the second coordinate point corresponding to the pixel point to the first dental model after similarity transformation according to the initial transformation parameters according to the detection result, and determine a first color value of the pixel point according to the second distance and the initial color field network; The color loss function is determined according to the first color value and the second color value of each pixel in the plurality of intra-oral images.

6. The method according to claim 5, characterized in that The detecting whether the ray passes through a first dental model in the first dental models to obtain a detection result includes: Collecting a plurality of sampling points on the ray; For each of the first tooth models, a plurality of first distances from the plurality of sampling points to the first tooth model are determined according to the signed distance function of the first tooth model, and whether the ray passes through the first tooth model is determined according to the plurality of first distances.

7. The method according to claim 5, characterized in that Determining, according to the detection result, a second distance from the second coordinate point corresponding to the pixel point to the first dental model after similarity transformation according to the initial transformation parameters, comprises: If the detection result is a first tooth model passing through the first dental model, a first distance from the second coordinate point corresponding to the pixel point to the first tooth model passed through by the ray corresponding to the pixel point is taken as the second distance; If the detection result is that it does not pass through the first dental model, the minimum value of the first distances from the second coordinate point corresponding to the pixel point to each of the first tooth models is used as the second distance.

8. An orthodontic treatment monitoring device, characterized in that: include: A first acquisition module, used for acquiring a first dental jaw model in a preset space, and a signed distance function of each first tooth model in the first dental jaw model, wherein the signed distance function is used for representing a first distance from each first coordinate point in the preset space to the first tooth model; A second acquisition module, used to acquire a plurality of intraoral images of the user and a color field network of a second dental model, wherein the second dental model corresponds to the plurality of intraoral images and includes a second tooth model, and the color field network is used to characterize the color value corresponding to each pair of second coordinate points and ray directions in the second dental model; A first determination module is used to determine initial transformation parameters corresponding to each pair of tooth models in the first jaw model and the second jaw model, wherein the tooth model pair includes the first tooth model and the second tooth model corresponding to the same tooth; a second determination module, configured to determine, according to the signed distance function, initial camera parameters, the initial transformation parameters, and a color loss function corresponding to the initial color field network, wherein the initial camera parameters correspond to the multiple intraoral images, and the color loss function is configured to characterize a total color deviation between the first tooth model and the second tooth model in each pair of the tooth models after the first dental jaw model is similarly transformed according to the initial transformation parameters; The third determination module is used to determine the initial transformation parameters as target transformation parameters if the color loss function corresponding to the initial camera parameters, the initial transformation parameters, and the initial color field network meets a first preset condition.

9. An electronic device, characterized in that: include: processor; A memory for storing executable instructions; Wherein, the processor is used to read the executable instructions from the memory and execute the executable instructions to implement the orthodontic treatment monitoring method described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the processor implements the orthodontic treatment monitoring method described in any one of claims 1 to 7.

Images