Tooth occlusion calculation method and system and storage medium

Abstract

The invention relates to the technical field of tooth occlusion calculation, in particular to a tooth occlusion calculation method and system and a storage medium. Acquiring specified position information based on the updated first preset model according to the to-be-detected tooth picture, the specified position information including a first position frame and a second position frame, and the first position frame and the second position frame being overlapped; acquiring a contact ratio according to the first position frame and the second position frame; and matching in preset types according to the contact proportion, and further obtaining the tooth occlusion type of the to-be-detected tooth corresponding to the corresponding to-be-detected tooth picture. The method is convenient to operate, time is saved, and the user can pay attention to the tooth problem of himself / herself in time.

Description

Technical Field

[0001] This application relates to the technical field of tooth occlusion calculation, and in particular to a tooth occlusion calculation method, system, and storage medium. Background Technology

[0002] In oral health, dental occlusion is a common and concerning issue. It can not only cause discomfort in oral function but also negatively impact overall health. Therefore, it is essential to pay close attention to dental occlusion.

[0003] The traditional detection method requires patients to go to the hospital for oral and dental X-ray examinations, then give the X-rays to the doctor for review, and finally determine the corresponding treatment plan based on the bite relationship of the teeth corresponding to the X-rays.

[0004] However, with the fast pace of life, most people are quite busy and do not go to the hospital for relevant examinations before dental problems occur. They also cannot obtain timely information about their bite by checking their teeth, which can lead to a failure to intervene in time when teeth are just beginning to deform, resulting in a lower overall personal image. Summary of the Invention

[0005] To facilitate operation and enable timely monitoring of dental issues, this application provides a method, system, and storage medium for calculating tooth occlusion.

[0006] In a first aspect, this application provides a method for calculating tooth occlusion, employing the following technical solution: A method for calculating tooth occlusion includes the following steps: Obtain an image of the tooth to be tested, wherein the image of the tooth to be tested is a photograph of the tooth to be tested; Based on the image of the tooth to be tested, specified position information is obtained according to the updated first preset model. The specified position information includes a first position border and a second position border, and the first position border overlaps with the second position border. The contact ratio is obtained based on the first position border and the second position border, and the contact ratio represents the contact ratio between the first position border and the second position border; Based on the contact ratio, a matching is performed in a preset type to obtain the occlusal type of the tooth to be tested corresponding to the image of the tooth to be tested.

[0007] By adopting the above technical solution, based on the acquired photo of the tooth to be tested, and inputting the photo into the first preset model, the first position border and the second position border can be obtained. The contact ratio can be obtained based on the first position border and the second position border, and the corresponding type can be obtained through the contact ratio. This method is easy to operate, saves time, and allows users to pay attention to their own dental problems in a timely manner.

[0008] The preferred method for obtaining the contact ratio based on the first position border and the second position border includes the following steps: The overlap width is obtained based on the first position border and the second position border, and the overlap width represents the horizontal length of the overlap between the first position border and the second position border; The ratio of the overlap width to the border length of the second position border is used as the contact ratio.

[0009] Preferably, obtaining the overlap width based on the first position border and the second position border includes the following steps: Based on the first preset model, obtain the coordinates of the first vertex corresponding to the first position border and the coordinates of the second vertex corresponding to the second position border; A first specified coordinate is obtained based on the first vertex coordinate and the second position bounding box, wherein the first specified coordinate represents the coordinate of the first vertex located within the second bounding box; A second specified coordinate is obtained based on the second vertex coordinate and the first position bounding box, wherein the second specified coordinate represents the coordinate of the second vertex located within the first bounding box; The overlap width is determined by the horizontal distance between the first specified coordinate and the second specified coordinate.

[0010] Preferably, obtaining the first position bounding box based on the updated first preset model using the image of the tooth to be tested includes the following steps: Based on the image of the tooth to be tested, a set of numbers is obtained according to the updated first preset model. The set of numbers includes the number value of the tooth to be tested and the tooth border corresponding to the number value. Matching is performed in the number set according to the first preset number, and the tooth border corresponding to the matched number value is used as the first position border.

[0011] By adopting the above technical solution, a set of numbers is obtained based on the first preset model, and the tooth border corresponding to the first preset number is selected from the set of numbers, thereby obtaining the corresponding first position border. This can obtain the first position border more accurately and improve the overall efficiency of calculating tooth occlusion.

[0012] Preferably, the first preset model includes a preset tooth recognition model, and the step of obtaining the second position bounding box based on the updated first preset model according to the image of the tooth to be tested includes the following steps: Based on the updated first preset model, obtain the reference model corresponding to the second preset number; Based on the number value corresponding to the first position border, a similar test border is obtained in the preset tooth recognition model, and the test border includes at least two tooth borders. The comparison values ​​between the tooth outline and the control model are obtained sequentially. The comparison values ​​represent the comparison values ​​between the tooth corresponding to the tooth outline and the tooth in the control model. The tooth border corresponding to the largest alignment value among all the alignment values ​​is selected as the second position border.

[0013] By adopting the above technical solution, the test bounding box is obtained in the preset tooth recognition model, and the tooth bounding box and the tooth comparison value of the control model are obtained in the test bounding box. By selecting the tooth bounding box with the largest tooth comparison value as the second position bounding box, the second position bounding box that overlaps with the first position bounding box can be obtained more accurately, thereby improving the efficiency of calculating tooth occlusion.

[0014] Preferably, before obtaining the specified location information based on the updated first preset model using the image of the tooth to be tested, the method further includes the following steps: The image of the tooth to be tested is input into the second preset model, the state of the tooth to be tested is obtained based on the second preset model, and it is determined whether the state of the tooth to be tested is a missing tooth state. If the tooth to be tested is in a missing tooth state, then the first preset model is updated according to the second preset model.

[0015] By adopting the above technical solution, the presence of missing teeth in the image of the tooth to be tested can be obtained through the second preset model. If missing teeth are found, the tooth number of the first preset model needs to be updated, thereby enabling more accurate acquisition of the first position border and the second position border, and further improving the calculation efficiency of tooth occlusion.

[0016] Preferably, updating the first preset model based on the second preset model includes the following steps: Obtain the missing tooth border based on the second preset model, and mark the position corresponding to the missing tooth border in the first preset model to generate several numbered borders; Several numbered borders are used as teeth to be numbered. The teeth to be numbered in the first preset model are renumbered in order to update the first preset model.

[0017] Preferably, after renumbering the teeth to be numbered in the first preset model, the following steps are also included: Based on the updated first preset model, obtain the first designated number and the first occlusion number, where the teeth corresponding to the first designated number and the first occlusion number are in an occlusal state. Determine whether there is a missing tooth in the tooth at the position corresponding to the first specified number or the tooth at the position corresponding to the first occlusal number; if there is a missing tooth in the tooth at the position corresponding to the first specified number or the tooth at the position corresponding to the first occlusal number, then obtain the second specified number and the second occlusal number according to the updated first preset model, and determine whether there is a missing tooth in the tooth at the position corresponding to the second specified number or the tooth at the position corresponding to the second occlusal number. If the tooth at the position corresponding to the second specified number and the tooth at the position corresponding to the second occlusal number exist, then the number border corresponding to the second specified number is used as the first position border, and the number border corresponding to the second occlusal number is used as the second position border, and the contact ratio is re-obtained based on the first position border and the second position border.

[0018] Secondly, this application provides a tooth occlusion calculation system, which adopts the following technical solution: A tooth occlusion calculation system, comprising: The photo acquisition module is used to acquire an image of the tooth to be tested, wherein the image of the tooth to be tested is a photograph of the tooth to be tested. The location acquisition module is used to acquire specified location information based on the image of the tooth to be tested and an updated first preset model. The specified location information includes a first location border and a second location border, and the first location border overlaps with the second location border. A data calculation module is used to obtain a contact ratio based on the first position border and the second position border, wherein the contact ratio represents the contact ratio between the first position border and the second position border. The type matching module is used to match the tooth occlusion type of the tooth to be tested in a preset type according to the contact ratio, thereby obtaining the tooth occlusion type of the tooth to be tested corresponding to the image of the tooth to be tested.

[0019] Thirdly, this application provides a storage medium, which adopts the following technical solution: A storage medium B stores at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the tooth occlusion calculation method described in the first aspect.

[0020] The tooth occlusion calculation method, system, and storage medium provided in this application embodiment, based on the acquired photograph of the tooth to be tested, input the photograph into a first preset model to obtain a first position border and a second position border, obtain the contact ratio based on the first and second position borders, and then obtain the corresponding type through the contact ratio, which is convenient for users to operate, saves time, and allows users to pay attention to their own dental problems in a timely manner; based on the first preset model, a set of numbers is obtained, and the tooth border corresponding to the preset number is filtered out from the set of numbers to obtain the corresponding first position border, which can obtain the first position border more accurately and improve the overall efficiency of tooth occlusion calculation. Attached Figure Description

[0021] Figure 1 This is a block diagram of the tooth occlusion calculation method provided in the embodiments of this application; Figure 2 This is a flowchart illustrating a method for obtaining the second position border provided in an embodiment of this application; Figure 3 This is a block diagram of a method for obtaining the overlap width provided in an embodiment of this application; Figure 4 This is a schematic diagram showing the positions of the first and second position borders; Figure 5 This embodiment provides a flowchart of the steps for determining missing teeth. Detailed Implementation

[0022] To better understand the purpose, technical solutions, and advantages of this application, it has been described and illustrated below with reference to the accompanying drawings and embodiments. However, those skilled in the art should understand that this application can be implemented without these details. In some cases, to avoid obscuring various aspects of this application due to unnecessary description, well-known methods, processes, systems, components, and / or circuits already described at a higher level will not be elaborated upon. It will be apparent to those skilled in the art that various modifications can be made to the embodiments disclosed in this application, and the general principles defined in this application can be applied to other embodiments and application scenarios without departing from the principles and scope of this application. Therefore, this application is not limited to the illustrated embodiments, but conforms to the broadest scope consistent with the scope of protection claimed in this application.

[0023] This application discloses a method for calculating tooth occlusion.

[0024] like Figure 1 As shown, the method for calculating tooth occlusion includes the following steps: S100, acquire images of the teeth to be tested.

[0025] The image of the tooth to be tested represents a photograph of the tooth to be tested. This photograph can be a photograph taken by the user himself or a photograph obtained by scanning at a hospital. In this embodiment, it is preferably set to a photograph taken by the user himself, but it is not limited to this.

[0026] S200: Based on the image of the tooth to be tested, obtain the information of the specified location according to the updated first preset model.

[0027] The specified location information includes a first location bounding box and a second location bounding box, and the first location bounding box overlaps with the second location bounding box. The first preset model can perform target detection on teeth within the image of the tooth to be tested. When the image of the tooth to be tested is input into the first preset model, it can obtain corresponding tooth images with numbers. The first preset model can be a conventional object detection network, such as RFCN / SSD / RCNN / FastRCNN / FasterRCNN / SPPNet / DPM / OverFeat / YOLO, etc.

[0028] In this embodiment, the first preset model is preferably YOLO, which is used to perform target detection on a large number of dental photographs. By training on multiple dental photographs, the first preset model can be obtained. The image of the tooth to be tested is input into the first preset model, and a dental photograph with a first position bounding box and a second position bounding box can be obtained. Both the first position bounding box and the second position bounding box are borders that mark the corresponding teeth on the dental photograph to be tested.

[0029] The specific detection process of the first preset model is as follows: First, the input image is divided into an S*S grid by continuously performing convolution operations. If the center of the target is located in one of the grids, then that grid is responsible for detecting the object. This function is designed in the loss function.

[0030] The network design and loss function are existing technologies and will not be described in detail here. This application only requires inputting the photograph of the tooth to be tested into a first preset model, and then outputting the corresponding specified location information based on the first preset model.

[0031] It should be noted that before obtaining the first position bounding box and the second position bounding box through the first preset model, it is necessary to perform target recognition on the teeth on the test tooth photograph according to the first preset model to obtain each recognized tooth, then select the recognized teeth to obtain the tooth bounding box, then number each tooth bounding box in sequence, and then obtain the specified position information according to the numbered tooth bounding box.

[0032] Furthermore, when acquiring the tooth outlines, coordinates are marked for each tooth outline based on its position. In the first preset model, the left incisor is numbered 1-1, and then the upper teeth are numbered sequentially along the direction away from the right incisor, based on the position of the left incisor. Similarly, the right incisor is numbered 3-1, and the upper teeth are numbered sequentially along the direction away from the left incisor, based on the position of the right incisor. The lower teeth are numbered sequentially in the same way, which will not be elaborated upon here.

[0033] It should be noted that the existing technology involves target recognition of the teeth in the photograph of the teeth to be tested, obtaining each identified tooth, then selecting the identified teeth to obtain tooth borders, and sequentially numbering each tooth border. This will not be described in detail here. Simply input the image of the teeth to be tested into the first preset model, and then you can obtain a photograph of the teeth with first position borders and second position borders.

[0034] In medicine, calculating an individual's bite is primarily based on the occlusion between two specific upper and lower teeth. This calculation determines the type of teeth an individual possesses. There are three types of bite relationships: neutral malocclusion, distal malocclusion, and mesial malocclusion.

[0035] Neutral malocclusion refers to a normal mesiodistal relationship between the maxilla and mandible, and the dental arch. Specifically, in centric occlusion, the mesiobuccal cusp of the maxillary first permanent molar occludes within the mesiobuccal groove of the mandibular first permanent molar. The second type of malocclusion is called distal malocclusion. In this case, the mandible is in a distal position, and if it retracts by 1 / 4 of a molar or half a premolar distance, meaning the mesiobuccal cusps of the maxillary and mandibular first permanent molars are opposite each other, it is called a mild distal malocclusion. If the mandible moves further towards both ends of the teeth, such that the mesiobuccal cusp of the maxillary first permanent molar occludes with the mandibular first permanent molar and second premolar, it is a complete distal malocclusion. The third type of mesial malocclusion is in a mesial position. In this case, if the mandible moves forward by 1 / 4 of a molar or half a premolar distance, meaning the mesiobuccal cusp of the maxillary first permanent molar is opposite the disbuccal cusp of the mandibular first permanent molar, it is called a mild mesial malocclusion. If the mandible is displaced mesially by half the distance of a molar or a premolar, such that the mesiobuccal cusp of the maxillary first permanent molar occludes between the mandibular first and second permanent molars, it is a complete mesial malocclusion.

[0036] Occlusal relationship is mainly used as an important indicator for judging malocclusion. Only with a stable occlusal structure can the teeth lock together, achieving long-term stability, easy cleaning, longer lifespan, and symmetrical facial muscles. A positive occlusal structure indicates that the teeth are normal and the occlusion is healthy.

[0037] For example, assuming this person has never had teeth extracted and has no missing teeth, numbering starts from the incisors: the left incisors are numbered 1-1, the right incisors 3-1, and so on inwards. The left maxillary first permanent molars are numbered 1-6, and the right maxillary first permanent molars are numbered 3-6. The left mandibular first permanent molars are numbered 2-6, and the right mandibular first permanent molars are numbered 4-6, symmetrical to the left mandibular first permanent molars. The tooth corresponding to the first position border mentioned here is the left maxillary first permanent molar, and the tooth corresponding to the second position border is the left mandibular first permanent molar.

[0038] To simplify the process of users obtaining their own dental occlusion information, the first and second position bounding boxes can be obtained by inputting an image of the teeth to be tested taken by the user into a first preset model. However, to ensure the accuracy of the overall dental occlusion calculation, it is necessary to accurately obtain the first and second position bounding boxes. Therefore, the specific process of obtaining the first position bounding box based on the first preset model is as follows.

[0039] In another embodiment, obtaining a first location bounding box based on an updated first preset model from an image of the tooth to be tested includes the following steps: S210, obtain a set of numbers based on the image of the tooth to be tested and the updated first preset model.

[0040] S220, Matching is performed in the number set according to the first preset number, and the tooth border corresponding to the matched number value is used as the first position border.

[0041] The number set includes the numerical values ​​of the teeth to be tested and the corresponding tooth borders. The first preset number represents teeth numbered 1-6, which are matched in the number set of the first preset model, and the borders of teeth numbered 1-6 in the number set are used as the first position borders.

[0042] Although the accurate first position bounding box can be obtained based on the number set of the first preset model, the lower row of teeth in the image of teeth biting together has inaccurate dentition, which means that teeth numbered 2-6 do not have a biting relationship with the teeth in the first position bounding box. Therefore, the second position bounding box cannot be obtained based on the first preset model.

[0043] Combination Figure 2 In another embodiment, obtaining the second position bounding box based on the image of the tooth to be tested using a first preset model includes the following steps: S230, obtain the reference model corresponding to the second preset number based on the updated first preset model.

[0044] S240, based on the number value corresponding to the first position border, obtain the similar test border in the preset tooth recognition model.

[0045] S250, sequentially obtain the comparison values ​​between the tooth outline and the control model.

[0046] S260, select the tooth border corresponding to the largest comparison value among all comparison values ​​as the second position border.

[0047] The first preset model includes a preset tooth recognition model, which is a pre-established target recognition model mainly used to identify the occlusal teeth corresponding to teeth numbered 1-6. The second preset number represents teeth numbered 2-6, and the control model represents the model of teeth numbered 2-6. The control model is mainly obtained based on the first preset model. During modeling, the first preset model stores a large amount of tooth data, enabling the acquisition of models for teeth numbered 2-6. The test bounding box includes at least two tooth bounding boxes, each corresponding to one tooth. The comparison value represents the comparison value between the tooth corresponding to the tooth bounding box and the tooth in the control model.

[0048] In addition, to simplify the calculation, the bounding box to be tested includes a maximum of three tooth bounding boxes. It should be noted that the bounding box to be tested is mainly obtained based on the preset tooth recognition model. Specifically, it is obtained by finding several mandibular tooth bounding boxes that are closest to the first position bounding box among the recognized mandibular tooth bounding boxes based on the first position bounding box. The number of mandibular tooth bounding boxes shall not exceed three. The bounding box to be tested includes several mandibular tooth bounding boxes.

[0049] Specifically, a first position bounding box is obtained based on a first preset model. Four coordinates are obtained based on the first position bounding box. In the tooth bounding box corresponding to the preset tooth recognition model, the three adjacent bounding boxes that are closest to the four coordinates are obtained. The vertex coordinates of the three adjacent bounding boxes are mainly compared. The bounding box whose coordinates are closest to the first position bounding box is used as the bounding box to be tested.

[0050] Then, the tooth pixels within each border are obtained and matched with the pixels of the reference model. The border with the highest matching degree is selected as the second position border.

[0051] The preset tooth recognition model is mainly trained by multiple tooth photos to obtain a model corresponding to each tooth. After inputting the image of the tooth to be tested into the preset tooth recognition model, the bounding box to be tested can be obtained, and the tooth bounding box with the highest similarity to the control model is selected as the second position bounding box.

[0052] It's important to note that the tooth detection uses existing technology. This involves inputting an image of the tooth to be detected into a pre-defined tooth recognition model to obtain the corresponding second-position bounding box. These pre-defined tooth recognition models are all object detection models, such as RFCN / SSD / RCNN / FastRCNN / FasterRCNN / SPPNet / DPM / OverFeat / YOLO, etc. These models are trained using multiple sets of tooth data and will not be described in detail here.

[0053] S300, the contact ratio is obtained based on the first position border and the second position border.

[0054] The contact ratio represents the contact ratio between the first position border and the second position border. The first position border corresponds to the borders of the maxillary teeth numbered 1-6, while the second position border corresponds to the borders of the mandibular teeth numbered 2-6. During normal tooth occlusion calculation, teeth numbered 1-6 and 2-6 need to be in an occlusal state. Therefore, after the user inputs the photographs of the teeth to be tested into the first preset model, the output tooth photographs will have the first and second position borders marked on the teeth numbered 1-6 and 2-6. Since teeth numbered 1-6 and 2-6 are in an occlusal state, the first and second position borders overlap. The corresponding contact ratio is calculated based on the overlapping portion of the first and second position borders. The specific steps for calculating the contact ratio are shown below.

[0055] The contact ratio is obtained based on the first and second position borders, including the following steps: S310, obtain the overlap width based on the first position border and the second position border.

[0056] S320 uses the ratio of the overlap width to the length of the second position border as the contact ratio.

[0057] The overlap width represents the horizontal length of the overlap between the first position border and the second position border, and the length of the second position border represents the length of the second position border along the tooth arrangement direction.

[0058] Combination Figure 3 The method of obtaining the overlap width based on the first position border and the second position border includes the following steps: S311, obtaining the first vertex coordinates corresponding to the first position border and the second vertex coordinates corresponding to the second position border based on the first preset model.

[0059] S312, obtain the first specified coordinates based on the first vertex coordinates and the second position bounding box.

[0060] S313, obtain the second specified coordinates based on the coordinates of the second vertex and the first position bounding box.

[0061] S314, the overlap width is determined by the horizontal distance between the first specified coordinate and the second specified coordinate.

[0062] Wherein, the first vertex coordinates represent the vertex position coordinates of the first position bounding box, and the first vertex coordinates include the coordinates of the four vertices of the first position bounding box. The second vertex coordinates represent the coordinates of the four vertices of the second position bounding box. The first specified coordinates represent the coordinates of the first vertex located within the second bounding box, and the second specified coordinates represent the coordinates of the second vertex located within the first bounding box.

[0063] Specifically, the key is that the first specified coordinate is within the second border, and the first specified coordinate is one of the four vertex positions of the first position border; the second specified coordinate is within the first border, and the second specified coordinate is one of the four vertex positions of the second position border. By obtaining the first specified coordinate and the second specified coordinate, the overlapping portion between the first position border and the second position border can be obtained.

[0064] The lateral distance refers to the horizontal distance between the first and second specified coordinates along the direction of tooth alignment. Specifically, the calculation method involves obtaining the coordinate furthest from the incisor based on the first and second specified coordinates, and then subtracting the horizontal axis coordinate of the coordinate furthest from the incisor from the horizontal axis coordinate of the coordinate closest to the incisor. The resulting value is the overlap width.

[0065] like Figure 4 As shown, the coordinates of the first vertices corresponding to the first position border A are A1, A2, A3, and A4, respectively, and the coordinates of the second vertices corresponding to the second position border B are B1, B2, B3, and B4, respectively. As shown in the figure, the first vertex coordinate A4 is the first specified coordinate, and the second vertex coordinate B1 is the second specified coordinate. Based on the positions of the first position border A and the second position border B, it can be seen that the second specified coordinate is far from the position of the incisor. Therefore, the overlap width is the x-coordinate of the first specified coordinate minus the x-coordinate of the second specified coordinate.

[0066] It should be noted that the actual values ​​of the coordinates of the first vertex and the second vertex will be reflected in the first preset model. For ease of understanding, letters are used to represent them here.

[0067] S400, based on the contact ratio, matches within a preset type to obtain the occlusal type of the tooth to be tested corresponding to the image of the tooth to be tested.

[0068] Among them, the preset type represents the set value for identifying the type of tooth occlusion. The preset type can be divided into three types: mesial, centric, and distal. However, the classification method of each type is mainly based on Table 1.

[0069] Table 1 Preset Types Contact ratio (N) Types of tooth occlusion N<0.25 Near and Middle 0.25<＝N<0.75 middle N>＝0.75 Far and Middle As shown in Table 1, when the contact ratio N is less than 0.25, the occlusion type is mesial; when the contact ratio N is greater than or equal to 0.75, the occlusion type is distal; and when the contact ratio N is greater than or equal to 0.25 and less than 0.75, the occlusion type is centric.

[0070] It's important to clarify that the values ​​of 0.25 and 0.75 are derived from extensive data collection, primarily based on medical imaging to determine mesial, centric, and distal occlusion types. The corresponding contact ratios were then calculated, and through multiple trials, the following results were obtained: a contact ratio less than 0.25 indicates a mesial occlusion; a contact ratio greater than or equal to 0.75 indicates a distal occlusion; and a contact ratio greater than or equal to 0.25 but less than 0.75 indicates a centric occlusion. The exact method for obtaining 0.25 and 0.75 will not be detailed here. Essentially, calculations were performed on different occlusal teeth to determine the actual occlusal type, and the preset type was obtained by continuously adjusting the contact ratio.

[0071] In another embodiment, before obtaining the specified location information based on the updated first preset model from the image of the tooth to be tested, the following steps are also included: S500: Input the image of the tooth to be tested into the second preset model, obtain the state of the tooth to be tested based on the second preset model, and determine whether the state of the tooth to be tested is a missing tooth state.

[0072] S600, if the tooth to be tested is in a missing tooth state, then update the first preset model according to the second preset model.

[0073] The second preset model represents a model for target detection of teeth, primarily used to detect whether a tooth is missing. In the first preset model, teeth are numbered by acquiring the corresponding tooth model, selecting a bounding box within that model, and then obtaining the corresponding number. However, when a tooth is missing, the first preset model cannot obtain a model of that location, making it impossible to number that position. Therefore, teeth numbered 1-6 based on the original first preset model are not the desired right maxillary first permanent molar. In this case, it is necessary to obtain the state of the tooth based on the image of the tooth under test using the second preset model, and then determine whether the tooth is missing. If so, the original first preset model needs to be updated based on the second preset model. The specific process for updating the first preset model is as follows.

[0074] It should be noted that the second preset model uses object detection models, such as RFCN / SSD / RCNN / FastRCNN / FasterRCNN / SPPNet / DPM / OverFeat / YOLO, etc. These are all object detection models trained on multiple sets of dental data, and will not be described in detail here. The second preset model only needs to be able to determine whether the tooth in the input image is missing.

[0075] In another embodiment, updating the first preset model based on the second preset model includes the following steps: S610: Obtain the missing tooth border according to the second preset model, and mark the position corresponding to the missing tooth border in the first preset model to generate several numbered borders.

[0076] S620, Several numbered borders are used as teeth to be numbered, and the teeth to be numbered in the first preset model are renumbered in order to update the first preset model.

[0077] The missing tooth border representation is based on the missing tooth border obtained from the second preset model. This is mainly achieved by inputting the image of the tooth to be tested into the second preset model, then using the second preset model to select all the missing tooth parts in the image of the tooth to be tested to obtain the missing tooth border. Then, the position corresponding to the missing tooth border is marked in the first preset model, thereby generating several numbered borders. The first preset model uses the numbered borders as the teeth to be numbered, and then readjusts the numbering of the teeth to be tested corresponding to the image of the tooth to be tested. Even if the tooth corresponding to the image of the tooth to be tested is missing, the position is numbered according to the numbered border, thereby ensuring that the first position border obtained according to the first preset model is the position border corresponding to the teeth numbered 1-6.

[0078] In some cases, if a user's teeth numbered 1-6 or 2-6 are missing, the user cannot test the occlusion relationship. Therefore, to ensure the validity of the user's occlusion test, in another embodiment, when teeth numbered 1-6 or 2-6 are missing, the teeth symmetrical to the original 1-6 can be obtained, namely the first permanent molars of the upper right jaw. The first permanent molars of the upper right jaw are numbered 3-6 using a first preset model. Then, a preset tooth recognition model is used to obtain the second position bounding box of the first permanent molars of the lower right jaw, numbered 4-6, corresponding to the teeth numbered 3-6.

[0079] The second preset model can also be trained on multiple dental photographs to obtain a model corresponding to each tooth. For example, based on the second preset model, a model of the incisors is obtained. When inputting a dental photograph, each tooth is first selected by bounding boxes. Then, the selected elements are compared with the model corresponding to each tooth to obtain the corresponding tooth and determine the position of the incisors. Then, based on the starting number of the incisors, the left incisor is numbered 1-1, the right incisor is numbered 3-1, and so on, according to the order of the two incisors. Then, it is determined whether the position between adjacent numbers is greater than a preset value. If so, it means that there is a possibility of a missing tooth between adjacent numbers. Then, the gap between adjacent numbers is bounded to obtain the missing tooth border. Next, the position corresponding to the missing tooth border is marked in the first preset model to generate several numbered borders. These numbered borders are used as teeth to be numbered, and the teeth to be numbered in the first preset model are renumbered.

[0080] Reference Figure 5 In another embodiment, after renumbering the teeth to be numbered in the first preset model, the following steps are also included: S621, obtain the first designated number and the first occlusion number according to the updated first preset model, and the teeth corresponding to the first designated number and the first occlusion number are in an occlusion state.

[0081] S622, determine whether there is a missing tooth in the tooth at the position corresponding to the first specified number or the tooth at the position corresponding to the first occlusal number.

[0082] S623, if there is a missing tooth in the tooth at the position corresponding to the first specified number or the tooth at the position corresponding to the first occlusal number, then obtain the second specified number and the second occlusal number according to the updated first preset model, and determine whether there is a missing tooth in the tooth at the position corresponding to the second specified number or the tooth at the position corresponding to the second occlusal number.

[0083] S624, if the tooth at the position corresponding to the second specified number and the tooth at the position corresponding to the second occlusal number exist, then the number border corresponding to the second specified number is used as the first position border, and the number border corresponding to the second occlusal number is used as the second position border, and the contact ratio is re-obtained based on the first position border and the second position border.

[0084] The second designated number represents the teeth numbered 3-6 obtained based on the first preset model, and the second occlusion number represents the number of the right mandibular first permanent molar that occludes with the teeth numbered 4-6.

[0085] Specifically, step S621 involves obtaining the first designated number and the first occlusal number that occludes with the first designated number based on the updated first preset model. Then, in step S622, it is determined whether there are missing teeth at the positions corresponding to the first designated number and the first occlusal number. Only when neither the first designated number nor the first occlusal number is missing can the user's occlusal relationship be calculated. Therefore, steps S623 and S624 result in two scenarios.

[0086] In step S623, if a tooth is missing at the position corresponding to the first specified number or the position corresponding to the first occlusal number, then the second specified number and the second occlusal number are obtained based on the updated first preset model, and it is determined whether a tooth is missing at the position corresponding to the second specified number or the position corresponding to the second occlusal number. In step S624, if a tooth is missing at both the position corresponding to the second specified number and the position corresponding to the second occlusal number, then the number border corresponding to the second specified number is used as the first position border, and the number border corresponding to the second occlusal number is used as the second position border, and steps S300-S400 are repeated.

[0087] Steps S300-S400 are repeated here, mainly to re-obtain the contact ratio based on the first position border and the second position border, and then re-obtain the corresponding tooth occlusion type.

[0088] It should be noted that in step S624, if the tooth at the position corresponding to the second specified number or the tooth at the position corresponding to the second occlusion number does not exist, it means that the user cannot calculate the tooth occlusion using this method. Therefore, the user can have an X-ray taken at a dental hospital to obtain the corresponding result.

[0089] The implementation principle is as follows: The system acquires images of the teeth to be tested, captured by the user, and inputs them into a first preset model to obtain corresponding numbered tooth images. Then, it retrieves a set of numbers based on these numbered images and matches them against preset numbers. The tooth border corresponding to the matched number is used as the first position border. Next, a reference model with a second preset number is obtained based on the first preset model. Similar test borders are retrieved from the preset tooth recognition model based on the number corresponding to the first position border. The comparison values ​​between the tooth borders and the reference models are then obtained, representing the comparison value between the tooth border and the tooth in the reference model. The tooth border corresponding to the largest comparison value is selected as the second position border. The overlap width is obtained based on the first and second position borders, and the ratio of the overlap width to the length of the second position border is used as the contact ratio. Matching is performed within a preset type based on the contact ratio to obtain the corresponding occlusal type of the tooth to be tested. This allows the system to obtain the current user's tooth occlusion type, providing convenience and enabling users to quickly identify their tooth type. When teeth are in a mesial or distal position, timely intervention can be implemented to improve dental health.

[0090] This application also discloses a tooth occlusion calculation system and a tooth occlusion calculation method.

[0091] A tooth occlusion calculation system includes an image acquisition module, a location acquisition module, a data calculation module, and a type matching module. The image acquisition module acquires an image of the tooth to be tested. The location acquisition module acquires specified location information based on the image of the tooth to be tested using a first preset model. The data calculation module acquires the contact ratio based on a first location border and a second location border. The type matching module matches the tooth to be tested within a preset type based on the contact ratio, thereby acquiring the tooth occlusion type corresponding to the image of the tooth to be tested.

[0092] Among them, the image of the tooth to be tested represents a photograph of the tooth to be tested, the specified position information includes a first position border and a second position border, and the first position border overlaps with the second position border, and the contact ratio represents the contact ratio between the first position border and the second position border.

[0093] For example, based on the image of the tooth to be tested, the specified position information is obtained based on the updated first preset model, the contact ratio is obtained based on the first position border and the second position border, and the contact ratio is matched in the preset type to obtain the corresponding image of the tooth to be tested. The tooth occlusion type of the tooth to be tested is the same as or similar to that disclosed in the tooth occlusion calculation method. Therefore, it will not be elaborated on here.

[0094] This application also discloses a storage medium that stores at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement a tooth occlusion calculation method.

[0095] The other functions performed in the above-mentioned photo acquisition module, location acquisition module, data calculation module, and type matching module, as well as the technical details of each function, are the same as or similar to the corresponding features in the tooth occlusion calculation method described above, so they will not be repeated here.

[0096] It should be understood that although the steps in the flowcharts in the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise expressly stated herein, there is no strict order in which these steps are performed, and they may be performed in other orders.

[0097] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for calculating tooth occlusion, characterized in that, Includes the following steps: Obtain an image of the tooth to be tested, wherein the image of the tooth to be tested is a photograph of the tooth to be tested; Based on the image of the tooth to be tested, specified position information is obtained according to the updated first preset model. The specified position information includes a first position border and a second position border, and the first position border overlaps with the second position border. The contact ratio is obtained based on the first position border and the second position border, and the contact ratio represents the contact ratio between the first position border and the second position border; Based on the contact ratio, a matching is performed in a preset type to obtain the occlusal type of the tooth to be tested corresponding to the image of the tooth to be tested.

2. The tooth occlusion calculation method according to claim 1, characterized in that, The step of obtaining the contact ratio based on the first position border and the second position border includes the following steps: The overlap width is obtained based on the first position border and the second position border, and the overlap width represents the horizontal length of the overlap between the first position border and the second position border; The ratio of the overlap width to the border length of the second position border is used as the contact ratio.

3. The tooth occlusion calculation method according to claim 2, characterized in that, The process of obtaining the overlap width based on the first position border and the second position border includes the following steps: Based on the first preset model, obtain the coordinates of the first vertex corresponding to the first position border and the coordinates of the second vertex corresponding to the second position border; A first specified coordinate is obtained based on the first vertex coordinate and the second position bounding box, wherein the first specified coordinate represents the coordinate of the first vertex located within the second bounding box; A second specified coordinate is obtained based on the second vertex coordinate and the first position bounding box, wherein the second specified coordinate represents the coordinate of the second vertex located within the first bounding box; The overlap width is determined by the horizontal distance between the first specified coordinate and the second specified coordinate.

4. The tooth occlusion calculation method according to claim 1, characterized in that, The step of obtaining the first position bounding box based on the updated first preset model of the image of the tooth to be tested includes the following steps: Based on the image of the tooth to be tested, a set of numbers is obtained according to the updated first preset model. The set of numbers includes the number value of the tooth to be tested and the tooth border corresponding to the number value. Matching is performed in the number set according to the first preset number, and the tooth border corresponding to the matched number value is used as the first position border.

5. The tooth occlusion calculation method according to claim 4, characterized in that, The first preset model includes a preset tooth recognition model. The step of obtaining the second position bounding box based on the updated first preset model using the image of the tooth to be tested includes the following steps: Based on the updated first preset model, obtain the reference model corresponding to the second preset number; Based on the number value corresponding to the first position border, a similar test border is obtained in the preset tooth recognition model, and the test border includes at least two tooth borders. The comparison values ​​between the tooth outline and the control model are obtained sequentially. The comparison values ​​represent the comparison values ​​between the tooth corresponding to the tooth outline and the tooth in the control model. The tooth border corresponding to the largest alignment value among all the alignment values ​​is selected as the second position border.

6. The method for calculating tooth occlusion according to claim 1, characterized in that, Before obtaining the specified location information based on the updated first preset model using the image of the tooth to be tested, the following steps are also included: The image of the tooth to be tested is input into the second preset model, the state of the tooth to be tested is obtained based on the second preset model, and it is determined whether the state of the tooth to be tested is a missing tooth state. If the tooth to be tested is in a missing tooth state, then the first preset model is updated according to the second preset model.

7. The tooth occlusion calculation method according to claim 6, characterized in that, The step of updating the first preset model based on the second preset model includes the following steps: Obtain the missing tooth border based on the second preset model, and mark the position corresponding to the missing tooth border in the first preset model to generate several numbered borders; Several numbered borders are used as teeth to be numbered. The teeth to be numbered in the first preset model are renumbered in order to update the first preset model.

8. The tooth occlusion calculation method according to claim 7, characterized in that, After renumbering the teeth to be numbered in the first preset model, the following steps are also included: Based on the updated first preset model, obtain the first designated number and the first occlusion number, where the teeth corresponding to the first designated number and the first occlusion number are in an occlusal state. Determine whether there is a missing tooth in the tooth at the position corresponding to the first specified number or the tooth at the position corresponding to the first occlusal number; If the tooth at the position corresponding to the first specified number or the tooth at the position corresponding to the first occlusal number is missing a tooth, then the second specified number and the second occlusal number are obtained according to the updated first preset model, and it is determined whether the tooth at the position corresponding to the second specified number or the tooth at the position corresponding to the second occlusal number is missing a tooth. If the tooth at the position corresponding to the second specified number and the tooth at the position corresponding to the second occlusal number exist, then the number border corresponding to the second specified number is used as the first position border, and the number border corresponding to the second occlusal number is used as the second position border, and the contact ratio is re-obtained based on the first position border and the second position border.

9. A tooth occlusion calculation system, characterized in that, include: The photo acquisition module is used to acquire an image of the tooth to be tested, wherein the image of the tooth to be tested is a photograph of the tooth to be tested. The location acquisition module is used to acquire specified location information based on the image of the tooth to be tested and an updated first preset model. The specified location information includes a first location border and a second location border, and the first location border overlaps with the second location border. A data calculation module is used to obtain a contact ratio based on the first position border and the second position border, wherein the contact ratio represents the contact ratio between the first position border and the second position border. The type matching module is used to match the tooth occlusion type of the tooth to be tested in a preset type according to the contact ratio, thereby obtaining the tooth occlusion type of the tooth to be tested corresponding to the image of the tooth to be tested.

10. A storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or instruction set is loaded and executed by a processor to implement the tooth occlusion calculation method as described in any one of claims 1-8.

Images