WO2023151549A1 - 面中线及牙齿矫治位置的确定方法、制造方法及系统 - Google Patents

面中线及牙齿矫治位置的确定方法、制造方法及系统 Download PDF

Info

Publication number
WO2023151549A1
WO2023151549A1 PCT/CN2023/074727 CN2023074727W WO2023151549A1 WO 2023151549 A1 WO2023151549 A1 WO 2023151549A1 CN 2023074727 W CN2023074727 W CN 2023074727W WO 2023151549 A1 WO2023151549 A1 WO 2023151549A1
Authority
WO
WIPO (PCT)
Prior art keywords
midline
target
dimensional
tooth
determining
Prior art date
Application number
PCT/CN2023/074727
Other languages
English (en)
French (fr)
Inventor
任远
王静
王雅静
冯洋
刘晓林
王明政
Original Assignee
上海时代天使医疗器械有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN202210117981.2A external-priority patent/CN116612505A/zh
Priority claimed from CN202211740579.6A external-priority patent/CN116035732A/zh
Application filed by 上海时代天使医疗器械有限公司 filed Critical 上海时代天使医疗器械有限公司
Publication of WO2023151549A1 publication Critical patent/WO2023151549A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions

Definitions

  • the present application relates to the technical field of orthodontics, and in particular to a method, manufacturing method and system for determining a mid-surface line and a position of orthodontics.
  • each tooth is adjusted, for example, taking the position of the facial midline of the person to be treated as a reference, and according to the relative relationship between the position of the tooth midline of the person to be treated and the facial midline, determine whether the teeth are adjusted to the left or right relative to the facial midline Adjustment.
  • the orthodontic designer needs to observe the two-dimensional photos of the exposed teeth of the person to be treated, and combine personal experience to roughly estimate the position of the mid-face midline in the dentition, and then manually place the crown and gums of the person to be treated In the three-dimensional dental model, the position of the midface midline in the three-dimensional dental model is estimated.
  • This method is not only a cumbersome process, but also the accuracy of the determined midface midline is low, and it has certain randomness and error.
  • the technical problem to be solved by the present invention is to overcome the need to manually estimate the position of the facial midline in the three-dimensional dental model in the process of orthodontic treatment in the prior art.
  • the determination process of the facial midline is cumbersome, and the accuracy of the determined facial midline Low, with a certain degree of randomness and error defects, providing a method for determining the mid-face and orthodontics, a manufacturing method and a system.
  • a method for determining the midface midline in a three-dimensional dental model includes:
  • the two-dimensional photo of the target includes the face of the target object showing the target teeth
  • the position of the second tooth midline in the three-dimensional dental model is obtained.
  • the obtaining the position of the first midline of the target object in the target two-dimensional photo and the position of the first tooth midline include:
  • the target teeth include at least one of the two maxillary central incisors, or at least one of the two mandibular central incisors.
  • the step of obtaining the facial feature anchor point of the facial region it also includes:
  • the step of obtaining the position of the first plane midline based on the facial feature anchor point includes:
  • the facial feature positioning points include eyes, eyebrows, bridge of the nose, lips, and corresponding positioning points of the facial contour;
  • the facial feature key points include eyebrow center point, eye corner point, nose tip point, nose base point, and midpoint At least three of the facial feature positioning points.
  • the step of acquiring the facial area in the target two-dimensional photo includes:
  • the step of obtaining the facial feature anchor points of the facial region includes:
  • the method for obtaining the position of the first plane midline based on the facial feature anchor point includes:
  • the step of obtaining the position of the midline of the first tooth based on the position of the target tooth in the target two-dimensional photo includes:
  • the position of the second tooth midline of the target object in the three-dimensional jaw model is acquired based on the target tooth, and the second size value of the target tooth in the three-dimensional jaw model is acquired.
  • the steps include:
  • the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the mandible is used as the position of the midline of the second tooth, and the corresponding The second dimension value of the central incisor.
  • the position of the second tooth midline of the target object in the three-dimensional jaw model is acquired based on the target tooth, and the second size value of the target tooth in the three-dimensional jaw model is acquired.
  • the steps include:
  • the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the mandible is used as the position of the second tooth midline, and based on the preset size The value corresponding to the second dimension value of the central incisor is obtained.
  • the step of acquiring a first size value of the target tooth in the target two-dimensional photo, and establishing a mapping relationship between the first size value and the second size value includes;
  • a mapping relationship between the first size value and the second size value is established based on the first size value and the second size value corresponding to the target tooth.
  • the step of acquiring a first size value of the target tooth in the target two-dimensional photo, and establishing a mapping relationship between the first size value and the second size value includes;
  • a mapping relationship between the first size value and the second size value is established based on the first size value and the second size value corresponding to the target tooth.
  • the step of obtaining the position of the second facial midline in the three-dimensional dental model based on the position of the second tooth midline in the three-dimensional dental model and the second deviation value includes :
  • a position of the second mid-face in the three-dimensional dental model is obtained based on the relative position.
  • a method for determining the position of orthodontic treatment includes:
  • the orthodontic position corresponding to each tooth of the target object is determined based on the three-dimensional coordinate system and the position of the second tooth midline of the target object in the three-dimensional dental model.
  • a manufacturing method of a dental appliance comprising:
  • the shape of the dental appliance of the target object is determined based on the target orthodontic position, so as to manufacture a dental appliance with a corresponding shape.
  • a system for determining the midface midline in a three-dimensional dental model comprising:
  • the first obtaining module is used to obtain the position of the first surface midline and the position of the first tooth midline of the target object in the target two-dimensional photo;
  • the two-dimensional photo of the target includes the face of the target object showing the target teeth
  • a second acquiring module configured to acquire the position of the second tooth midline of the target object in the three-dimensional jaw model based on the target tooth, and acquire a second size of the target tooth in the three-dimensional jaw model value;
  • a mapping relationship establishment module configured to obtain a first size value of the target tooth in the target two-dimensional photo, and establish a mapping relationship between the first size value and the second size value;
  • a first deviation value acquisition module configured to obtain a first deviation value between the first plane center line and the first tooth center line based on the position of the first plane center line and the position of the first tooth center line;
  • a second deviation value acquiring module configured to obtain a second deviation between the second centerline of the target object in the three-dimensional dental model and the second tooth centerline based on the mapping relationship and the first deviation value value;
  • a third obtaining module configured to obtain the position of the second mid-face midline in the three-dimensional dental model based on the position of the second tooth midline in the three-dimensional dental model and the second deviation value.
  • the first acquisition module includes:
  • a two-dimensional photo acquisition unit configured to acquire the target two-dimensional photo of the target object
  • a facial area acquiring unit configured to acquire the facial area in the target two-dimensional photo
  • An anchor point acquisition unit configured to acquire facial feature anchor points of the facial region
  • a first plane midline acquisition unit configured to obtain the position of the first plane midline based on the facial feature positioning point
  • a first tooth midline acquisition unit configured to obtain the position of the first tooth midline based on the position of the target tooth in the target two-dimensional photo
  • the target teeth include at least one of the two maxillary central incisors, or at least one of the two mandibular central incisors.
  • the first acquisition module further includes:
  • a face tilt judging unit configured to judge whether the face of the target object in the target two-dimensional photo is tilted based on the position of the facial feature anchor point;
  • a correction unit configured to perform a rotation correction operation on the facial feature anchor point when the face tilt judging unit judges that the face of the target object in the target two-dimensional photo is tilted, and run the first Surface midline acquisition unit;
  • the face inclination judging unit judges that the face of the target object in the target two-dimensional photo has no inclination, the first midline acquiring unit is executed.
  • the first surface midline acquisition unit is specifically configured to select facial feature key points from the facial feature anchor points, and obtain the position of the first facial feature key point based on the facial feature key points;
  • the facial feature positioning points include eyes, eyebrows, bridge of the nose, lips, and corresponding positioning points of the facial contour;
  • the facial feature key points include eyebrow center point, eye corner point, nose tip point, nose base point, and midpoint At least three of the facial feature positioning points.
  • the facial region acquisition unit is specifically configured to perform facial region recognition algorithm based on the Image recognition is performed on the target 2D photo to obtain the facial area.
  • the positioning point acquisition unit is specifically configured to identify the facial region based on a facial feature recognition algorithm, so as to obtain the facial feature positioning point.
  • the first plane midline acquisition unit is specifically configured to perform a fitting process on the facial feature positioning points based on a fitting algorithm, so as to obtain the position of the first plane midline.
  • the first tooth midline obtaining unit is specifically configured to obtain the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw, and use the midline as the the first tooth midline to obtain the position of the first tooth midline.
  • the second acquisition module includes:
  • a three-dimensional jaw model acquisition unit configured to acquire the three-dimensional jaw model of the target object including crowns and gingiva;
  • the second size value acquisition unit is used to use the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the mandible as the second tooth in the three-dimensional dental model position of the midline, and obtain the corresponding second dimension value of the central incisor.
  • the second acquisition module includes:
  • a three-dimensional dental model acquisition unit which acquires the three-dimensional dental model of the target object including the crown and the gingiva;
  • the second size value acquisition unit is used to use the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the mandible as the second tooth in the three-dimensional dental model position of the midline, and obtain a corresponding second size value of the central incisor based on the preset size value.
  • mapping relationship establishment module includes:
  • a first size value acquisition unit configured to acquire the pixel value of the target tooth in the target two-dimensional photo based on an image recognition algorithm, and use the pixel value as the first size value
  • a mapping relationship establishing unit configured to establish a mapping relationship between the first size value and the second size value based on the first size value and the second size value corresponding to the target tooth.
  • mapping relationship establishment module includes:
  • a first size value acquisition unit configured to acquire based on a preset scale corresponding to the target two-dimensional photo The scale value of the target tooth in the target two-dimensional photo, and use the scale value as the first size value;
  • a mapping relationship establishing unit configured to establish a mapping relationship between the first size value and the second size value based on the first size value and the second size value corresponding to the target tooth.
  • the third acquisition module includes:
  • a relative position determining unit configured to obtain the relative position of the second mid-plane relative to the second tooth mid-line based on the second deviation value
  • a second midline determining unit configured to obtain a position of the second midline in the three-dimensional dental model based on the relative position.
  • a system for determining the position of orthodontic treatment includes:
  • a mid-face acquisition module configured to obtain the position of the mid-face of the target object in the three-dimensional dental model by using the above-mentioned system for determining the mid-face midline of the three-dimensional dental model;
  • a coordinate system establishment module configured to establish a three-dimensional coordinate system in the three-dimensional dental model based on the position of the mid-surface;
  • a treatment position determining module configured to determine the treatment position corresponding to each tooth of the target object based on the three-dimensional coordinate system and the position of the centerline of the target object's teeth in the three-dimensional dental model.
  • a manufacturing system for a dental appliance comprising:
  • the orthodontic position acquisition module is used to obtain the target orthodontic position of the teeth of the target object by using the above-mentioned system for determining the orthodontic position;
  • the appliance manufacturing module is configured to determine the shape of the appliance of the target object based on the target location for orthodontic treatment, so as to manufacture the appliance of the corresponding shape.
  • an electronic device which includes a memory, a processor, and a computer program stored on the memory and used to run on the processor.
  • the processor executes the computer program, it realizes the above-mentioned three-dimensional tooth and jaw model.
  • a computer-readable storage medium on which a computer program is stored, When the computer program is executed by the processor, it realizes the above-mentioned method for determining the midface midline of the three-dimensional dental model, or the above-mentioned method for determining the position of orthodontic treatment, or the above-mentioned method for manufacturing a dental appliance.
  • the method, manufacturing method and system for determining the centerline of the face and the orthodontic position of the present invention can automatically obtain the position of the first centerline of the target object in the target two-dimensional photo and the position of the first tooth centerline, according to the target object’s target
  • the first size value of the tooth in the target two-dimensional photo, and the second size value of the target tooth in the three-dimensional jaw model establish a mapping relationship between the first size value and the second size value, and then according to the first surface midline
  • the second deviation value between the second face midline and the second tooth midline of the target object in the three-dimensional jaw model is obtained from the first deviation value between the first tooth midline and the first tooth midline, and finally according to the second deviation value of the second tooth midline in the three-dimensional dental jaw model
  • the position and the second deviation value obtain the position of the second facial midline in the three-dimensional jaw model; the position of the facial midline is automatically obtained in the three-dimensional dental model, and the accuracy and efficiency of determining the facial midline are improved.
  • Yet another aspect of the present application provides a computer-implemented method for determining the position of the midline, which includes: acquiring a photo of the patient's frontal face with teeth; performing facial feature point recognition on the photo of the patient's frontal face with teeth; Generate a mid-face line based on the corresponding facial feature points; generate a mid-dental line based on the intralip region image in the patient's frontal tooth-showing photo; identify the iris region based on the identified corresponding facial feature points; and based on a given iris diameter size , the number of pixels of the identified diameter of the iris area and the pixel number deviation between the mid-surface line and the mid-dental line, and calculate the actual deviation between the mid-plane midline and the mid-dental line.
  • the method for determining the mid-face position further includes: performing facial region detection on the patient's frontal tooth-showing photo, and the facial feature point recognition is performed on the detected facial region image.
  • the facial region detection adopts one of the following methods: a traditional machine learning method and a deep neural network-based method.
  • the facial feature point detection adopts one of the following methods: a model-based method, a cascaded shape regression method, and a deep neural network-based method.
  • the mid-plane is obtained by fitting based on the corresponding facial feature points.
  • the method for determining the mid-face position further includes: identifying the inner lip area based on the identified corresponding facial feature points; and performing image processing on the inner lip area image to generate the Midline.
  • the method for determining the position of the mid-plane further includes: identifying the eye region based on the identified corresponding facial feature points; performing edge extraction on the eye region image; and determining based on the extracted edge the iris area.
  • the method for determining the centerline position of the plane further includes: performing Hough transform based on the extracted edge to determine the iris area.
  • the method for determining the position of the midplane further includes: obtaining a three-dimensional digital model of the patient's dentition; determining the position of the mid-dentine on the three-dimensional digital model of the dentition; and based on the The actual deviation between the plane midline and the tooth midline and the position of the midline on the three-dimensional digital model of the dentition determine the position of the mid-plane on the three-dimensional digital model of the dentition.
  • Yet another aspect of the present application provides a computer system for determining the position of the centerline of a plane, which includes a storage device and a processor, the storage device stores a computer program, and when it is executed by the processor, it will execute the The method of determining the position of the centerline of the surface is described.
  • Fig. 1 is the first schematic flow chart of the method for determining the midface midline in the three-dimensional dental model provided by Embodiment 1 of the present invention
  • Fig. 2 is the second schematic flow chart of the method for determining the midface midline of the three-dimensional dental model provided by Embodiment 1 of the present invention
  • FIG. 3 is a schematic diagram of facial feature positioning points in the method for determining the midface midline in the three-dimensional dental and jaw model provided by Embodiment 1 of the present invention
  • Fig. 4 is the target in the determination method of the midface midline in the three-dimensional dental model provided by Example 1 of the present invention Schematic diagram of the position of the first surface midline and the tooth midline in the 2D photograph;
  • Example 5 is a schematic diagram of a three-dimensional dental model in the method for determining the midface midline of the three-dimensional dental model provided in Example 1 of the present invention
  • Fig. 6 is a schematic flow chart of the method for determining the position of orthodontics provided by Embodiment 2 of the present invention.
  • Fig. 7 is a schematic flow chart of the manufacturing method of the dental appliance provided by Embodiment 3 of the present invention.
  • Fig. 8 is a schematic structural diagram of a system for determining the midface midline of the three-dimensional dental model provided by Embodiment 4 of the present invention.
  • Fig. 9 is a schematic structural diagram of a system for determining the position of orthodontic teeth provided by Embodiment 5 of the present invention.
  • FIG. 10 is a schematic structural diagram of an electronic device provided by Embodiment 7 of the present invention.
  • Fig. 11 is a schematic flow chart of a method for determining the position of a plane midline in another embodiment of the present application.
  • Fig. 12 is an example of a smiling photo of the patient's positive face shown in an interface of a computer program for determining the position of the mid-face in an embodiment of the present application;
  • Figure 13A schematically shows the distribution of 68 facial feature points
  • Figure 13B schematically shows the distribution of 51 facial feature points
  • Fig. 14 is an example of a patient's frontal smile photo shown by an interface of the computer program and the midface and midline generated based on it;
  • Fig. 15 is a three-dimensional digital model of the teeth and jaws in which the mid-face and mid-dental lines are marked in an example displayed on an interface of the computer program.
  • This embodiment provides a method for determining the mid-face midline in a three-dimensional dental model, as shown in Figure 1, the determination method includes:
  • the two-dimensional photo of the target includes the face of the target object showing the target teeth.
  • the first tooth midline is the tooth midline of the target object in the target two-dimensional photo
  • the first face midline is the target object's face midline in the target two-dimensional photo
  • the second tooth midline is the target object's tooth midline in the three-dimensional jaw model
  • the second midline is the midline of the target object in the 3D dental model.
  • the second size value is the actual physical size value of the target tooth of the target object, which can be obtained through a three-dimensional jaw model, or through statistical laws of teeth.
  • the method for determining the centerline of the three-dimensional dental model in this embodiment can automatically obtain the position of the first centerline of the target object in the target two-dimensional photo and the position of the first tooth centerline, according to the position of the target tooth of the target object in the target
  • the first size value in the two-dimensional photo, and the second size value of the target tooth in the three-dimensional dental model establish a mapping relationship between the first size value and the second size value, and then according to the first surface midline and the first
  • the first deviation value between the tooth midlines obtains the second deviation value between the second midline of the target object in the three-dimensional dental model and the second tooth midline, and finally according to the position of the second tooth midline in the three-dimensional dental model and the second deviation value
  • the second deviation value obtains the position of the second facial midline in the three-dimensional dental model; realizes the automatic determination of the position of the facial midline in the three-dimensional dental model, and improves the accuracy and efficiency of determining the facial midline.
  • step S101 includes:
  • the target tooth includes at least one of the two maxillary central incisors, or at least one of the two mandibular central incisors.
  • the target two-dimensional photo may be a frontal smiling photo showing the target teeth.
  • the central incisor is commonly known as the incisor, and the midline of the tooth can also be called the midline of the dentition.
  • the midline of the tooth is an imaginary line passing between the two upper or lower central incisors.
  • step S1012 image recognition is performed on the target two-dimensional photo based on the face region recognition algorithm to obtain the face region.
  • Image recognition is performed on the target two-dimensional photo through the face area recognition algorithm, and whether there is a face in the target two-dimensional photo is detected, and the face area is marked.
  • the facial area recognition algorithm can be a traditional machine learning method, such as Haar Cascade (Haar cascade), Eigenfaces (eigenface), Fisherfaces (face recognizer), etc., or a method based on a convolutional neural network.
  • the Haar Cascade algorithm is an object detection algorithm for locating objects on an image. Based on machine learning, the algorithm learns from a large number of positive samples and negative samples. Anything but positive and negative images, train a cascade function from a large number of positive and negative images, and then use it to detect objects in other images.
  • the Eigenfaces algorithm can determine a set of "normalized facial components" through statistical analysis of a large number of facial images. Facial features are assigned constant values because this algorithm does not use digital pictures, but uses statistical databases. Any facial image is of these values. Combinations of different percentages. It uses the principal component analysis (Principal Component Analysis, PAC) method to process the high-dimensional facial data into low-dimensional data, and then performs data analysis and processing to obtain the recognition results, and uses the PCA method to reduce the dimensionality of the original data to obtain the Principal component information, thus realizing the method of face recognition.
  • Principal component analysis Principal Component Analysis
  • the Fisherfaces algorithm is an eigenface-based method that finds the feature that maximizes the variance in the data Eigenface is the name of a set of feature vectors used in computer vision problems of face recognition, and Eigenfaces is based on PCA (Principal Component Analysis).
  • the facial region is specifically identified based on the facial feature recognition algorithm to obtain facial feature anchor points.
  • the facial feature anchor points may also be referred to as facial key feature points.
  • Facial feature recognition algorithms can be model-based ASM (Active Shape Model, active shape model) and AAM (Active Appearnce Model, active appearance model), or an algorithm based on cascade shape regression, or an algorithm based on deep learning.
  • ASM Active Shape Model, active shape model
  • AAM Active Appearnce Model, active appearance model
  • ASM is a variable model that models the shape in the image (such as facial feature anchor points).
  • the obtained shape model can be used to analyze new shapes, such as fitting the model to obtain new shapes, and can also be used to generate shapes. Such as searching for a shape in a given image.
  • AAM is an image segmentation algorithm based on the active appearance model. This method can be divided into two parts.
  • the first part is the training part, which is to construct an active appearance model of a thing. This part requires a training set, and its function is to let the program remember The shape feature and appearance feature of the image that needs to be cut;
  • the second part is the image segmentation stage, this part is to search for the features about the object collected in the training set just now in the image that needs to be segmented, and find the object image similar to the training set contour and appearance, and then segment it from the entire image.
  • FIG. 3 is a schematic diagram of facial feature positioning points provided in this embodiment.
  • facial feature anchor points 1-17 are facial feature anchor points corresponding to the outer contour of the face
  • facial feature anchor points 18-27 are facial feature anchor points corresponding to two eyebrows
  • facial feature anchor points 28-36 The facial feature anchor points corresponding to the nose bridge
  • the facial feature anchor points 37-48 are the facial feature anchor points corresponding to the two eyes
  • the facial feature anchor points 49-68 are the facial feature anchor points corresponding to the lips.
  • step S1014 the facial feature positioning points are specifically fitted based on the fitting algorithm to obtain the position of the first face midline.
  • Fitting algorithms include, but are not limited to, the least square method and the Hough transform method.
  • the basic principle of the Hough transform is to transform the curves (including straight lines) in the image space into the parameter space, and determine the description parameters of the curve by detecting the extreme points in the parameter space, so as to extract the regular curves in the image.
  • the facial feature anchor points are symmetrically distributed on both sides of the first plane midline with the first plane midline as the axis of symmetry.
  • step S1015 specifically, the position of the first tooth midline is obtained by obtaining the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw, and using the midline as the first tooth midline.
  • the midline of the first tooth can be the midline between the two central incisors of the upper jaw, or the midline between the two central incisors of the lower jaw.
  • the boundary line of the central incisor adjacent to another central incisor in the same jaw direction is called the mesial boundary line, and the mesial boundary line is used as the first tooth midline.
  • the boundary line of the central incisor away from the other central incisor in the same jaw direction is called the distal boundary line.
  • the acquisition of the first tooth midline may be based on a traditional image recognition method, such as edge detection; it may also be based on a deep learning method, such as a convolutional neural network.
  • the first size value of the corresponding central incisor in the target two-dimensional photo can be calculated according to the mesial borderline and the distal borderline.
  • Fig. 4 is a schematic diagram of the positions of the first midline of the face and the midline of the teeth in the target two-dimensional photo provided by this embodiment.
  • Fig. 4 shows the positions of the midline of the first tooth and the midline of the first face, and the position of the midline of the first tooth is The midline of the two central incisors of the upper jaw.
  • the method for determining the mid-face midline of the three-dimensional dental model in this embodiment automatically obtains the position of the first mid-face midline and the position of the first tooth midline of the target object in the target two-dimensional photo, and according to the position of the target object
  • the first size value of the target tooth in the target two-dimensional photo, and the second size value of the target tooth in the three-dimensional jaw model establish a mapping relationship between the first size value and the second size value, and then according to the first surface
  • the first deviation value between the center line and the first tooth center line obtains the second deviation value between the second surface center line of the target object in the three-dimensional dental model and the second tooth center line, and finally according to the second tooth center line in the three-dimensional dental model
  • the position of the second deviation value and the position of the second facial midline in the three-dimensional dental model are realized; the position of the facial midline is automatically obtained in the three-dimensional dental model, Improved accuracy and efficiency in determining surface centerlines.
  • step S1013 after the above step S1013, it also includes:
  • S10131 Determine whether the face of the target object in the target two-dimensional photo is tilted based on the position of the facial feature anchor point.
  • step S10132 If yes, execute step S10132; if not, execute step S1014.
  • Step S1014 is executed after the rotation correction operation is performed.
  • the rotation correction can be performed based on the facial feature anchor points obtained above.
  • the inner corners of the left and right eyes are connected The center is the origin of the coordinates, establish the coordinate system of the face image, rotate the image so that the inner corners of the two eyes remain horizontal, and then obtain the rotation matrix, so as to realize the rotation correction operation of the facial feature positioning points.
  • the method for determining the midface midline of the three-dimensional dental model in this embodiment performs a rotation correction operation on the facial feature positioning points, so that the facial area of the target object is no longer tilted, so that the face in the target two-dimensional photo remains correct, which is convenient for follow-up Mark the position of the midline of the first tooth and the position of the midline of the first surface in the target 2D photo.
  • step S1014 includes:
  • the facial feature key points are selected from the facial feature anchor points, and the position of the first surface midline is obtained based on the facial feature key points.
  • the facial feature anchor points include eyes, eyebrows, nose bridge, lips, and facial contour corresponding anchor points; facial feature key points include at least three of the eyebrow center point, eye corner point, nose tip point, nose base point, and midpoint. indivual.
  • facial feature anchor points 20 and 25 are facial feature key points corresponding to the brow point; facial feature anchor points 37 and 46 are facial feature key points corresponding to outer corner points; facial feature anchor points 40 and 43 are inner The facial feature key point corresponding to the canthus point; the facial feature anchor point 31 is the facial feature key point corresponding to the tip of the nose point; point as the midpoint of the person.
  • At least three facial feature key points are selected, and the facial feature anchor points are fitted based on a fitting algorithm to obtain the position of the first surface midline.
  • step S102 includes:
  • the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the mandible is used as the position of the midline of the second tooth, and the second size of the corresponding central incisor is obtained value.
  • a three-dimensional dental model including the crown and gums of the target object is established.
  • the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw is used as the position of the midline of the second tooth.
  • the distribution of each tooth of the target object and the size of each tooth can be obtained according to the three-dimensional jaw model, and then the second size value of the central incisor can be obtained.
  • Fig. 5 is a schematic diagram of a three-dimensional jaw model provided by this embodiment, and Fig. 5 shows the positions of the second tooth midline and the second facial midline.
  • the method for determining the mid-face midline of the three-dimensional dental model in this embodiment realizes the automatic determination of the position of the second mid-face midline on the three-dimensional dental model, and improves the accuracy and efficiency of determining the mid-face midline.
  • step S102 includes:
  • the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the mandible is used as the position of the second tooth midline, and the corresponding said The second dimension value of the central incisor.
  • a three-dimensional dental model including the crown and gums of the target object is established.
  • the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw is used as the position of the midline of the second tooth.
  • the size information of the teeth can be obtained, for example, according to the age, gender, and general shape and size of the teeth of the target subject, the second size value of the central incisor can be obtained according to the preset size values stored in the tooth database.
  • the method for determining the mid-face midline of the three-dimensional dental model in this embodiment realizes the automatic determination of the position of the second mid-face midline on the three-dimensional dental model, and improves the accuracy and efficiency of determining the mid-face midline.
  • step S103 includes:
  • the pixel value of the central incisor in the target two-dimensional photo is calculated based on the image recognition algorithm, and the pixel value is used as the target tooth in the target two-dimensional image.
  • the first size value in the three-dimensional photo meanwhile, the second size value of the central incisor can be obtained in the three-dimensional dental model.
  • a mapping relationship between the first size value and the second size value is established based on the corresponding first size value of the central incisor in the target two-dimensional photo and the corresponding second size value in the three-dimensional jaw model.
  • the method for determining the midface midline of the three-dimensional dental model in this embodiment realizes the automatic acquisition of the first size value of the target tooth in the target two-dimensional photo, and establishes based on the first size value and the second size value corresponding to the target tooth
  • the mapping relationship between the first size value and the second size value can automatically determine the position of the second midline in the three-dimensional dental model, which improves the accuracy and efficiency of determining the midline.
  • step S103 includes:
  • the scale value of the target tooth in the target two-dimensional photo can be obtained according to the numerical value defined by the preset scale, and the scale value is used as the first size value.
  • the method for determining the midface midline of the three-dimensional dental model in this embodiment realizes the automatic acquisition of the first size value of the target tooth in the target two-dimensional photo, and establishes based on the first size value and the second size value corresponding to the target tooth
  • the mapping relationship between the first size value and the second size value can automatically determine the position of the second midline in the three-dimensional dental model, which improves the accuracy and efficiency of determining the midline.
  • step S105 includes:
  • the first deviation value After obtaining the first deviation value between the first surface center line and the first tooth center line, the first deviation value can be converted into a corresponding second deviation value according to the mapping relationship between the first size value and the second size value value, since the position of the second tooth midline has been determined, the relative position of the second facial midline relative to the second tooth midline can be obtained according to the second deviation value, and then based on the relative position, the second facial midline can be obtained in the three-dimensional dental model s position.
  • the method for determining the mid-face midline of the three-dimensional dental model in this embodiment through the second deviation value between the second mid-face midline of the target object and the second tooth midline, and the mapping relationship between the first size value and the second size value, and the position of the second tooth midline in the three-dimensional dental model, and automatically obtain the position of the second facial midline in the three-dimensional dental model, which improves the accuracy and efficiency of determining the facial midline.
  • This embodiment provides a method for determining the position of the orthodontic treatment. As shown in FIG. 6, the method for determining the position of the orthodontic treatment includes:
  • S203 Determine the orthodontic position corresponding to each tooth of the target object based on the three-dimensional coordinate system and the position of the second tooth midline of the target object in the three-dimensional jaw model.
  • the method for determining the orthodontic position in this embodiment is based on the method for determining the mid-face midline of the three-dimensional dental model in Example 1, and automatically obtains the position of the target object’s mid-face in the three-dimensional dental model, and then based on the second
  • the position of the centerline of the surface establishes a three-dimensional coordinate system, and according to the position of the centerline of the second tooth of the target object in the three-dimensional dental model in the three-dimensional coordinate system, determine the corresponding orthodontic position of each tooth of the target object, for example, a certain tooth relative to the first tooth Whether the median line of two surfaces moves to the left or to the right.
  • the method for determining the orthodontic position of this embodiment can quickly determine the orthodontic position corresponding to each tooth of the target object, and then flexibly design and adjust the orthodontic plan.
  • This embodiment provides a manufacturing method of a dental appliance, as shown in Figure 7, the manufacturing method of a dental appliance includes:
  • the obtained data of the target orthodontic position of the teeth can be transmitted to the manufacturing equipment of the orthodontic appliance, and the orthodontic appliance manufacturing equipment can form a mold (such as a positive mold) of the orthodontic appliance according to the data according to a general engineering method, thereby A dental appliance with a corresponding shape is produced from the mold.
  • a mold such as a positive mold
  • the appliance manufacturing equipment can also use 3D printing technology to form a male mold of the dental appliance according to the data of the target orthodontic position of the teeth, and further manufacture a dental appliance with a corresponding shape from the male mold.
  • the manufacturing equipment of the orthodontic appliance can also use the three-dimensional printing technology to determine the data of the corresponding orthodontic appliance according to the data of the target orthodontic position of the teeth, and directly form the orthodontic appliance according to the data of the orthodontic appliance.
  • the dental appliance is manufactured from a flexible, transparent polymer material, thereby forming an invisible appliance without brackets.
  • the manufacturing method of the orthodontic appliance of this embodiment can quickly and accurately determine the shape of the orthodontic appliance of the target subject according to the target orthodontic position of the teeth of the target subject, and then manufacture the orthodontic appliance with a corresponding shape.
  • This embodiment provides a system for determining the mid-face midline of a three-dimensional dental model.
  • the system for determining the mid-face midline of the three-dimensional dental model includes:
  • the first acquisition module 1 is used to acquire the position of the first midline of the target object in the target two-dimensional photo and the position of the first tooth midline;
  • the two-dimensional photo of the target includes the face of the target subject showing the target teeth
  • the second acquisition module 2 is configured to acquire the position of the second tooth midline of the target object in the three-dimensional jaw model based on the target tooth, and acquire the second size value of the target tooth in the three-dimensional jaw model;
  • a mapping relationship establishment module 3 configured to obtain the first size value of the target tooth in the target two-dimensional photo, and establish a mapping relationship between the first size value and the second size value;
  • the first deviation value acquisition module 4 is used to obtain the first deviation value between the first plane center line and the first tooth center line based on the position of the first plane center line and the position of the first tooth center line;
  • the second deviation value acquisition module 5 is used to obtain a second deviation value between the second midline of the target object in the three-dimensional dental model and the second tooth midline based on the mapping relationship and the first deviation value;
  • the third acquiring module 6 is configured to obtain the position of the second facial midline in the three-dimensional dental model based on the position of the second tooth midline in the three-dimensional dental model and the second deviation value.
  • the first acquisition module 1 includes a two-dimensional photo acquisition unit 11, which is used to acquire a target two-dimensional photo of a target object; a facial area acquisition unit 12, which is used to acquire a facial area in the target two-dimensional photo
  • An anchor point acquisition unit 13 is used to obtain the facial feature anchor point of the face region;
  • the first midline acquisition unit 14 is used to obtain the position of the first midline based on the facial feature anchor point;
  • the first tooth midline acquisition unit 15 uses Obtaining the position of the midline of the first tooth based on the position of the target tooth in the target two-dimensional photo; wherein the target tooth includes at least one of the two maxillary central incisors, or at least one of the two mandibular central incisors.
  • the first acquisition module 1 further includes a face tilt judging unit 16, configured to judge whether the face of the target object in the target two-dimensional photo is tilted based on the position of the facial feature anchor point; the correction unit 17, When the facial inclination judging unit 16 judges that the face of the target object in the target two-dimensional photo has an inclination, perform a rotation correction operation on the facial feature anchor point, and run the first midline acquisition unit 14; if the facial inclination judging unit 16 When it is judged that the face of the target object in the target two-dimensional photo does not have an inclination, the first midline acquisition unit 14 is executed.
  • a face tilt judging unit 16 configured to judge whether the face of the target object in the target two-dimensional photo is tilted based on the position of the facial feature anchor point
  • the correction unit 17 When the facial inclination judging unit 16 judges that the face of the target object in the target two-dimensional photo has an inclination, perform a rotation correction operation on the facial feature anchor point, and run the first midline acquisition unit
  • the first midline acquisition unit 14 is specifically used to Select the key points of facial features from the anchor points, and obtain the position of the midline of the first face based on the key points of facial features; wherein, the anchor points of facial features include the anchor points corresponding to the eyes, eyebrows, bridge of the nose, lips, and the outer contour of the face; The feature key points include at least three of the brow center point, eye corner point, nose tip point, nose base point, and human midpoint.
  • the facial region acquisition unit 12 is specifically configured to perform image recognition on the target two-dimensional photo based on a facial region recognition algorithm to obtain the facial region.
  • the anchor point obtaining unit 13 is specifically configured to identify the facial region based on a facial feature recognition algorithm, so as to obtain the facial feature anchor point.
  • the first midline obtaining unit 14 is specifically configured to perform fitting processing on facial feature anchor points based on a fitting algorithm to obtain the position of the first midline.
  • the first tooth midline acquisition unit 15 is specifically used to obtain the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw, and use the midline as the first tooth midline , to get the position of the midline of the first tooth.
  • the second acquisition module 2 includes a three-dimensional jaw model acquisition unit 21, which is used to acquire a three-dimensional jaw model including a crown and a gingiva of the target object; a second dimension value acquisition unit 22, which is used to In the three-dimensional dental model, the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw is used as the position of the second tooth midline, and the corresponding second size value of the central incisor is obtained.
  • a three-dimensional jaw model acquisition unit 21 which is used to acquire a three-dimensional jaw model including a crown and a gingiva of the target object
  • a second dimension value acquisition unit 22 which is used to In the three-dimensional dental model, the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw is used as the position of the second tooth midline, and the corresponding second size value of the central incisor is obtained.
  • the second acquisition module 2 includes a three-dimensional jaw model acquisition unit 21, which is used to acquire a three-dimensional jaw model including a crown and a gingiva of the target object; a second dimension value acquisition unit 22, which is used to In the three-dimensional dental model, the midline between the two central incisors of the upper jaw or the midline between the two central incisors of the lower jaw is used as the position of the second tooth midline, and the corresponding first central incisor is obtained based on the preset size value.
  • Two dimension values Two dimension values.
  • the mapping relationship establishment module 3 includes a first size value acquisition unit 31, which is used to acquire the first size value of the target tooth in the target two-dimensional photo based on an image recognition algorithm; the mapping relationship establishment unit 32, It is used to establish a mapping relationship between the first size value and the second size value based on the first size value and the second size value corresponding to the target tooth.
  • the mapping relationship establishment module 3 includes a first size value acquisition unit 31, for obtaining the first size value of the target tooth in the target two-dimensional photo based on the preset scale corresponding to the target two-dimensional photo; the mapping relationship establishment unit 32, for based on the first size value and the second size corresponding to the target tooth value establishes a mapping relationship between the first size value and the second size value.
  • the third acquisition module 6 includes a relative position determination unit 61, configured to obtain the relative position of the second midline relative to the second tooth midline based on the second deviation value; the second midline determination unit 62 , used to obtain the position of the second midline in the 3D jaw model based on the relative position.
  • the working principle of the system for determining the midface midline of the 3D dental model in this embodiment is the same as that of the method for determining the midface midline of the 3D dental model in Embodiment 1, and will not be repeated here.
  • the system for determining the centerline of the three-dimensional dental model in this embodiment can automatically obtain the position of the first centerline of the target object in the target two-dimensional photo and the position of the first tooth centerline, according to the position of the target tooth of the target object
  • the first size value in the target two-dimensional photo, and the second size value of the target tooth in the three-dimensional dental model establish a mapping relationship between the first size value and the second size value, and then according to the first surface midline and the second
  • a first deviation value between the midlines of the teeth obtains a second deviation value between the second midline of the target object in the three-dimensional jaw model and the second tooth midline, and finally according to the position and the position of the second tooth midline in the three-dimensional jaw model
  • the second deviation value obtains the position of the second facial midline in the three-dimensional dental model; it realizes the automatic determination of the position of the facial midline in the three-dimensional dental model, and improves the accuracy and efficiency of determining the facial midline.
  • This embodiment provides a system for determining the position of orthodontic treatment.
  • the system for determining the position of orthodontic treatment includes:
  • the mid-face acquisition module 7 is used to obtain the position of the mid-face of the target object in the three-dimensional dental model by using the determination system of the mid-face mid-line in the three-dimensional dental and maxillary model in embodiment 4;
  • a coordinate system establishment module 8 configured to establish a three-dimensional coordinate system in the three-dimensional dental model based on the position of the midline;
  • the treatment position determination module 9 is configured to determine the treatment position corresponding to each tooth of the target object based on the three-dimensional coordinate system and the position of the tooth midline of the target object in the three-dimensional dental model.
  • the working principle of the system for determining the orthodontic position in this embodiment is the same as that of the method for determining the orthodontic position in Embodiment 2, and will not be repeated here.
  • the system for determining the orthodontic position of this embodiment based on the system for determining the midface midline of the three-dimensional dental model in Embodiment 4, automatically acquires the position of the target object's midface in the three-dimensional dental model, and can quickly determine The orthodontic position corresponding to each tooth of the target object, and then flexibly design and adjust the orthodontic plan.
  • This embodiment provides a manufacturing system for a dental appliance, and the manufacturing system for a dental appliance includes:
  • the orthodontic position acquisition module is used to obtain the target orthodontic position of the teeth of the target object by using the system for determining the orthodontic position in Embodiment 5;
  • the appliance manufacturing module is configured to determine the shape of the appliance of the target object based on the target location for orthodontic treatment, so as to manufacture the appliance of the corresponding shape.
  • the working principle of the manufacturing system of the dental appliance in this embodiment is the same as the working principle of the manufacturing method of the dental appliance in Embodiment 3, and will not be repeated here.
  • the manufacturing system of the orthodontic device of this embodiment determines the target orthodontic position of the teeth of the target object based on the system for determining the orthodontic position in Embodiment 5, and quickly and accurately determines the shape of the orthodontic appliance of the target object, and then A dental aligner with a corresponding shape is manufactured.
  • FIG. 10 is a schematic structural diagram of the electronic device provided by this embodiment.
  • the electronic device includes a memory, a processor, and a computer program stored on the memory and operable on the processor.
  • the processor executes the computer program.
  • the method for determining the midface midline of the three-dimensional dental model in the above-mentioned embodiment 1, or the method for determining the position of the orthodontic treatment in the above-mentioned embodiment 2, or the manufacturing method of the dental appliance in the above-mentioned embodiment 3 is realized.
  • the electronic device 70 shown in FIG. 10 is only an example, and should not limit the functions and scope of use of this embodiment of the present invention. As shown in FIG.
  • the electronic device 70 may be in the form of a general-purpose computing device, for example, it may be a server device.
  • the components of the electronic device 70 may include, but are not limited to: the above-mentioned at least one processor 71, the above-mentioned at least one memory 72, and connections to different system components (including the memory 72 and the processor 71). bus 73.
  • the bus 73 includes a data bus, an address bus and a control bus.
  • the memory 72 may include a volatile memory, such as a random access memory (RAM) 721 and/or a cache memory 722 , and may further include a read only memory (ROM) 723 .
  • RAM random access memory
  • ROM read only memory
  • Memory 72 may also include a program tool 725 (or utility) having a set (at least one) of program modules 724, such program modules 724 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include the realization of the network environment.
  • program tool 725 or utility
  • program modules 724 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include the realization of the network environment.
  • the processor 71 executes various functional applications and data processing by running the computer program stored in the memory 72, such as the determination method of the mid-surface midline in the three-dimensional dental model in the above-mentioned embodiment 1, or the tooth in the above-mentioned embodiment 2.
  • Electronic device 70 may also communicate with one or more external devices 74 . Such communication may occur through input/output (I/O) interface 75 . Also, the model-generating electronic device 70 can also communicate with one or more networks (eg, a local area network (LAN), a wide area network (WAN) and/or a public network, such as the Internet) via a network adapter 76 . As shown in FIG. 10 , the network adapter 76 communicates with other modules of the electronic device 70 through the bus 73 .
  • networks eg, a local area network (LAN), a wide area network (WAN) and/or a public network, such as the Internet
  • electronic device 70 may be used in conjunction with electronic device 70, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (array of disks) systems, tape drives, and data backup storage systems.
  • This embodiment provides a computer-readable storage medium, on which a computer program is stored, and the calculation When the computer program is executed by the processor, the method for determining the midface midline of the three-dimensional dental and jaw model in the above-mentioned embodiment 1, or the method for determining the position of the orthodontic treatment in the above-mentioned embodiment 2, or the dental appliance in the above-mentioned embodiment 3 manufacturing method.
  • the readable storage medium may more specifically include but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device or any of the above-mentioned the right combination.
  • the present invention can also be implemented in the form of a program product, which includes program code, and when the program product runs on the terminal device, the program code is used to make the terminal device execute the implementation of the above-mentioned embodiment 1.
  • program code for executing the present invention may be written in any combination of one or more programming languages, and the program code may be completely executed on the user equipment, partially executed on the user equipment, or used as an independent software Package execution, partly on the user device and partly on the remote device, or entirely on the remote device.
  • Yet another aspect of the present application provides a method for determining the position of the midline of the face, which is to detect the feature points of the patient's face with teeth, and to obtain the midline of the face and the midline of the teeth based on the detected specific feature points, and then, based on the size of the iris Determines the actual distance between the surface centerline and the tooth centerline.
  • the method for determining the position of the plane midline in one embodiment of the present application will be described in detail below with reference to the accompanying drawings.
  • FIG. 11 is a schematic flow chart of a method 100 for determining a plane centerline position in an embodiment of the present application.
  • the method 100 for determining the centerline position of a plane is executed by a computer.
  • a computer system for determining the position of the mid-plane which includes a storage device and a processor, wherein the storage device stores a computer program, and when it is executed by the processor, it will execute The method 100 for determining the position of the centerline of a plane.
  • the patient's frontal toothy photo may be a frontal smiling photo.
  • FIG. 12 is a frontal smiling photo of a patient shown in an example of an interface of a computer program for determining the mid-face position in an embodiment of the present application.
  • the eyes of the patient's photo shown in Figure 12 are covered.
  • face region detection is performed on the photo of the patient's frontal face showing teeth.
  • facial region detection can be performed on it.
  • it can be used to filter out non-standard photos of smiling faces, for example, photos with incomplete facial features, face twists, and moiré patterns.
  • subsequent calculations can be based on detected facial regions rather than the entire frontal smile photo to improve computational efficiency and reliability.
  • a machine learning method can be used for facial area detection. If a human face cannot be detected, it is determined that the image is irregular, for example, facial features are incomplete, the face is twisted, and the like.
  • the recognition of the moiré pattern may be performed after the human face is detected. First, samples with moiré and no moiré in the face image are obtained, and then a model capable of recognizing moiré is trained using the SVM (support vector machine) method, and the model is used to identify moiré on the face image.
  • SVM support vector machine
  • traditional machine learning methods can be used for face region detection, for example, Haar Cascade, Eigenfaces, Fisherfaces, etc.
  • a method based on a deep neural network for example, a method based on a convolutional neural network, may also be used for face region detection.
  • the user may be prompted to re-enter the front face smile photo.
  • facial feature point detection is performed on the face area.
  • facial feature point detection can be performed on the face area picture.
  • Facial landmarks identify special locations on the face, such as eyes, eyebrows, bridge of the nose, lips, etc.
  • a model-based method can be used for feature point detection, for example, ASM (Active Shape Model) and AAM (Active Appearance Model).
  • ASM Active Shape Model
  • AAM Active Appearance Model
  • the method of cascading shape regression may also be used for feature point detection.
  • a method based on a deep neural network can be used for feature point detection.
  • FIG. 13A schematically shows the distribution of 68 facial feature points.
  • FIG. 13B which schematically shows the distribution of 51 facial feature points.
  • the face midline is fitted based on the corresponding facial feature points.
  • the facial midline and dental midline can be fitted based on the corresponding facial feature points.
  • the centerline of the face can be fitted based on the feature points substantially located on the symmetrical centerline of the face (ie, the centerline of the face), for example, the center of the eyebrows, the point of the tip of the nose, the point of the base of the nose, the midpoint of the person, and the like.
  • the least square method can be used to fit the centerline of the surface.
  • the lip region can be located based on the corresponding feature points, and then, based on the exposed teeth in the lip region picture, the two central incisors can be recognized through image processing (image segmentation), and then The contour information determines the gap between the two central incisors and uses it as the midline.
  • frontal smile photos may have unexposed teeth, too few exposed teeth, or missing teeth, which will affect the identification of the midline. Therefore, before identifying the midline, these situations can be identified and It sifts out.
  • a method based on deep learning can be used to classify lip submaps to screen out the above-mentioned unqualified cases.
  • other machine learning methods can also be used to perform this operation, for example, decision tree, SVM, AdaBoost, etc.
  • FIG. 14 is an example of a frontal smile photo of a patient and the midface midline and dental midline generated based on it as an example of an interface of the computer program.
  • the eyes of the patient photos are masked.
  • the inventors of the present application found that the iris diameter of different people does not vary much, which is about 12.3 ⁇ 0.7mm.
  • 12.3mm can be taken as the iris diameter, and the actual size of a single pixel can be estimated based on this size, and then the actual deviation between the mid-plane midline and the mid-dental line can be calculated.
  • the actual deviation of the facial midline and dental midline is calculated based on the given iris size.
  • the eye region pictures can be obtained based on the corresponding facial feature points first, then the edges are extracted and optimized, and the possible iris regions are detected by Hough transform, and finally, these possible regions are screened to obtain Determine the final iris area. Then, based on the selected iris diameter and the number of pixels of the iris area diameter, the physical size of each pixel is calculated. Finally, based on the pixel number deviation and the physical size of the pixels, the physical deviation between the surface center line and the tooth center line is calculated.
  • the position of the midline of the face refers to the position projected along the direction of the frontal face.
  • Midline alignment ie alignment of the tooth midline with the facial midline
  • the three-dimensional digital model of the teeth and jaws can be obtained by intraoral scanning, or by scanning impressions or tooth solid models.
  • the computer can mark the surface on the 3D digital model of the jaw when the position of the midline is known.
  • the position of the midline which is used as a reference for dental professionals when designing orthodontic treatment plans.
  • FIG. 15 is an example of an interface of the computer program showing a three-dimensional digital model of the tooth and jaw with the mid-face midline and the mid-dental line marked.
  • the midline is represented by a point located between the upper two incisors.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Dental Tools And Instruments Or Auxiliary Dental Instruments (AREA)

Abstract

本发明公开了一种面中线及牙齿矫治位置的确定方法、制造方法及系统，面中线的确定方法包括获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置；获取目标对象在三维牙颌模型中的第二牙齿中线的位置，获取目标牙齿在三维牙颌模型中的第二尺寸值；获取目标牙齿在目标二维照片中的第一尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系；得到第一面中线和第一牙齿中线之间的第一偏差值；得到目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值；基于第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置。本发明实现了自动的在三维牙颌模型得出面中线的位置，提高了确定面中线的准确性和效率。

Description

面中线及牙齿矫治位置的确定方法、制造方法及系统 技术领域
本申请涉及牙齿矫治技术领域，尤其涉及一种面中线及牙齿矫治位置的确定方法、制造方法及系统。
背景技术
基于口腔美学的概念，越来越多的人对自身的牙齿形态有了更高的追求，牙齿矫治(或称为口腔正畸)应运而生，在牙齿矫治过程中，需要根据一定的规则对每个牙齿的位置进行调整，例如以待矫治者脸部面中线的位置作为基准，根据待矫治者牙齿中线的位置与面中线的相对关系，确定牙齿相对于面中线是向左调整还向右调整。
现有的牙齿矫治过程，需要矫治设计师观察待矫治者的二维露牙照片，并结合个人经验，大致估算出面中线在牙列中的位置，然后人工在包含有待矫治者的牙冠和牙龈的三维牙颌模型中，估算出面中线在三维牙颌模型中的位置，该方法不仅过程繁琐，而且确定出的面中线的准确性较低，具有一定的随机性和误差。
发明内容
本发明要解决的技术问题是为了克服现有技术中牙齿矫治过程中，需要人工估算面中线在三维牙颌模型中的位置，面中线的确定过程繁琐，而且确定出的面中线的准确性较低，具有一定的随机性和误差的缺陷，提供一种面中线的确定及牙齿矫治方法、制造方法及系统。
本发明是通过下述技术方案来解决上述技术问题：
第一方面，提供一种三维牙颌模型中面中线的确定方法，所述确定方法包括：
获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线 的位置；
其中，所述目标二维照片包括所述目标对象露出目标牙齿的面部；
基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值；
获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系；
基于所述第一面中线的位置和所述第一牙齿中线的位置得到所述第一面中线和所述第一牙齿中线之间的第一偏差值；
基于所述映射关系和所述第一偏差值得到所述目标对象在所述三维牙颌模型中的第二面中线与所述第二牙齿中线的第二偏差值；
基于所述三维牙颌模型中的所述第二牙齿中线的位置和所述第二偏差值得到所述第二面中线在所述三维牙颌模型中的位置。
较佳地，所述获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置包括：
获取所述目标对象的所述目标二维照片；
获取所述目标二维照片中的面部区域；
获取所述面部区域的面部特征定位点；
基于所述面部特征定位点得到所述第一面中线的位置；
基于所述目标牙齿在所述目标二维照片中的位置得到所述第一牙齿中线的位置；
其中，所述目标牙齿包括上颌两颗中切牙中的至少一颗，或下颌两颗中切牙中的至少一颗。
较佳地，所述获取所述面部区域的面部特征定位点的步骤之后还包括：
基于所述面部特征定位点的位置判断所述目标二维照片中的所述目标对象的面部是否存在倾斜；
若是，则对所述面部特征定位点进行旋转校正操作，并执行所述基于所述面部特征定位点得到所述第一面中线的位置的操作；
若否，则执行所述基于所述面部特征定位点得到所述第一面中线的位置的操作。
较佳地，所述基于所述面部特征定位点得到所述第一面中线的位置的步骤包括：
在所述面部特征定位点中选取出面部特征关键点，并基于所述面部特征关键点得到所述第一面中线的位置；
其中，所述面部特征定位点包括眼睛、眉毛、鼻梁、唇部、脸部外轮廓对应的定位点；所述面部特征关键点包括眉心点、眼角点，鼻尖点、鼻底点、人中点中的至少三个。
较佳地，所述获取所述目标二维照片中的面部区域的步骤包括：
基于面部区域识别算法对所述目标二维照片进行图像识别，以得到所述面部区域；
和/或，所述获取所述面部区域的面部特征定位点的步骤包括：
基于面部特征识别算法对所述面部区域进行识别，以得到所述面部特征定位点；
和/或，所述基于所述面部特征定位点得到所述第一面中线的位置的方法包括：
基于拟合算法对所述面部特征定位点进行拟合处理，以得到所述第一面中线的位置；
和/或，所述基于所述目标牙齿在所述目标二维照片中的位置得到所述第一牙齿中线的位置的步骤包括：
获取上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线，并将所述中线作为所述第一牙齿中线，以得到所述第一牙齿中线的位置。
较佳地，所述基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值的步骤包括：
获取所述目标对象包含有牙冠和牙龈的所述三维牙颌模型；
在所述三维牙颌模型中，以上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线作为所述第二牙齿中线的位置，并得到对应的所述中切牙的第二尺寸值。
较佳地，所述基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值的步骤包括：
获取所述目标对象包含有牙冠和牙龈的所述三维牙颌模型；
在所述三维牙颌模型中，以上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线作为所述第二牙齿中线的位置，并基于预设尺寸值得到对应的所述中切牙的第二尺寸值。
较佳地，所述获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系的步骤包括；
基于图像识别算法获取所述目标牙齿在所述目标二维照片中的第一尺寸值；
基于所述目标牙齿对应的所述第一尺寸值和所述第二尺寸值建立所述第一尺寸值与所述第二尺寸值之间的映射关系。
较佳地，所述获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系的步骤包括；
基于所述目标二维照片对应的预设标尺获取所述目标牙齿在所述目标二维照片中的第一尺寸值；
基于所述目标牙齿对应的所述第一尺寸值和所述第二尺寸值建立所述第一尺寸值与所述第二尺寸值之间的映射关系。
较佳地，所述基于所述三维牙颌模型中的所述第二牙齿中线的位置和所述第二偏差值得到所述第二面中线在所述三维牙颌模型中的位置的步骤包括：
基于所述第二偏差值得到所述第二面中线相对于所述第二牙齿中线的相对位置；
基于所述相对位置得到所述第二面中线在所述三维牙颌模型中的位置。
第二方面，还提供一种牙齿矫治位置的确定方法，所述确定方法包括：
采用上述所述的三维牙颌模型中面中线的确定方法得到目标对象在三维牙颌模型中的第二面中线的位置；
基于所述第二面中线的位置在所述三维牙颌模型中建立三维坐标系；
基于所述三维坐标系和所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置确定所述目标对象每个牙齿对应的矫治位置。
第三方面，还提供一种牙齿矫治器的制造方法，所述制造方法包括：
采用上述所述的牙齿矫治位置的确定方法获得目标对象的牙齿的目标矫治位置；
基于所述目标矫治位置确定所述目标对象的牙齿矫治器的形状，以制造出具有相应形状的牙齿矫治器。
第四方面，还提供一种三维牙颌模型中面中线的确定系统，所述确定系统包括：
第一获取模块，用于获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置；
其中，所述目标二维照片包括所述目标对象露出目标牙齿的面部；
第二获取模块，用于基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值；
映射关系建立模块，用于获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系；
第一偏差值获取模块，用于基于所述第一面中线的位置和所述第一牙齿中线的位置得到所述第一面中线和所述第一牙齿中线之间的第一偏差值；
第二偏差值获取模块，用于基于所述映射关系和所述第一偏差值得到所述目标对象在所述三维牙颌模型中的第二面中线与所述第二牙齿中线的第二偏差值；
第三获取模块，用于基于所述三维牙颌模型中的所述第二牙齿中线的位置和所述第二偏差值得到所述第二面中线在所述三维牙颌模型中的位置。
较佳地，所述第一获取模块包括：
二维照片获取单元，用于获取所述目标对象的所述目标二维照片；
面部区域获取单元，用于获取所述目标二维照片中的面部区域；
定位点获取单元，用于获取所述面部区域的面部特征定位点；
第一面中线获取单元，用于基于所述面部特征定位点得到所述第一面中线的位置；
第一牙齿中线获取单元，用于基于所述目标牙齿在所述目标二维照片中的位置得到所述第一牙齿中线的位置；
其中，所述目标牙齿包括上颌两颗中切牙中的至少一颗，或下颌两颗中切牙中的至少一颗。
较佳地，所述第一获取模块还包括：
面部倾斜判断单元，用于基于所述面部特征定位点的位置判断所述目标二维照片中的所述目标对象的面部是否存在倾斜；
校正单元，用于当所述面部倾斜判断单元判断出所述目标二维照片中的所述目标对象的面部存在倾斜时，对所述面部特征定位点进行旋转校正操作，并运行所述第一面中线获取单元；
若所述面部倾斜判断单元判断出所述目标二维照片中的所述目标对象的面部不存在倾斜时，则运行所述第一面中线获取单元。
较佳地，所述第一面中线获取单元具体用于在所述面部特征定位点中选取出面部特征关键点，并基于所述面部特征关键点得到所述第一面中线的位置；
其中，所述面部特征定位点包括眼睛、眉毛、鼻梁、唇部、脸部外轮廓对应的定位点；所述面部特征关键点包括眉心点、眼角点，鼻尖点、鼻底点、人中点中的至少三个。
较佳地，所述面部区域获取单元具体用于基于面部区域识别算法对所述 目标二维照片进行图像识别，以得到所述面部区域。
较佳地，所述定位点获取单元具体用于基于面部特征识别算法对所述面部区域进行识别，以得到所述面部特征定位点。
较佳地，所述第一面中线获取单元具体用于基于拟合算法对所述面部特征定位点进行拟合处理，以得到所述第一面中线的位置。
较佳地，所述第一牙齿中线获取单元具体用于获取上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线，并将所述中线作为所述第一牙齿中线，以得到所述第一牙齿中线的位置。
较佳地，所述第二获取模块包括：
三维牙颌模型获取单元，用于获取所述目标对象包含有牙冠和牙龈的所述三维牙颌模型；
第二尺寸值获取单元，用于在所述三维牙颌模型中，以上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线作为所述第二牙齿中线的位置，并得到对应的所述中切牙的第二尺寸值。
较佳地，所述第二获取模块包括：
三维牙颌模型获取单元，获取所述目标对象包含有牙冠和牙龈的所述三维牙颌模型；
第二尺寸值获取单元，用于在所述三维牙颌模型中，以上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线作为所述第二牙齿中线的位置，并基于预设尺寸值得到对应的所述中切牙的第二尺寸值。
较佳地，所述映射关系建立模块包括：
第一尺寸值获取单元，用于基于图像识别算法获取所述目标牙齿在所述目标二维照片中的像素值，并将所述像素值作为所述第一尺寸值；
映射关系建立单元，用于基于所述目标牙齿对应的所述第一尺寸值和所述第二尺寸值建立所述第一尺寸值与所述第二尺寸值之间的映射关系。
较佳地，所述映射关系建立模块包括：
第一尺寸值获取单元，用于基于所述目标二维照片对应的预设标尺获取 所述目标牙齿在所述目标二维照片中的标尺值，并将所述标尺值作为所述第一尺寸值；
映射关系建立单元，用于基于所述目标牙齿对应的所述第一尺寸值和所述第二尺寸值建立所述第一尺寸值与所述第二尺寸值之间的映射关系。
较佳地，所述第三获取模块包括：
相对位置确定单元，用于基于所述第二偏差值得到所述第二面中线相对于所述第二牙齿中线的相对位置；
第二面中线确定单元，用于基于所述相对位置得到所述第二面中线在所述三维牙颌模型中的位置。
第五方面，还提供一种牙齿矫治位置的确定系统，所述确定系统包括：
面中线获取模块，用于采用上述所述的三维牙颌模型中面中线的确定系统得到目标对象在三维牙颌模型中的面中线的位置；
坐标系建立模块，用于基于所述面中线的位置在所述三维牙颌模型中建立三维坐标系；
矫治位置确定模块，用于基于所述三维坐标系和所述目标对象在所述三维牙颌模型中的牙齿中线的位置确定所述目标对象每个牙齿对应的矫治位置。
第六方面，还提供一种牙齿矫治器的制造系统，所述制造系统包括：
矫治位置获取模块，用于采用上述所述的牙齿矫治位置的确定系统获得目标对象的牙齿的目标矫治位置；
矫治器制造模块，用于基于所述目标矫治位置确定所述目标对象的牙齿矫治器的形状，以制造出具有相应形状的牙齿矫治器。
第七方面，还提供一种电子设备，包括存储器、处理器及存储在存储器上并用于在处理器上运行的计算机程序，所述处理器执行计算机程序时实现上述所述的三维牙颌模型中面中线的确定方法，或上述所述的牙齿矫治位置的确定方法，或上述所述的牙齿矫治器的制造方法。
第八方面，还提供一种计算机可读存储介质，其上存储有计算机程序， 所述计算机程序被处理器执行时实现上述所述的三维牙颌模型中面中线的确定方法，或上述所述的牙齿矫治位置的确定方法，或上述所述的牙齿矫治器的制造方法。
在符合本领域常识的基础上，上述各优选条件，可任意组合，即得本发明各较佳实例。
本发明的积极进步效果在于：
本发明的面中线及牙齿矫治位置的确定方法、制造方法及系统，可以自动的获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置，根据目标对象的目标牙齿在目标二维照片中的第一尺寸值，和目标牙齿在三维牙颌模型中的第二尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系，进而根据第一面中线和第一牙齿中线之间的第一偏差值得出目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值，最终根据三维牙颌模型中的第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置；实现了自动的在三维牙颌模型得出面中线的位置，提高了确定面中线的准确性和效率。
本申请的又一方面提供了一种计算机执行的确定面中线位置的方法，其包括：获取患者正脸露齿照片；对所述患者正脸露齿照片进行面部特征点识别；基于识别出的相应面部特征点，产生面中线；基于所述患者正脸露齿照片中的唇内区域图像，产生牙中线；基于识别出的相应面部特征点，识别虹膜区域；以及基于给定的虹膜直径尺寸、所述识别出的虹膜区域直径的像素数以及所述面中线和牙中线之间的像素数偏差，计算得到所述面中线和牙中线之间的实际偏差。
在一些实施方式中，所述的确定面中线位置的方法还包括：对所述患者正脸露齿照片进行面部区域检测，所述面部特征点识别是针对检测到的面部区域图像进行。
在一些实施方式中，所述面部区域检测采用以下方法之一：基于传统的机器学习方法以及基于深度神经网络的方法。
在一些实施方式中，所述面部特征点检测采用以下方法之一：基于模型的方法、级联形状回归的方法以及基于深度神经网络的方法。
在一些实施方式中，所述面中线是基于所述相应面部特征点拟合得到。
在一些实施方式中，所述的确定面中线位置的方法还包括：基于识别出的相应面部特征点，识别所述唇内区域；以及针对所述唇内区域图像进行图像处理，以产生所述牙中线。
在一些实施方式中，所述的确定面中线位置的方法还包括：基于识别出的相应面部特征点，识别眼部区域；针对所述眼部区域图像进行边缘提取；以及基于提取到的边缘确定所述虹膜区域。
在一些实施方式中，所述的确定面中线位置的方法还包括：基于所述提取到的边缘进行霍夫变换，确定所述虹膜区域。
在一些实施方式中，所述的确定面中线位置的方法还包括：获取所述患者的牙列三维数字模型；确定所述牙中线在所述牙列三维数字模型上的位置；以及基于所述面中线和牙中线之间的实际偏差以及所述中线在所述牙列三维数字模型上的位置，确定所述面中线在所述牙列三维数字模型上的位置。
本申请的又一方面提供了一种用于确定面中线位置的计算机系统，其包括存储装置和处理器，所述存储装置存储有一计算机程序，当其被所述处理器执行后，将执行所述的确定面中线位置的方法。
附图说明
图1为本发明实施例1提供的三维牙颌模型中面中线的确定方法的第一流程示意图；
图2为本发明实施例1提供的三维牙颌模型中面中线的确定方法的第二流程示意图；
图3为本发明实施例1提供的三维牙颌模型中面中线的确定方法中面部特征定位点的示意图；
图4为本发明实施例1提供的三维牙颌模型中面中线的确定方法中目标 二维照片中第一面中线和牙齿中线的位置的示意图；
图5为本发明实施例1提供的三维牙颌模型中面中线的确定方法中三维牙颌模型的示意图；
图6为本发明实施例2提供的牙齿矫治位置的确定方法的流程示意图；
图7为本发明实施例3提供的牙齿矫治器的制造方法的流程示意图；
图8为本发明实施例4提供的三维牙颌模型中面中线的确定系统的结构示意图；
图9为本发明实施例5提供的牙齿矫治位置的确定系统的结构示意图；
图10为本发明实施例7提供的电子设备的结构示意图；
图11为本申请又一实施例中的确定面中线位置的方法的示意性流程图；
图12为本申请一个实施例中用于确定面中线位置的计算机程序的一个界面所展示的一个例子中的患者正脸微笑照片；
图13A示意性地展示了68个面部特征点的分布；
图13B示意性地展示了51个面部特征点的分布；
图14为所述计算机程序的一个界面所展示的一个例子中的患者正脸微笑照片以及基于其产生的面中线和牙中线；以及
图15为所述计算机程序的一个界面所展示的一个例子中标出面中线和牙中线的牙颌三维数字模型。
具体实施方式
下面通过实施例的方式进一步说明本发明，但并不因此将本发明限制在所述的实施例范围之中。
实施例1
本实施例提供一种三维牙颌模型中面中线的确定方法，如图1所示，确定方法包括：
S101、获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置。
其中，目标二维照片包括目标对象露出目标牙齿的面部。
S102、基于目标牙齿获取目标对象在三维牙颌模型中的第二牙齿中线的位置，获取目标牙齿在三维牙颌模型中的第二尺寸值。
S103、获取目标牙齿在目标二维照片中的第一尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系。
S104、基于第一面中线的位置和第一牙齿中线的位置得到第一面中线和第一牙齿中线之间的第一偏差值。
S105、基于映射关系和第一偏差值得到目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值。
S106、基于三维牙颌模型中的第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置。
第一牙齿中线即目标对象在目标二维照片中的牙齿中线，第一面中线即目标对象在目标二维照片中的面中线；第二牙齿中线即目标对象在三维牙颌模型的牙齿中线，第二面中线即目标对象在三维牙颌模型的面中线。
第二尺寸值是目标对象的目标牙齿的实际物理尺寸值，其获取方法可以是通过三维牙颌模型获取，也可以是通过牙齿的统计学规律获取。
本实施例的三维牙颌模型中面中线的确定方法，可以自动的获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置，根据目标对象的目标牙齿在目标二维照片中的第一尺寸值，和目标牙齿在三维牙颌模型中的第二尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系，进而根据第一面中线和第一牙齿中线之间的第一偏差值得出目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值，最终根据三维牙颌模型中的第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置；实现了自动的在三维牙颌模型得出面中线的位置，提高了确定面中线的准确性和效率。
在一可选的实施方式中，如图2所示，上述步骤S101包括：
S1011、获取目标对象的目标二维照片。
S1012、获取目标二维照片中的面部区域。
S1013、获取面部区域的面部特征定位点。
S1014、基于面部特征定位点得到第一面中线的位置。
S1015、基于目标牙齿在目标二维照片中的位置得到第一牙齿中线的位置。
其中，目标牙齿包括上颌两颗中切牙中的至少一颗，或下颌两颗中切牙中的至少一颗。
若目标对象为人，则目标二维照片可以是漏出目标牙齿的正面微笑照片。
中切牙即俗称的门牙，牙齿中线也可以称为牙列中线，例如，牙齿中线是穿过上颌两颗或下颌两颗中切牙之间的一条假想线。
上述步骤S1012中，具体基于面部区域识别算法对目标二维照片进行图像识别，以得到面部区域。
通过面部区域识别算法对目标二维照片进行图像识别，检测目标二维照片中是否存在面部，并标出面部区域。面部区域识别算法可以是传统的机器学习方法，如Haar Cascade(哈尔级联)、Eigenfaces(特征脸)、Fisherfaces(脸部识别器)等，也可以使基于卷积神经网络的方法。
Haar Cascade算法是一种用于在图像上定位对象的对象检测算法，基于机器学习，该算法从大量正样本和负样本中学习，前者包含感兴趣的对象，而后者包含除要查找的对象之外的任何内容，从大量积极和消极的图像里训练一个级联函数，然后用来在其他图像里检测对象。
Eigenfaces算法可以通过对大量面部图像的统计分析确定“标准化面部成分”的集合，面部特征被赋予常数值，因为这种算法不使用数字图片，而是使用统计数据库，任何面部图像都是这些值的不同百分比的组合。它使用主成分分析(Principal Component Analysis，PAC)方法将高维的面部数据处理为低维数据后，在进行数据分析和处理，获取识别结果，对原始数据使用PCA方法进行降维，获取其中的主成分信息，从而实现面部识别的方法。
Fisherfaces算法是一种基于特征脸的方法，找到使数据中最大方差的特 征线性组合，Eigenface是在人脸识别的计算机视觉问题中使用的一组特征向量的名称，Eigenfaces是基于PCA(主成分分析)的。
若目标二维照片中无法识别到面部区域，则需要重新获取新的目标二维照片。
上述步骤S1013中，具体基于面部特征识别算法对面部区域进行识别，以得到面部特征定位点。
面部特征定位点也可以称为面部关键特征点。
在识别出面部区域后，通过面部特征识别算法对面部区域进行识别，得到面部特征定位点。面部特征识别算法可以是基于模型的ASM(Active Shape Model，主动形状模型)和AAM(Active Appearnce Model，主动外观模型)，或基于级联形状回归的算法，也可以是基于深度学习的算法。
ASM是对图像中的形状(如面部特征定位点)进行建模的可变模型，得到的形状模型既可以用来分析新的形状，如拟合模型得到新形状，也可以用于生成形状，如在给定图像中搜索形状。
AAM是基于主动外观模型的图像分割算法，该方法可以分为两大部分，第一部分是训练部分，即构造一个事物的主动外观模型，这部分需要一个训练集，它的作用是让程序记住需要切割的图像的形状特征和外观特征；第二部分是图像分割阶段，这个部分是将刚才的训练集收集的关于该事物的特征在需要分割的图像里面寻找，找到与训练集相似的物体图像的轮廓和外观，然后将其从整幅的图像中分割出来。
图3为本实施方式提供的面部特征定位点的示意图。如图3所示，面部特征定位点1-17为脸部外轮廓对应的面部特征定位点，面部特征定位点18-27为两个眉毛对应的面部特征定位点，面部特征定位点28-36为鼻梁对应的面部特征定位点，面部特征定位点37-48为两个眼睛对应的面部特征定位点，面部特征定位点49-68为唇部对应的面部特征定位点。
上述步骤S1014中，具体基于拟合算法对面部特征定位点进行拟合处理，以得到第一面中线的位置。
拟合算法包括但不限于最小二乘法和Hough(霍夫)变换法。
Hough变换的基本原理是将影像空间中的曲线(包括直线)变换到参数空间中，通过检测参数空间中的极值点，确定出该曲线的描述参数，从而提取影像中的规则曲线。
面部特征定位点以第一面中线为对称轴，对称分布在第一面中线的两侧。
上述步骤S1015中，具体通过获取上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线，并将中线作为第一牙齿中线，以得到第一牙齿中线的位置。
即第一牙齿中线可以是上颚两颗中切牙之间的中线，也可以是下颚两颗中切牙之间的中线。
若目标牙齿仅为一颗中切牙，则以该中切牙同一牙颌方向上靠近相邻的另一中切牙的边界线称为近中边界线，并将近中边界线作为第一牙齿中线。中切牙同一牙颌方向上远离另一中切牙的边界线称为远中边界线。
第一牙齿中线的获取可以是基于传统图像识别的方式，如边缘检测；也可以是基于深度学习的方式，如卷积神经网络。同时，根据近中边界线和远中边界线可以计算出相应的中切牙在目标二维照片中的第一尺寸值。
图4为本实施方式提供的目标二维照片中第一面中线和牙齿中线的位置的示意图，图4中示出了第一牙齿中线和第一面中线的位置，第一牙齿中线的位置为上颚两个中切牙的中线。
本实施方式的三维牙颌模型中面中线的确定方法，根据相应的算法，自动的得到目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置，根据目标对象的目标牙齿在目标二维照片中的第一尺寸值，和目标牙齿在三维牙颌模型中的第二尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系，进而根据第一面中线和第一牙齿中线之间的第一偏差值得出目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值，最终根据三维牙颌模型中的第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置；实现了自动的在三维牙颌模型得出面中线的位置， 提高了确定面中线的准确性和效率。
在一可选的实施方式中，上述步骤S1013之后还包括：
S10131、基于面部特征定位点的位置判断目标二维照片中的目标对象的面部是否存在倾斜。
若是，则执行步骤S10132，若否，则执行步骤S1014。
S10132、对面部特征定位点进行旋转校正操作。
进行旋转校正操作后执行步骤S1014。
为防止拍摄角度或面部姿态变化导致的二维目标照片中的目标对象的面部倾斜，可基于以上获取的面部特征定位点进行旋转矫正，以目标对象为人为例，以左右两眼内角点连线中心为坐标原点，建立人脸图像坐标系，旋转图像使两眼内角点保持水平，即可获得旋转矩阵，从而实现对面部特征定位点的旋转校正操作。
本实施方式的三维牙颌模型中面中线的确定方法，通过对面部特征定位点进行旋转校正操作，使得目标对象的面部区域不再倾斜，使得目标二维照片中的脸部保持端正，便于后续在目标二维照片中标注出第一牙齿中线的位置和第一面中线的位置。
在一可选的实施方式中，上述步骤S1014包括：
在面部特征定位点中选取出面部特征关键点，并基于面部特征关键点得到第一面中线的位置。
其中，面部特征定位点包括眼睛、眉毛、鼻梁、唇部、脸部外轮廓对应的定位点；面部特征关键点包括眉心点、眼角点，鼻尖点、鼻底点、人中点中的至少三个。
如图3所示，面部特征定位点20和25为眉心点对应的面部特征关键点；面部特征定位点37和46为外眼角点对应的面部特征关键点；面部特征定位点40和43为内眼角点对应的面部特征关键点；面部特征定位点31为鼻尖点对应的面部特征关键点；面部特征定位点34为鼻底点对应的面部特征关键点，以面部特征定位点34和52的中点作为人中点。
选取至少三个面部特征关键点，基于拟合算法对面部特征定位点进行拟合处理，以得到第一面中线的位置。
在一可选的实施方式中，上述步骤S102包括：
S1021、获取目标对象包含有牙冠和牙龈的三维牙颌模型。
S1022、在三维牙颌模型中，以上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线作为第二牙齿中线的位置，并得到对应的中切牙的第二尺寸值。
通过对目标对象的口腔进行扫描，建立出目标对象包含有牙冠和牙龈的三维牙颌模型。在三维牙颌模型中，以上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线作为第二牙齿中线的位置。
可以根据三维牙颌模型可以得到目标对象各个牙齿的分布情况和每个牙齿的尺寸，进而得知中切牙的第二尺寸值。
图5为本实施方式提供的三维牙颌模型的示意图，图5中示出了第二牙齿中线和第二面中线的位置。
本实施方式的三维牙颌模型中面中线的确定方法，实现了自动的在三维牙颌模型确定出第二面中线的位置，提高了确定面中线的准确性和效率。
在一可选的实施方式中，上述步骤S102包括：
S1023、获取目标对象包含有牙冠和牙龈的三维牙颌模型。
S1024、在三维牙颌模型中，以上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线作为第二牙齿中线的位置，并基于预设尺寸值得到对应的所述中切牙的第二尺寸值。
通过对目标对象的口腔进行扫描，建立出目标对象包含有牙冠和牙龈的三维牙颌模型。在三维牙颌模型中，以上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线作为第二牙齿中线的位置。
根据统计学规律，可以得出牙齿的尺寸信息，例如，根据目标对象的年龄、性别、牙齿的大致形状和大小，根据牙齿数据库中存储的预设尺寸值得到中切牙的第二尺寸值。
本实施方式的三维牙颌模型中面中线的确定方法，实现了自动的在三维牙颌模型确定出第二面中线的位置，提高了确定面中线的准确性和效率。
在一可选的实施方式中，上述步骤S103包括：
S1031、基于图像识别算法获取目标牙齿在目标二维照片中的像素值，并将像素值作为第一尺寸值。
S1032、基于目标牙齿对应的第一尺寸值和第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系。
在得到某一中切牙的近中边界线和远中边界线后，基于图像识别算法计算出该中切牙在目标二维照片中的像素值，并将该像素值作为目标牙齿在目标二维照片中的第一尺寸值，同时，在三维牙颌模型中可以获取到该中切牙的第二尺寸值。基于中切牙在目标二维照片中对应的第一尺寸值和在三维牙颌模型中对应的第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系。
本实施方式的三维牙颌模型中面中线的确定方法，实现了自动的获取目标牙齿在目标二维照片中的第一尺寸值，并基于目标牙齿对应的第一尺寸值和第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系，进而自动的在三维牙颌模型确定出第二面中线的位置，提高了确定面中线的准确性和效率。
在一可选的实施方式中，上述步骤S103包括：
S1033、基于目标二维照片对应的预设标尺获取目标牙齿在目标二维照片中的标尺值，并将标尺值作为第一尺寸值。
S1034、基于目标牙齿对应的第一尺寸值和第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系。
当目标二维照片具有对应的预设标尺时，根据预设标尺定义的数值大小即可得出目标牙齿在目标二维照片中的标尺值，并将标尺值作为第一尺寸值。
本实施方式的三维牙颌模型中面中线的确定方法，实现了自动的获取目标牙齿在目标二维照片中的第一尺寸值，并基于目标牙齿对应的第一尺寸值和第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系，进而自动的在三维牙颌模型确定出第二面中线的位置，提高了确定面中线的准确性和效率。
在一可选的实施方式中，上述步骤S105包括：
S1051、基于第二偏差值得到第二面中线相对于第二牙齿中线的相对位置。
S1052、基于相对位置得到第二面中线在三维牙颌模型中的位置。
在获取到第一面中线和第一牙齿中线之间的第一偏差值后，可以根据第一尺寸值与第二尺寸值之间的映射关系，将第一偏差值转换为对应的第二偏差值，由于第二牙齿中线的位置已经确定，根据第二偏差值即可得到第二面中线相对于第二牙齿中线的相对位置，进而基于该相对位置得到第二面中线在三维牙颌模型中的位置。
本实施方式的三维牙颌模型中面中线的确定方法，通过目标对象的第二面中线与第二牙齿中线的第二偏差值，和第一尺寸值与第二尺寸值之间的映射关系，及三维牙颌模型中的第二牙齿中线的位置，自动得到第二面中线在三维牙颌模型中的位置，提高了确定面中线的准确性和效率。
实施例2
本实施例提供一种牙齿矫治位置的确定方法，如图6所示，牙齿矫治位置的确定方法包括：
S201、采用实施例1中的三维牙颌模型中面中线的确定方法得到目标对象在三维牙颌模型中的第二面中线的位置。
S202、基于第二面中线的位置在三维牙颌模型中建立三维坐标系。
S203、基于三维坐标系和目标对象在三维牙颌模型中的第二牙齿中线的位置确定目标对象每个牙齿对应的矫治位置。
本实施例中的牙齿矫治位置的确定方法，基于实施例1中的三维牙颌模型中面中线的确定方法自动的获取到目标对象在三维牙颌模型中的面中线的位置，进而基于第二面中线的位置建立三维坐标系，根据目标对象在三维牙颌模型中的第二牙齿中线在三维坐标系中的位置，确定出目标对象每个牙齿对应的矫治位置，例如某一牙齿相对于第二面中线是向左移动或是向右移动。
本实施例的牙齿矫治位置的确定方法，可以快速的确定出目标对象每个牙齿对应的矫治位置，进而灵活的设计和调整牙齿矫治方案。
实施例3
本实施例提供一种牙齿矫治器的制造方法，如图7所示，牙齿矫治器的制造方法包括：
S301、采用实施例2中的牙齿矫治位置的确定方法获得目标对象的牙齿的目标矫治位置；
S302、基于目标矫治位置确定目标对象的牙齿矫治器的形状，以制造出具有相应形状的牙齿矫治器。
具体地，所得到的牙齿的目标矫治位置的数据可以被传送至牙齿矫治器制造设备，牙齿矫治器制造设备可根据该数据按照一般的工程方法形成牙齿矫治器的模具(例如阳模)，从而由该模具制造出具有相应形状的牙齿矫治器。
当然，牙齿矫治器制造设备也可以利用三维打印技术，根据牙齿的目标矫治位置的数据形成牙齿矫治器的阳模，并且进一步由该阳模制造出具有相应形状的牙齿矫治器。
并且，牙齿矫治器制造设备还可以利用三维打印技术，根据牙齿的目标矫治位置的数据确定相应的牙齿矫治器的数据，并根据牙齿矫治器的数据直接形成牙齿矫治器。
在一可选的实施方式中，牙齿矫治器由具有韧性的、透明的聚合物材料制造，从而形成无托槽的隐形矫治器。
本实施例的牙齿矫治器的制造方法，可以根据目标对象的牙齿的目标矫治位置，快速且精准的确定出目标对象的牙齿矫治器的形状，进而制造出具有相应形状的牙齿矫治器。
实施例4
本实施例提供一种三维牙颌模型中面中线的确定系统，如图8所示，三维牙颌模型中面中线的确定系统包括：
第一获取模块1，用于获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置；
其中，目标二维照片包括目标对象露出目标牙齿的面部；
第二获取模块2，用于基于目标牙齿获取目标对象在三维牙颌模型中的第二牙齿中线的位置，获取目标牙齿在三维牙颌模型中的第二尺寸值；
映射关系建立模块3，用于获取目标牙齿在目标二维照片中的第一尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系；
第一偏差值获取模块4，用于基于第一面中线的位置和第一牙齿中线的位置得到第一面中线和第一牙齿中线之间的第一偏差值；
第二偏差值获取模块5，用于基于映射关系和第一偏差值得到目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值；
第三获取模块6，用于基于三维牙颌模型中的第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置。
在一可选的实施方式中，第一获取模块1包括二维照片获取单元11，用于获取目标对象的目标二维照片；面部区域获取单元12，用于获取目标二维照片中的面部区域；定位点获取单元13，用于获取面部区域的面部特征定位点；第一面中线获取单元14，用于基于面部特征定位点得到第一面中线的位置；第一牙齿中线获取单元15，用于基于目标牙齿在目标二维照片中的位置得到第一牙齿中线的位置；其中，目标牙齿包括上颌两颗中切牙中的至少一颗，或下颌两颗中切牙中的至少一颗。
在一可选的实施方式中，第一获取模块1还包括面部倾斜判断单元16，用于基于面部特征定位点的位置判断目标二维照片中的目标对象的面部是否存在倾斜；校正单元17，用于当面部倾斜判断单元16判断出目标二维照片中的目标对象的面部存在倾斜时，对面部特征定位点进行旋转校正操作，并运行第一面中线获取单元14；若面部倾斜判断单元16判断出目标二维照片中的目标对象的面部不存在倾斜时，则运行第一面中线获取单元14。
在一可选的实施方式中，第一面中线获取单元14具体用于在面部特征 定位点中选取出面部特征关键点，并基于面部特征关键点得到第一面中线的位置；其中，面部特征定位点包括眼睛、眉毛、鼻梁、唇部、脸部外轮廓对应的定位点；面部特征关键点包括眉心点、眼角点，鼻尖点、鼻底点、人中点中的至少三个。
在一可选的实施方式中，面部区域获取单元12具体用于基于面部区域识别算法对目标二维照片进行图像识别，以得到面部区域。
在一可选的实施方式中，定位点获取单元13具体用于基于面部特征识别算法对面部区域进行识别，以得到面部特征定位点。
在一可选的实施方式中，第一面中线获取单元14具体用于基于拟合算法对面部特征定位点进行拟合处理，以得到第一面中线的位置。
在一可选的实施方式中，第一牙齿中线获取单元15具体用于获取上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线，并将中线作为第一牙齿中线，以得到第一牙齿中线的位置。
在一可选的实施方式中，第二获取模块2包括三维牙颌模型获取单元21，用于获取目标对象包含有牙冠和牙龈的三维牙颌模型；第二尺寸值获取单元22，用于在三维牙颌模型中，以上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线作为第二牙齿中线的位置，并得到对应的中切牙的第二尺寸值。
在一可选的实施方式中，第二获取模块2包括三维牙颌模型获取单元21，用于获取目标对象包含有牙冠和牙龈的三维牙颌模型；第二尺寸值获取单元22，用于在三维牙颌模型中，以上颚两颗中切牙之间的中线或下颚两颗中切牙之间的中线作为第二牙齿中线的位置，基于预设尺寸值得到对应的中切牙的第二尺寸值。
在一可选的实施方式中，映射关系建立模块3包括第一尺寸值获取单元31，用于基于图像识别算法获取目标牙齿在目标二维照片中的第一尺寸值；映射关系建立单元32，用于基于目标牙齿对应的第一尺寸值和第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系。
在一可选的实施方式中，映射关系建立模块3包括第一尺寸值获取单元 31，用于基于目标二维照片对应的预设标尺获取目标牙齿在目标二维照片中的第一尺寸值；映射关系建立单元32，用于基于目标牙齿对应的第一尺寸值和第二尺寸值建立第一尺寸值与第二尺寸值之间的映射关系。
在一可选的实施方式中，第三获取模块6包括相对位置确定单元61，用于基于第二偏差值得到第二面中线相对于第二牙齿中线的相对位置；第二面中线确定单元62，用于基于相对位置得到第二面中线在三维牙颌模型中的位置。
本实施例中的三维牙颌模型中面中线的确定系统的工作原理与实施例1中的三维牙颌模型中面中线的确定方法的工作原理相同，此处就不再赘述。
本实施例中的三维牙颌模型中面中线的确定系统，可以自动的获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置，根据目标对象的目标牙齿在目标二维照片中的第一尺寸值，和目标牙齿在三维牙颌模型中的第二尺寸值，建立第一尺寸值与第二尺寸值之间的映射关系，进而根据第一面中线和第一牙齿中线之间的第一偏差值得出目标对象在三维牙颌模型中的第二面中线与第二牙齿中线的第二偏差值，最终根据三维牙颌模型中的第二牙齿中线的位置和第二偏差值得到第二面中线在三维牙颌模型中的位置；实现了自动的在三维牙颌模型得出面中线的位置，提高了确定面中线的准确性和效率。
实施例5
本实施例提供一种牙齿矫治位置的确定系统，如图9所示，牙齿矫治位置的确定系统包括：
面中线获取模块7，用于采用实施例4中的三维牙颌模型中面中线的确定系统得到目标对象在三维牙颌模型中的面中线的位置；
坐标系建立模块8，用于基于面中线的位置在三维牙颌模型中建立三维坐标系；
矫治位置确定模块9，用于基于三维坐标系和目标对象在三维牙颌模型中的牙齿中线的位置确定目标对象每个牙齿对应的矫治位置。
本实施例的牙齿矫治位置的确定系统的工作原理与实施2中的牙齿矫治位置的确定方法的工作原理相同，此处就不再赘述。
本实施例的牙齿矫治位置的确定系统，基于实施例4中的三维牙颌模型中面中线的确定系统自动的获取到目标对象在三维牙颌模型中的面中线的位置，可以快速的确定出目标对象每个牙齿对应的矫治位置，进而灵活的设计和调整牙齿矫治方案。
实施例6
本实施例提供一种牙齿矫治器的制造系统，牙齿矫治器的制造系统包括：
矫治位置获取模块，用于采用实施例5中的牙齿矫治位置的确定系统获得目标对象的牙齿的目标矫治位置；
矫治器制造模块，用于基于所述目标矫治位置确定所述目标对象的牙齿矫治器的形状，以制造出具有相应形状的牙齿矫治器。
本实施例的牙齿矫治器的制造系统的工作原理与实施3中的牙齿矫治器的制造方法的工作原理相同，此处就不再赘述。
本实施例的牙齿矫治器的制造系统，基于实施例5中的牙齿矫治位置的确定系统确定出目标对象的牙齿的目标矫治位置，快速且精准的确定出目标对象的牙齿矫治器的形状，进而制造出具有相应形状的牙齿矫治器。
实施例7
本实施例提供一种电子设备，图10为本实施例提供的电子设备的结构示意图，电子设备包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，处理器执行计算机程序时实现上述实施例1中的三维牙颌模型中面中线的确定方法，或上述实施例2中的牙齿矫治位置的确定方法，或上述实施例3中的牙齿矫治器的制造方法。图10显示的电子设备70仅仅是一个示例，不应对本发明实施例的功能和使用范围带来任何限制。如图10所示，电子设备70可以以通用计算设备的形式表现，例如其可以为服务器设备。电子设备70的组件可以包括但不限于：上述至少一个处理器71、上述至少一个存储器72、连接不同系统组件(包括存储器72和处理器71)的 总线73。
总线73包括数据总线、地址总线和控制总线。
存储器72可以包括易失性存储器，例如随机存取存储器(RAM)721和/或高速缓存存储器722，还可以进一步包括只读存储器(ROM)723。
存储器72还可以包括具有一组(至少一个)程序模块724的程序工具725(或实用工具)，这样的程序模块724包括但不限于：操作系统、一个或者多个应用程序、其它程序模块以及程序数据，这些示例中的每一个或某种组合中可能包括网络环境的实现。
处理器71通过运行存储在存储器72中的计算机程序，从而执行各种功能应用以及数据处理，例如上述实施例1中的三维牙颌模型中面中线的确定方法，或上述实施例2中的牙齿矫治位置的确定方法，或上述实施例3中的牙齿矫治器的制造方法。
电子设备70也可以与一个或多个外部设备74通信。这种通信可以通过输入/输出(I/O)接口75进行。并且，模型生成的电子设备70还可以通过网络适配器76与一个或者多个网络(例如局域网(LAN)，广域网(WAN)和/或公共网络，例如因特网)通信。如图10所示，网络适配器76通过总线73与电子设备70的其它模块通信。应当明白，尽管图中未示出，可以结合电子设备70使用其它硬件和/或软件模块，包括但不限于：微代码、设备驱动器、冗余处理器、外部磁盘驱动阵列、RAID(磁盘阵列)系统、磁带驱动器以及数据备份存储系统等。
应当注意，尽管在上文详细描述中提及了电子设备的若干单元/模块或子单元/模块，但是这种划分仅仅是示例性的并非强制性的。实际上，根据本发明的实施方式，上文描述的两个或更多单元/模块的特征和功能可以在一个单元/模块中具体化。反之，上文描述的一个单元/模块的特征和功能可以进一步划分为由多个单元/模块来具体化。
实施例8
本实施例提供一种计算机可读存储介质，其上存储有计算机程序，计算 机程序在由处理器执行时实现上述实施例1中的三维牙颌模型中面中线的确定方法，或上述实施例2中的牙齿矫治位置的确定方法，或上述实施例3中的牙齿矫治器的制造方法。
其中，可读存储介质可以采用的更具体可以包括但不限于：便携式盘、硬盘、随机存取存储器、只读存储器、可擦拭可编程只读存储器、光存储器件、磁存储器件或上述的任意合适的组合。
在可能的实施方式中，本发明还可以实现为一种程序产品的形式，其包括程序代码，当程序产品在终端设备上运行时，程序代码用于使终端设备执行实现上述实施例1中的三维牙颌模型中面中线的确定方法中的步骤，或上述实施例2中的牙齿矫治位置的确定方法中的步骤，或上述实施例3中的牙齿矫治器的制造方法中的步骤。
其中，可以以一种或多种程序设计语言的任意组合来编写用于执行本发明的程序代码，程序代码可以完全地在用户设备上执行、部分地在用户设备上执行、作为一个独立的软件包执行、部分在用户设备上部分在远程设备上执行或完全在远程设备上执行。
本申请的又一方面提供了一种确定面中线位置的方法，对患者正脸露齿照片进行特征点检测，基于检测到的特定特征点拟合得到面中线和牙中线，然后，基于虹膜尺寸确定面中线和牙中线的实际间距。以下结合附图对本申请一个实施例中的确定面中线位置的方法进行详细说明。
请参图11，为本申请一个实施例中的确定面中线位置的方法100的示意性流程图。
在一个实施例中，所述确定面中线位置的方法100是由计算机执行。本申请又一方面提供了一种用于确定面中线位置的计算机系统，其包括存储装置和处理器，其中，所述存储装置存储有一计算机程序，当其被所述处理器执行后，将执行所述确定面中线位置的方法100。
在101中，获取患者正脸露齿照片。
在一个实施例中，患者正脸露齿照片可以是正脸微笑照片。
请参图12，为本申请一个实施例中用于确定面中线位置的计算机程序的一个界面所展示的一个例子中的患者正脸微笑照片。为了保护患者隐私，将图12所示患者照片的眼睛部位遮蔽。
在103中，对所述患者正脸露齿照片进行面部区域检测。
在一个实施例中，在获得患者正脸露齿照片后，可以对其进行面部区域检测。一方面，可以藉此筛除不规范的正脸微笑照片，例如，五官不全、面部扭转、具有摩尔纹的照片等。另一方面，后续计算可以基于检测到的面部区域而非整张正脸微笑照片，以提高计算效率和可靠性。
在一个实施例中，可以利用机器学习的方法进行面部区域检测，若未能检测到人脸，则判定该图像不规范，例如，五官不全、面部扭转等。
在一个实施例中，可以在检测到人脸之后进行对摩尔纹的识别。首先，获取人脸图像中有摩尔纹和无摩尔纹的样本，然后，利用SVM(支持向量机)的方法训练一个能够识别摩尔纹的模型，利用该模型对人脸图像进行摩尔纹识别。
在一个实施例中，可以采用传统的机器学习方法进行面部区域检测，例如，Haar Cascade、Eigenfaces、Fisherfaces等。在又一实施例中，也可以采用基于深度神经网络的方法进行面部区域检测，例如，基于卷积神经网络的方法。
在一个实施例中，若系统未能检测到规范的脸部区域，可以提示用户重新输入正脸微笑照片。
请再参图12，其中方框区域是检测到的脸部区域。
在105中，对所述脸部区域进行面部特征点检测。
在检测到面部区域之后，可以对面部区域图片进行面部特征点检测。
面部特征点标识面部的特殊位置，例如眼睛、眉毛、鼻梁、唇部等。
在一个实施例中，可以采用基于模型的方法进行特征点检测，例如，ASM(Active Shape Model)和AAM(Active Appearance Model)。在又一实施例中，也可以采用级联形状回归的方法进行特征点检测。在又一实施例中，还 可以采用基于深度神经网络的方法进行特征点检测。
请参图13A，示意性地展示了68个面部特征点的分布。请参图13B，示意性地展示了51个面部特征点的分布。
在107中，基于相应的面部特征点拟合面中线。
在检测出面部特征点之后，就可以基于相应的面部特征点拟合面中线和牙中线。
在一个实施例中，可以基于基本位于面部对称中心线(即面中线)的特征点拟合面中线，例如，眉心点、鼻尖点、鼻底点、人中点等。在一个实施例中，可以采用最小二乘法拟合面中线。
在一个实施例中，可以基于相应的特征点定位唇部区域，然后，基于唇部区域图片中裸露的牙齿，通过图像处理(图像分割)的方法识识别出两个中切牙，再通过其轮廓信息确定两个中切牙的牙缝，并将其作为牙中线。
在一些情况下，正脸微笑照片可能存在牙齿未裸露、牙齿裸露过少或缺牙等情况，这些会对牙中线的识别产生影响，因此，在识别牙中线之前，可以先识别这些情况并将其筛除。在一个实施例中，可以利用基于深度学习的方法对唇部子图进行分类，以筛除上述不合格的情况。在又一实施例中，也可以利用其他机器学习的方法进行此操作，例如，决策树、SVM、AdaBoost等。
请参图14，为所述计算机程序的一个界面所展示的一个例子中的患者正脸微笑照片以及基于其产生的面中线和牙中线。为了保护患者隐私，将患者照片的眼睛部位遮蔽。
在获得面中线和牙中线之后，虽然能够计算得到两者之间以像素为单位的距离(偏差)，但由于不同的照片的单个像素的实际尺寸可能不同，因此，此时还无法获得面中线和牙中线之间的实际偏差。
通过统计，本申请的发明人发现，不同人的虹膜直径变化不大，大约为12.3±0.7mm。在一个实施例中，可以取12.3mm作为虹膜直径，并基于该尺寸估算单个像素的实际尺寸，进而计算面中线和牙中线之间的实际偏差。
在109中，基于给定虹膜尺寸计算所述面中线和牙中线的实际偏差。
在一个实施例中，可以先基于相应的面部特征点获取眼部区域图片，然后，提取边缘并进行优化处理，再通过霍夫变换检测出可能的虹膜区域，最后，对这些可能区域进行筛选以确定最终的虹膜区域。接着，基于所述选定的虹膜直径与所述虹膜区域直径的像素数，计算出每个像素的物理尺寸。最后，基于所述面中线和牙中线的像素数偏差和像素的物理尺寸，计算出所述面中线和牙中线之间的物理偏差。
在111中，基于所述物理偏差，确定面中线在牙颌三维数字模型上位置。
可以理解，面中线的位置是指沿正脸方向投影的位置。
如今，越来越多牙科专业人员基于患者牙颌三维数字模型设计牙齿正畸治疗方案。中线对齐(即牙中线与面中线对齐)是牙科正畸治疗的目标之一。因此，在患者牙颌三维数字模型上标出面中线和牙中线便于牙科专业人员设计牙齿正畸治疗方案。
在一个实施例中，可以通过口内扫描，或扫描印模或牙齿实体模型等方式，获得牙颌三维数字模型。
由于面中线和牙中线的物理偏差为已知，而牙颌三维数字模型的物理尺寸也为已知，因此，在牙中线位置已知的情况下，计算机能够在牙颌三维数字模型上标出面中线的位置，以供牙科专业人员设计牙齿正畸治疗方案时参考。
请参图15，为所述计算机程序的一个界面所展示的一个例子中标出面中线和牙中线的牙颌三维数字模型。其中，牙中线由位于上颌两颗门牙之间的一个点表示。
虽然以上描述了本发明的具体实施方式，但是本领域的技术人员应当理解，这仅是举例说明，本发明的保护范围是由所附权利要求书限定的。本领域的技术人员在不背离本发明的原理和实质的前提下，可以对这些实施方式做出多种变更或修改，但这些变更和修改均落入本发明的保护范围。

Claims (27)

  1. 一种三维牙颌模型中面中线的确定方法，其特征在于，所述确定方法包括：
    获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置；
    其中，所述目标二维照片包括所述目标对象露出目标牙齿的面部；
    基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值；
    获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系；
    基于所述第一面中线的位置和所述第一牙齿中线的位置得到所述第一面中线和所述第一牙齿中线之间的第一偏差值；
    基于所述映射关系和所述第一偏差值得到所述目标对象在所述三维牙颌模型中的第二面中线与所述第二牙齿中线的第二偏差值；
    基于所述三维牙颌模型中的所述第二牙齿中线的位置和所述第二偏差值得到所述第二面中线在所述三维牙颌模型中的位置。
  2. 根据权利要求1所述的确定方法，其特征在于，所述获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置包括：
    获取所述目标对象的所述目标二维照片；
    获取所述目标二维照片中的面部区域；
    获取所述面部区域的面部特征定位点；
    基于所述面部特征定位点得到所述第一面中线的位置；
    基于所述目标牙齿在所述目标二维照片中的位置得到所述第一牙齿中线的位置；
    其中，所述目标牙齿包括上颌两颗中切牙中的至少一颗，或下颌两颗中切牙中的至少一颗。
  3. 根据权利要求2所述的确定方法，其特征在于，所述获取所述面部区域的面部特征定位点的步骤之后还包括：
    基于所述面部特征定位点的位置判断所述目标二维照片中的所述目标对象的面部是否存在倾斜；
    若是，则对所述面部特征定位点进行旋转校正操作，并执行所述基于所述面部特征定位点得到所述第一面中线的位置的操作；
    若否，则执行所述基于所述面部特征定位点得到所述第一面中线的位置的操作。
  4. 根据权利要求2所述的确定方法，其特征在于，所述基于所述面部特征定位点得到所述第一面中线的位置的步骤包括：
    在所述面部特征定位点中选取出面部特征关键点，并基于所述面部特征关键点得到所述第一面中线的位置；
    其中，所述面部特征定位点包括眼睛、眉毛、鼻梁、唇部、脸部外轮廓对应的定位点；所述面部特征关键点包括眉心点、眼角点，鼻尖点、鼻底点、人中点中的至少三个。
  5. 根据权利要求2所述的确定方法，其特征在于，所述获取所述目标二维照片中的面部区域的步骤包括：
    基于面部区域识别算法对所述目标二维照片进行图像识别，以得到所述面部区域；
    和/或，所述获取所述面部区域的面部特征定位点的步骤包括：
    基于面部特征识别算法对所述面部区域进行识别，以得到所述面部特征定位点；
    和/或，所述基于所述面部特征定位点得到所述第一面中线的位置的方法包括：
    基于拟合算法对所述面部特征定位点进行拟合处理，以得到所述第一面中线的位置；
    和/或，所述基于所述目标牙齿在所述目标二维照片中的位置得到所述第一牙齿中线的位置的步骤包括：
    获取上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中 线，并将所述中线作为所述第一牙齿中线，以得到所述第一牙齿中线的位置。
  6. 根据权利要求2所述的确定方法，其特征在于，所述基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值的步骤包括：
    获取所述目标对象包含有牙冠和牙龈的所述三维牙颌模型；
    在所述三维牙颌模型中，以上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线作为所述第二牙齿中线的位置，并得到对应的所述中切牙的第二尺寸值。
  7. 根据权利要求2所述的确定方法，其特征在于，所述基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值的步骤包括：
    获取所述目标对象包含有牙冠和牙龈的所述三维牙颌模型；
    在所述三维牙颌模型中，以上颚两颗所述中切牙之间的中线或下颚两颗所述中切牙之间的中线作为所述第二牙齿中线的位置，并基于预设尺寸值得到对应的所述中切牙的第二尺寸值。
  8. 根据权利要求1所述的确定方法，其特征在于，所述获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系的步骤包括；
    基于图像识别算法获取所述目标牙齿在所述目标二维照片中的像素值，并将所述像素值作为所述第一尺寸值；
    基于所述目标牙齿对应的所述第一尺寸值和所述第二尺寸值建立所述第一尺寸值与所述第二尺寸值之间的映射关系。
  9. 根据权利要求1所述的确定方法，其特征在于，所述获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系的步骤包括；
    基于所述目标二维照片对应的预设标尺获取所述目标牙齿在所述目标二维照片中的标尺值，并将所述标尺值作为所述第一尺寸值；
    基于所述目标牙齿对应的所述第一尺寸值和所述第二尺寸值建立所述第一尺寸值与所述第二尺寸值之间的映射关系。
  10. 根据权利要求1所述的确定方法，其特征在于，所述基于所述三维牙颌模型中的所述第二牙齿中线的位置和所述第二偏差值得到所述第二面中线在所述三维牙颌模型中的位置的步骤包括：
    基于所述第二偏差值得到所述第二面中线相对于所述第二牙齿中线的相对位置；
    基于所述相对位置得到所述第二面中线在所述三维牙颌模型中的位置。
  11. 一种牙齿矫治位置的确定方法，其特征在于，所述确定方法包括：
    采用如权利要求1-10中任一所述的三维牙颌模型中面中线的确定方法得到目标对象在三维牙颌模型中的第二面中线的位置；
    基于所述第二面中线的位置在所述三维牙颌模型中建立三维坐标系；
    基于所述三维坐标系和所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置确定所述目标对象每个牙齿对应的矫治位置。
  12. 一种牙齿矫治器的制造方法，其特征在于，所述制造方法包括：
    采用如权利要求11所述的牙齿矫治位置的确定方法获得目标对象的牙齿的目标矫治位置；
    基于所述目标矫治位置确定所述目标对象的牙齿矫治器的形状，以制造出具有相应形状的牙齿矫治器。
  13. 一种三维牙颌模型中面中线的确定系统，其特征在于，所述确定系统包括：
    第一获取模块，用于获取目标对象在目标二维照片中的第一面中线的位置和第一牙齿中线的位置；
    其中，所述目标二维照片包括所述目标对象露出目标牙齿的面部；
    第二获取模块，用于基于所述目标牙齿获取所述目标对象在所述三维牙颌模型中的第二牙齿中线的位置，获取所述目标牙齿在所述三维牙颌模型中的第二尺寸值；
    映射关系建立模块，用于获取所述目标牙齿在所述目标二维照片中的第一尺寸值，建立所述第一尺寸值与所述第二尺寸值之间的映射关系；
    第一偏差值获取模块，用于基于所述第一面中线的位置和所述第一牙齿中线的位置得到所述第一面中线和所述第一牙齿中线之间的第一偏差值；
    第二偏差值获取模块，用于基于所述映射关系和所述第一偏差值得到所述目标对象在所述三维牙颌模型中的第二面中线与所述第二牙齿中线的第二偏差值；
    第三获取模块，用于基于所述三维牙颌模型中的所述第二牙齿中线的位置和所述第二偏差值得到所述第二面中线在所述三维牙颌模型中的位置。
  14. 一种牙齿矫治位置的确定系统，其特征在于，所述确定系统包括：
    面中线获取模块，用于采用如权利要求13所述的三维牙颌模型中面中线的确定系统得到目标对象在三维牙颌模型中的面中线的位置；
    坐标系建立模块，用于基于所述面中线的位置在所述三维牙颌模型中建立三维坐标系；
    矫治位置确定模块，用于基于所述三维坐标系和所述目标对象在所述三维牙颌模型中的牙齿中线的位置确定所述目标对象每个牙齿对应的矫治位置。
  15. 一种牙齿矫治器的制造系统，其特征在于，所述制造系统包括：
    矫治位置获取模块，用于采用如权利要求14所述的牙齿矫治位置的确定系统获得目标对象的牙齿的目标矫治位置；
    矫治器制造模块，用于基于所述目标矫治位置确定所述目标对象的牙齿矫治器的形状，以制造出具有相应形状的牙齿矫治器。
  16. 一种电子设备，包括存储器、处理器及存储在存储器上并用于在处理器上运行的计算机程序，其特征在于，所述处理器执行计算机程序时实现如权利要求1-10中任一项所述的三维牙颌模型中面中线的确定方法，或如权利要求11所述的牙齿矫治位置的确定方法，或如权利要求12所述的牙齿矫治器的制造方法。
  17. 一种计算机可读存储介质，其上存储有计算机程序，其特征在于，所述计算机程序被处理器执行时实现如权利要求1-10中任一项所述的三维牙颌模型中面中线的确定方法，或如权利要求11所述的牙齿矫治位置的确定方法，或如权利要求12所述的牙齿矫治器的制造方法。
  18. 一种计算机执行的确定面中线位置的方法，其包括：
    获取患者正脸露齿照片；
    对所述患者正脸露齿照片进行面部特征点识别；
    基于识别出的相应面部特征点，产生面中线；
    基于所述患者正脸露齿照片中的唇内区域图像，产生牙中线；
    基于识别出的相应面部特征点，识别虹膜区域；以及
    基于给定的虹膜直径尺寸、所述识别出的虹膜区域直径的像素数以及所述面中线和牙中线之间的像素数偏差，计算得到所述面中线和牙中线之间的实际偏差。
  19. 如权利要求18所述的确定面中线位置的方法，其特征在于，它还包括：对所述患者正脸露齿照片进行面部区域检测，所述面部特征点识别是针对检测到的面部区域图像进行。
  20. 如权利要求19所述的确定面中线位置的方法，其特征在于，所述面部区域检测采用以下方法之一：基于传统的机器学习方法以及基于深度神经网络的方法。
  21. 如权利要求18所述的确定面中线位置的方法，其特征在于，所述面部特征点检测采用以下方法之一：基于模型的方法、级联形状回归的方法以及基于深度神经网络的方法。
  22. 如权利要求18所述的确定面中线位置的方法，其特征在于，所述面中线是基于所述相应面部特征点拟合得到。
  23. 如权利要求18所述的确定面中线位置的方法，其特征在于，它还包括：基于识别出的相应面部特征点，识别所述唇内区域；以及针对所述唇内区域图像进行图像处理，以产生所述牙中线。
  24. 如权利要求18所述的确定面中线位置的方法，其特征在于，它还包括：基于识别出的相应面部特征点，识别眼部区域；针对所述眼部区域图像进行边缘提取；以及基于提取到的边缘确定所述虹膜区域。
  25. 如权利要求24所述的确定面中线位置的方法，其特征在于，它还包括：基于所述提取到的边缘进行霍夫变换，确定所述虹膜区域。
  26. 如权利要求18所述的确定面中线位置的方法，其特征在于，它还包括：获取所述患者的牙列三维数字模型；确定所述牙中线在所述牙列三维数字模型上的位置；以及基于所述面中线和牙中线之间的实际偏差以及所述中线在所述牙列三维数字模型上的位置，确定所述面中线在所述牙列三维数字模型上的位置。
  27. 一种用于确定面中线位置的计算机系统，其包括存储装置和处理器，所述存储装置存储有一计算机程序，当其被所述处理器执行后，将执行如权利要求18所述的确定面中线位置的方法。
PCT/CN2023/074727 2022-02-08 2023-02-07 面中线及牙齿矫治位置的确定方法、制造方法及系统 WO2023151549A1 (zh)

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN202210117981.2A CN116612505A (zh) 2022-02-08 2022-02-08 确定面中线位置的方法
CN202210117981.2 2022-02-08
CN202211740579.6 2022-12-30
CN202211740579.6A CN116035732A (zh) 2022-12-30 2022-12-30 面中线及牙齿矫治位置的确定方法、制造方法及系统

Publications (1)

Publication Number Publication Date
WO2023151549A1 true WO2023151549A1 (zh) 2023-08-17

Family

ID=87563634

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/074727 WO2023151549A1 (zh) 2022-02-08 2023-02-07 面中线及牙齿矫治位置的确定方法、制造方法及系统

Country Status (1)

Country Link
WO (1) WO2023151549A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5431562A (en) * 1990-01-19 1995-07-11 Ormco Corporation Method and apparatus for designing and forming a custom orthodontic appliance and for the straightening of teeth therewith
CN208511238U (zh) * 2017-09-07 2019-02-19 四川大学 中线测量装置
CN110916779A (zh) * 2019-12-10 2020-03-27 南京医科大学附属口腔医院 一种上颌骨三维精确定位器
CN112515787A (zh) * 2020-11-05 2021-03-19 上海牙典软件科技有限公司 一种三维牙颌数据分析方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5431562A (en) * 1990-01-19 1995-07-11 Ormco Corporation Method and apparatus for designing and forming a custom orthodontic appliance and for the straightening of teeth therewith
CN208511238U (zh) * 2017-09-07 2019-02-19 四川大学 中线测量装置
CN110916779A (zh) * 2019-12-10 2020-03-27 南京医科大学附属口腔医院 一种上颌骨三维精确定位器
CN112515787A (zh) * 2020-11-05 2021-03-19 上海牙典软件科技有限公司 一种三维牙颌数据分析方法

Similar Documents

Publication Publication Date Title
US11957541B2 (en) Machine learning scoring system and methods for tooth position assessment
WO2021004112A1 (zh) 异常人脸检测方法、异常识别方法、装置、设备及介质
US10405796B2 (en) Estimating and predicting tooth wear using intra-oral 3D scans
CN1045129C (zh) 个人识别装置
US11403813B2 (en) Systems and methods for constructing a three-dimensional model from two-dimensional images
JP6872556B2 (ja) 母集団に基づく表面メッシュの再構成
AU2020391247A1 (en) Systems and methods for constructing a three-dimensional model from two-dimensional images
JP4951498B2 (ja) 顔画像認識装置、顔画像認識方法、顔画像認識プログラムおよびそのプログラムを記録した記録媒体
CN111598038B (zh) 脸部特征点检测方法、装置、设备及存储介质
US20230196570A1 (en) Computer-implemented method and system for predicting orthodontic results based on landmark detection
CN108491850B (zh) 三维牙齿网格模型的特征点自动提取方法及装置
US20230225832A1 (en) Photo-based dental attachment detection
WO2020164284A1 (zh) 基于平面检测的活体识别方法、装置、终端及存储介质
WO2023242757A1 (en) Geometry generation for dental restoration appliances, and the validation of that geometry
WO2023151549A1 (zh) 面中线及牙齿矫治位置的确定方法、制造方法及系统
WO2022153340A2 (en) System and method for facial and dental photography, landmark detection and mouth design generation
WO2024141074A1 (zh) 面中线的生成方法、矫治器及其制造方法
CN116035732A (zh) 面中线及牙齿矫治位置的确定方法、制造方法及系统
TWI802309B (zh) 臨床相片智能分類與導正的方法及影像處理裝置、電腦可讀取的記錄媒體
CN113139908A (zh) 一种三维牙列分割与标注方法
CN116612505A (zh) 确定面中线位置的方法
EP4307229A1 (en) Method and system for tooth pose estimation
WO2022173055A1 (ja) 骨格推定方法、装置、プログラム、システム、学習済みモデル生成方法、および学習済みモデル
US20240122463A1 (en) Image quality assessment and multi mode dynamic camera for dental images
WO2024114758A1 (zh) 三维牙齿模型分割方法和装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23752329

Country of ref document: EP

Kind code of ref document: A1