An oral scanner and an oral scanning method

Abstract

This invention relates to the fields of medical devices and information technology, and discloses an oral scanner and an oral scanning method. The adaptive flexible shell is a spherical structure adapted to the geometry of the human oral cavity. Made of a biocompatible elastic material, the shell contains an internal rubber support to define the relative spatial relationships of the imaging array. The shell has a preset Shore hardness to maintain the relative topological reference of the internal components when deformed under pressure. Utilizing the spherical array topology and a 140° wide-angle macro lens, 360° full coverage of the oral cavity wall is achieved, solving the blind spots of traditional handheld devices in the posterior molar region, lingual region, and gingival margin. The adaptive flexible sphere replaces handheld guidance, eliminating the need for users to possess professional anatomical knowledge or steady-state control capabilities, greatly reducing the difficulty of operation in home settings.

Description

Technical Field

[0001] This invention relates to the field of medical devices and information technology, specifically to an oral scanner and an oral scanning method. Background Technology

[0002] Traditional oral scanning primarily relies on handheld intraoral endoscopes. However, due to the narrow oral cavity, handheld devices struggle to reach the posterior molar region and can easily cause nausea in patients. Furthermore, handheld scanning depends on the operator's steady-state control, making it impossible for non-professional users to perform home self-examinations. Regarding illumination, high reflectivity noise caused by saliva in the oral cavity often leads to the loss of image features, resulting in reconstruction failure.

[0003] Therefore, there is an urgent need for a scanning device with flexible adaptive capabilities that can automatically process the light field and compensate for deformation displacement. Summary of the Invention

[0004] This invention provides an oral scanner and an oral scanning method, which solves the problems mentioned in the background art.

[0005] The present invention provides the following technical solution: an oral scanner with an adaptive flexible housing: its shape is set as a spherical structure adapted to the geometry of the human oral cavity, the adaptive flexible housing is made of a biocompatible elastic material, and its interior is provided with a rubber support to define the relative spatial relationship of the imaging array. The adaptive flexible housing has a preset Shore hardness to maintain the relative topological reference of the internal components through the rubber support when deformed under pressure.

[0006] Imaging array: Contains multiple miniature vision sensors, which are arranged in a spherical array on the rubber support and located inside the adaptive flexible shell. The fields of view of each vision sensor overlap to achieve full coverage of the oral cavity wall.

[0007] Dynamic pose sensing unit: integrated inside the adaptive flexible shell, used to acquire the three-dimensional pose data of the imaging array in the oral cavity in real time;

[0008] Complementary lighting unit: configured corresponding to the imaging array, used to provide a controlled light field;

[0009] A data processing terminal is communicatively connected to the imaging array and the dynamic pose sensing unit. The data processing terminal includes a processor and a memory storing a computer program. When the computer program is executed by the processor, it includes:

[0010] Pose-guided initial registration module: configured to construct an initial rotation matrix and displacement vector based on the three-dimensional pose data, and perform coarse registration on the image stream of the vision sensor;

[0011] Curvature-based point cloud registration module: configured to generate local point clouds based on image features, and construct a digital oral cavity 3D model based on point cloud or mesh data.

[0012] As a preferred embodiment of the present invention, the adaptive flexible shell has a Shore hardness between 20A and 40A and a diameter between 25mm and 50mm.

[0013] As a preferred embodiment of the present invention, the surface of the adaptive flexible shell is coated with a nano-antibacterial coating conforming to ISO 10993 standard, and its Shore hardness is further limited to between 20A and 30A, and its diameter is further limited to between 30mm and 45mm.

[0014] As a preferred embodiment of the present invention, the imaging array includes at least 6 visual sensors with macro imaging capabilities, and there is a field-of-view overlap of not less than 20% between adjacent visual sensors.

[0015] As a preferred technical solution of the present invention, the oral scanner has a fully sealed waterproof structure and integrates a wireless inductive charging module and a BLE low-power Bluetooth transmission module.

[0016] An oral cavity scanning method includes the following steps:

[0017] Step S100: Synchronously acquire the raw oral cavity image streams collected by each visual sensor in the imaging array, as well as the three-dimensional pose data collected by the dynamic pose perception unit.

[0018] Step S200: Extract feature points from the original oral cavity image stream and construct initial registration parameters based on the three-dimensional pose data to perform coarse spatial pose registration;

[0019] Step S300: Using a multi-viewpoint geometric algorithm, local point cloud data of the corresponding tooth structure and periodontal tissues are calculated from the original oral image stream;

[0020] Step S400: The iterative nearest point algorithm is used to perform fine registration of the local point cloud data generated by the adjacent visual sensors to construct a digital oral cavity 3D model under a unified world coordinate system;

[0021] Step S500: Based on a preset standard healthy oral cavity model database, perform image parameter analysis on the digital oral cavity 3D model using a preset feature extraction model, and generate spatial identification data corresponding to texture difference features or color deviation parameters in the digital oral cavity 3D model.

[0022] As a preferred embodiment of the present invention, step S100 further includes a dynamic supplementary lighting adjustment step:

[0023] Real-time monitoring of the grayscale gradient in the high-reflectance region of the raw oral imaging stream;

[0024] When the grayscale value of a local area exceeds the preset saturation threshold, the duty cycle of the drive current of the corresponding visual sensor supplementary light unit is reduced by pulse width modulation (PWM).

[0025] As a preferred embodiment of the present invention, in step S400, the iterative nearest point algorithm includes:

[0026] The initial pose calculated from the 3D pose data is used as the initial value to dynamically correct the relative external parameter changes between the visual sensors during the registration process. Feature points with a curvature change rate greater than a preset value in the overlapping area are selected as registration sources. The calculation is terminated when the mean square error of the iterative registration is less than the preset convergence threshold.

[0027] The present invention has the following beneficial effects:

[0028] This oral scanner and scanning method utilize a spherical array topology and a 140° wide-angle macro lens to achieve 360° full coverage of the oral cavity wall, solving the blind spots of traditional handheld devices in the posterior molar region, lingual region, and gingival margin. By replacing handheld guidance with an adaptive flexible sphere-based sublingual approach, it eliminates the need for users to possess professional anatomical knowledge or steady-state control capabilities, greatly reducing the operational difficulty in home settings. Employing a "soft-on-hard" physical structure combined with a dynamic extrinsic parameter correction algorithm, it maintains a reconstruction accuracy within 0.12mm even when the flexible shell with a Shore hardness of 25A undergoes nonlinear deformation under a 15N biting force, ensuring the medical-grade reference value of the scanned data. Dynamic PWM dimming technology suppresses high-light reflection caused by saliva in real time, ensuring clear extraction of tooth surface texture while effectively preventing feature loss in 3D modeling. Based on the differential identification logic of a standard database, complex pathological analysis is transformed into intuitive 3D spatial data identification, avoiding the legal risks of remote medical diagnosis while providing users with highly sensitive health warnings. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the structure of the present invention;

[0030] Figure 2 This is a schematic diagram of the oral cavity scanning method of the present invention.

[0031] In the figure: 1. Adaptive flexible shell; 2. Imaging array; 3. Light supplement unit; 4. Dynamic pose sensing unit. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Please see Figure 1 - Figure 2 An oral scanner, with an adaptive flexible housing 1: its shape is set as a spherical structure adapted to the geometry of the human oral cavity, the adaptive flexible housing 1 is made of a biocompatible elastic material, and its interior is provided with a rubber support to define the relative spatial relationship of the imaging array 2. The adaptive flexible housing 1 has a preset Shore hardness to maintain the relative topological reference of the internal components through the rubber support when deformed under pressure.

[0034] Imaging array 2: Contains multiple miniature vision sensors, which are arranged in a spherical array on a rubber support and located inside the adaptive flexible shell 1. The fields of view of each vision sensor overlap to achieve full coverage of the oral cavity wall.

[0035] Dynamic pose sensing unit 4: integrated inside the adaptive flexible shell 1, used to acquire the three-dimensional pose data of the imaging array 2 in the oral cavity in real time;

[0036] Complementary lighting unit 3: Set to correspond to imaging array 2, used to provide a controlled light field;

[0037] The data processing terminal is communicatively connected to the imaging array 2 and the dynamic pose sensing unit 4. The data processing terminal includes a processor and a memory storing a computer program. When the computer program is executed by the processor, it includes:

[0038] Pose-guided initial registration module: configured to construct an initial rotation matrix and displacement vector based on 3D pose data, and perform coarse registration on the image stream of the vision sensor;

[0039] Curvature-based point cloud registration module: configured to generate local point clouds based on image features, and construct a digital oral cavity 3D model based on point cloud or mesh data.

[0040] The adaptive flexible shell 1 has a Shore hardness between 20A and 40A and a diameter between 25mm and 50mm.

[0041] The surface of the adaptive flexible shell 1 is coated with a nano antibacterial coating that conforms to ISO10993 standards, with its Shore hardness further limited to between 20A and 30A and its diameter further limited to between 30mm and 45mm.

[0042] The imaging array 2 contains at least 6 visual sensors with macro imaging capabilities, and there is a field of view overlap of no less than 20% between adjacent visual sensors.

[0043] The dental scanner has a fully sealed waterproof structure and integrates a wireless inductive charging module and a BLE low-power Bluetooth transmission module.

[0044] An oral cavity scanning method includes the following steps:

[0045] Step S100: Synchronously acquire the raw oral cavity image streams collected by each visual sensor in the imaging array 2, as well as the three-dimensional pose data collected by the dynamic pose perception unit 4.

[0046] Step S200: Extract feature points from the original oral cavity image stream and construct initial registration parameters based on the three-dimensional pose data to perform coarse spatial pose registration;

[0047] Step S300: Using a multi-viewpoint geometric algorithm, local point cloud data of the corresponding tooth structure and periodontal tissues are generated from the original oral image stream;

[0048] Step S400: The iterative nearest point algorithm is used to perform fine registration of the local point cloud data generated by adjacent visual sensors to construct a digital oral cavity 3D model under a unified world coordinate system;

[0049] Step S500: Based on a preset standard healthy oral cavity model database, perform image parameter analysis on the digital oral cavity 3D model using a preset feature extraction model, and generate spatial identification data corresponding to texture difference features or color deviation parameters in the digital oral cavity 3D model.

[0050] Step S100 also includes a dynamic fill light adjustment step:

[0051] Real-time monitoring of the grayscale gradient in the high-reflectance region of the raw oral imaging stream;

[0052] When the gray value of a local area exceeds the preset saturation threshold, the duty cycle of the drive current of the corresponding visual sensor supplementary light unit 3 is reduced by pulse width modulation (PWM).

[0053] In step S400, the iterative nearest point algorithm includes:

[0054] The initial pose calculated from the 3D pose data is used as the initial value to dynamically correct the relative extrinsic parameter changes between the visual sensors during the registration process. Feature points with a curvature change rate greater than a preset value in the overlapping region are selected as registration sources. The calculation terminates when the mean square error of the iterative registration is less than a preset convergence threshold.

[0055] In this embodiment, the outer diameter of the adaptive flexible housing 1 of the dental scanner is set to 38 mm. The adaptive flexible housing 1 is made of medical-grade liquid silicone LSR, and its Shore A hardness is controlled at 25A.

[0056] The internal rubber support is made of high-modulus polycarbonate (PC) with an elastic modulus ≥2300MPa, injection molded. Imaging array 2 consists of eight 1 / 4-inch CMOS image sensors, each module measuring 5mm × 5mm × 3.5mm. Each sensor is rigidly bonded to the recessed mounting groove of the rubber support using UV adhesive.

[0057] A buffer gap filled with damping oil is provided between the support and the adaptive flexible housing 1. When the adaptive flexible housing 1 is subjected to a bite pressure of 15N and deforms, the inner wall of the adaptive flexible housing 1 contacts the rubber support and generates friction. Because the rigidity of the rubber support is greater than that of the adaptive flexible housing 1, the optical axis deflection angle of the sensor is limited to within ±1.5°. The dynamic pose sensing unit 4 is a low-drift six-axis IMU with a sampling frequency of 200Hz. It is connected to the main control board through a flexible circuit board (FPC) to collect the translational and rotational increments of the housing in the oral cavity in real time.

[0058] In this embodiment, the multi-viewpoint geometry algorithm of step S300 is specifically implemented as follows:

[0059] Parallax calculation: The data processing terminal calls the pre-calibrated camera intrinsic parameter matrix K from memory. For the overlapping region of any two adjacent cameras C1 and C2, ORB feature point pairs are extracted. and .

[0060] 3D back projection: Utilizing the principle of triangulation and combining the fundamental matrix F in the initial registration parameters, the 3D coordinates are calculated. :

[0061]

[0062] Where f is the effective focal length, which is 2.8mm in this embodiment; B is the baseline distance between the optical centers of adjacent cameras, which is 12.5mm in this embodiment; and d is the pixel parallax.

[0063] Point cloud density control: For areas with rich tooth surface texture, such as the edge of the tooth cusp, the system generates dense point clouds in steps of 0.1mm, while for smooth enamel surfaces, point cloud sparse processing is used to reduce memory load.

[0064] Since the occlusion causes the stent to exhibit slight submicron-level bending, this embodiment introduces dynamic extrinsic parameter correction in step S400:

[0065] Pose coarse registration step S200: Using the quaternion Q acquired by the IMU at the current time, generate a rotation matrix through Rodrigues transformation. Project the local point clouds of adjacent cameras onto the world coordinate system.

[0066] ICP fine-tuning is a further refinement of step S400: with These are the initial values. To compensate for changes in the relative positions between cameras caused by shell deformation, the algorithm uses the originally fixed extrinsic parameter matrix. Set as a variable to be optimized.

[0067] Objective function optimization:

[0068]

[0069] in, The curvature weighting factor is used. The system's convergence threshold is set to root mean square error (RMSE) ≤ 0.05. Experimental data show that, with a 10% deformation of the 25A hardness adaptive flexible shell 1, this method can reduce the global error of dental arch reconstruction from 1.5 mm to 0.12 mm.

[0070] Regarding step S500: The system stores a "standard healthy oral cavity model database", which contains color mapping values ​​and morphological parameters of multiple sets of standard anatomical structures.

[0071] Color space conversion: The processor converts the surface texture of the reconstructed 3D model from RGB to CIELAB color space.

[0072] Difference detection: The processor calls upon a standard healthy oral cavity model database to calculate the difference in the area to be tested. The component represents the Euclidean distance between the yellow offset and the standard features of the corresponding anatomical location in the database. .

[0073] Spatial identifier generation: If For clinically obvious calculus deposits or early caries demineralization, the processor writes a "Warning" flag at the corresponding vertex index of the 3D mesh.

[0074] Terminal display: The APP reads the model file with the flag and renders a semi-transparent red bubble prompt at the corresponding position.

[0075] It should be noted that a gyroscope, accelerometer, Bluetooth module and micro lithium battery can also be installed inside the adaptive flexible housing 1. The components are connected by a flexible circuit board. The adaptive flexible housing 1 will not affect the circuit connection due to deformation in the conformal state.

[0076] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0077] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended technical solutions and their equivalents.

Claims

1. An oral scanner, characterized in that, include: Adaptive flexible shell (1): Its shape is set as a spherical structure that adapts to the geometry of the human oral cavity. The adaptive flexible shell (1) is made of biocompatible elastic material and has a rubber support inside to limit the relative spatial relationship of the imaging array (2). The adaptive flexible shell (1) has a preset Shore hardness. When the adaptive flexible housing (1) is subjected to a biting pressure of 15N and deforms, the inner wall of the adaptive flexible housing (1) contacts the rubber bracket and generates friction. Since the rigidity of the rubber bracket is greater than that of the adaptive flexible housing (1), the optical axis deflection angle of the sensor is limited to within ±1.5°. This is to maintain the relative topological reference of the internal components through the rubber bracket when deformation occurs under pressure. Imaging array (2): contains multiple miniature vision sensors, which are arranged in a spherical array on the rubber support and located inside the adaptive flexible shell (1). The field of view of each vision sensor overlaps with each other to achieve full coverage of the oral cavity wall. Dynamic pose sensing unit (4): integrated inside the adaptive flexible shell (1), used to collect the three-dimensional pose data of the imaging array (2) in the oral cavity in real time; Complementary lighting unit (3): Set in accordance with the imaging array (2), used to provide a controlled light field; A data processing terminal is communicatively connected to the imaging array (2) and the dynamic pose sensing unit (4). The data processing terminal includes a processor and a memory storing a computer program. When the computer program is executed by the processor, it includes: Pose-guided initial registration module: configured to construct an initial rotation matrix and displacement vector based on the three-dimensional pose data, and perform coarse registration on the image stream of the vision sensor; The curvature-based point cloud registration module is configured to generate local point clouds based on image features. It uses the initial pose calculated from the 3D pose data as the initial value to dynamically correct the relative extrinsic parameter changes between the visual sensors during the registration process. It selects feature points with a curvature change rate greater than a preset value in the overlapping area as registration sources. The calculation is terminated when the mean square error of the iterative registration is less than a preset convergence threshold, so as to construct a digital oral 3D model based on point cloud or mesh data.

2. The oral scanner of claim 1, wherein: The adaptive flexible shell (1) has a Shore hardness between 20A and 40A and a diameter between 25mm and 50mm.

3. The oral scanner of claim 2, wherein: The surface of the adaptive flexible shell (1) is covered with a nano antibacterial coating conforming to ISO 10993 standard, with its Shore hardness further limited to between 20A and 30A and its diameter further limited to between 30mm and 45mm.

4. The oral scanner of claim 1, wherein: The imaging array (2) includes at least 6 visual sensors with macro imaging capabilities, and there is a field of view overlap of not less than 20% between adjacent visual sensors.

5. The oral scanner of claim 1, wherein: The oral scanner has a fully sealed waterproof structure and integrates a wireless inductive charging module and a BLE low-power Bluetooth transmission module.

6. A method of oral scanning, using the oral scanner of any one of claims 1-5, wherein, Includes the following steps: Step S100: Synchronously acquire the original oral cavity image stream collected by each visual sensor in the imaging array (2) and the three-dimensional pose data collected by the dynamic pose perception unit (4), and monitor the grayscale gradient of the high-light reflection area in the original oral cavity image stream in real time. When the gray value of a local area exceeds the preset saturation threshold, the duty cycle of the driving current of the corresponding visual sensor side-light unit (3) is reduced by pulse width modulation (PWM). Step S200: Extract feature points from the original oral cavity image stream and construct initial registration parameters based on the three-dimensional pose data to perform coarse spatial pose registration; Step S300: Using a multi-viewpoint geometric algorithm, local point cloud data of the corresponding tooth structure and periodontal tissues are calculated from the original oral image stream; Step S400: The iterative nearest point algorithm is used to perform fine registration of the local point cloud data generated by the adjacent vision sensors. The initial pose calculated from the three-dimensional pose data is used as the initial value to dynamically correct the relative external parameter changes between the vision sensors during the registration process. Feature points with a curvature change rate greater than a preset value in the overlapping area are selected as registration sources. The calculation is terminated when the mean square error of the iterative registration is less than a preset convergence threshold, and a digital oral three-dimensional model under a unified world coordinate system is constructed. Step S500: Based on a preset standard healthy oral cavity model database, perform image parameter analysis on the digital oral cavity 3D model using a preset feature extraction model, and generate spatial identification data corresponding to texture difference features or color deviation parameters in the digital oral cavity 3D model.

Images