Dental 3D Scanning Method and System Based on High-Efficiency Laser Source

Abstract

This invention belongs to the technical field of dental 3D modeling, specifically relating to a dental 3D scanning method and system using a high-efficiency laser light source. The method acquires images of a planar calibration plate, a stepped calibration plate, and a cylindrical arc calibration plate to establish a calibration parameter table. During scanning, uniform reference frames and low-brightness stripe frames are acquired to generate specular reflection masks, occlusion masks, gingival masks, and tooth body masks. A first stripe frame is acquired for the tooth body region, and a second stripe frame with low peak current is acquired for the reflective and occluded edge regions. Spatial depth values ​​are then selected by combining the auxiliary reference frames to generate tooth surface point cloud fragments with region origin markers. Finally, based on pose, cusp edges, and adjacent tooth contact area features, a 3D surface model of the dental arch is obtained through fusion. This invention can reduce the impact of reflection and occlusion on the 3D reconstruction of the tooth surface.

Description

Technical Field

[0001] This invention belongs to the technical field of dental 3D modeling, specifically relating to a dental 3D scanning method and system using a high-efficiency laser light source. Background Technology

[0002] Dental 3D scanning technology is mainly used to acquire three-dimensional morphological data of intraoral structures such as dentition, tooth preparations, gingival margins, occlusal fissures, and contact areas between adjacent teeth. This data provides fundamental information for digital restorations, orthodontic design, implant guide design, occlusal analysis, and chairside fabrication. Intraoral scanning typically requires structured light projection, image acquisition, and 3D reconstruction within a confined oral space, placing high demands on light source brightness, fringe stability, exposure control, local detail resolution, and point cloud stitching continuity.

[0003] Current dental 3D scanning technologies mostly employ white light structured light, LED stripe projection, single-band laser stripe projection, or multi-frame phase-shift acquisition methods. These methods involve acquiring images of tooth surface stripes via a camera, followed by phase analysis and triangulation to obtain a 3D model of the dental arch. These techniques can achieve basic scanning in conventional tooth surface areas.

[0004] However, in a real oral environment, enamel, restorations, moist tooth surfaces, and areas with residual saliva are prone to specular reflection, leading to local image saturation, fringe center shift, or fringe breakage. Furthermore, occlusion dark areas easily appear in the contact areas between adjacent teeth, the cusp walls, and the pits and fissures, making phase unfolding lack a stable basis. Some solutions address this by reducing overall exposure, increasing multi-frame acquisition, or post-processing point cloud smoothing. However, reducing overall exposure weakens the fringe contrast in ordinary tooth areas, increasing multi-frame acquisition exacerbates the impact of pose changes during handheld scanning, and post-processing point cloud smoothing may weaken details such as cusp margins, shoulder lines, and proximal boundaries, making it difficult to simultaneously achieve reflection suppression, dark area compensation, and preservation of tooth details.

[0005] Therefore, it is necessary to propose a dental 3D scanning method and system suitable for high-efficiency laser light sources to address practical acquisition problems such as tooth surface reflection, local occlusion, and unstable stripes in dental intraoral scanning, so as to improve the stability and integrity of the dental arch 3D data acquisition process. Summary of the Invention

[0006] To address the above problems, the present invention aims to provide a dental three-dimensional scanning method using a high-efficiency laser light source, comprising the following steps: S1. Establishment of light source stripes: A laser stripe light source is set inside the intraoral scanning head. The laser stripe light source includes a first laser diode group, a second laser diode group, a collimating lens group, an optical fiber coupler, an integrating homogenizing rod, a movable scattering plate, and a diffraction stripe element. The diffraction stripe element includes a stripe diffraction region and a non-stripe light transmission region. The first laser diode group and the second laser diode group output visible lasers of different wavelengths. After collimation, coupling, and integrating homogenization, the two visible lasers enter the stripe diffraction region to form structured light stripes projected onto the tooth surface area. S2. Optical calibration establishment: Acquire fringe images of a planar calibration plate with a planar grid pattern, a stepped calibration plate with at least three known height steps, and a cylindrical arc surface calibration plate with a known radius of curvature. Establish a calibration parameter table between projection coordinates, imaging coordinates, scanning head pose coordinates, phase value, and spatial depth value. Write the pulse current, pulse width, on / off sequence, and displacement of the movable scattering sheet into the calibration parameter table. S3. Tooth surface region recognition: Collect uniform reference frames and low-brightness stripe frames of the dentition to be scanned. Based on the saturated pixels, dark area pixels, gingival color pixels and tooth boundary pixels in the uniform reference frame, and combined with the stripe continuity in the low-brightness stripe frame, generate specular reflection mask, occlusion mask, gingival mask and tooth mask. S4. Segmented stripe acquisition: The first stripe frame is acquired for the corresponding area of ​​the tooth mask, the second stripe frame is acquired for the outer expansion area of ​​the specular reflection mask and the edge area of ​​the mask, and an auxiliary reference frame is acquired when the laser passes through the non-striped light-transmitting area. The peak current of the laser pulse in the second stripe frame is less than the peak current of the laser pulse in the first stripe frame. S5. 3D point generation: Extract the fringe center, unwrap the phase and perform triangulation on the first and second fringe frames respectively. Based on the auxiliary reference frame and each mask, select the spatial depth value under the same imaging coordinates to generate a tooth surface point cloud fragment with region source markers. S6. Tooth row model fusion: The continuously acquired tooth surface point cloud fragments are registered according to the scanning head pose coordinates, tooth cusp edge features, adjacent tooth contact area features, and region source markers. Non-tooth body points corresponding to the gingival mask are deleted, and tooth body boundary points in adjacent tooth surface point cloud fragments are connected to obtain the three-dimensional surface model of the tooth row to be scanned.

[0007] As a preferred technical solution, in step S1, the output wavelength of the first laser diode group is 450nm to 470nm, and the output wavelength of the second laser diode group is 510nm to 540nm. The first laser diode group and the second laser diode group are mounted on the same heat-conducting base. The collimating lens group is respectively positioned to correspond to the two laser diode groups. The fiber coupler has two light-incident ends and one light-outcident end. The light-incident end of the integrating light-diffusing rod is coaxially positioned with the light-outcident end of the fiber coupler. The light-outcident end of the integrating light-diffusing rod is coaxially positioned with the light-incident surface of the diffraction fringe element.

[0008] As a preferred technical solution, in step S1, the movable scattering sheet is disposed between the integrating homogenizing rod and the diffraction fringe element; the movable scattering sheet reciprocates in a direction perpendicular to the fringe extension direction during a single frame exposure time, with a displacement amplitude of 0.05mm to 0.8mm and a moving frequency of 80Hz to 800Hz. When acquiring the first and second stripe frames, the displacement phase of the movable scatterer is written into the frame header data of the corresponding stripe frame.

[0009] As a preferred technical solution, in step S2, the planar calibration plate is used to determine the mapping relationship between imaging coordinates and planar grid coordinates; the adjacent steps in the stepped calibration plate have a known height difference, which is used to determine the correspondence between phase values ​​and spatial depth values; The cylindrical arc surface calibration plate is used to record the distortion parameters of the field of view at the edge of the intraoral scanning window; The calibration parameter table includes fringe period, fringe tilt angle, camera intrinsic parameters, projection extrinsic parameters, phase-depth conversion coefficient, edge field-of-view distortion parameters, and movable scatterer displacement phase data.

[0010] As a preferred technical solution, in step S3, the uniform reference frame is acquired when the laser passes through the non-striped light-passing area of ​​the diffraction stripe element, and the low-brightness stripe frame is acquired under the condition that the pulse peak current of the first stripe frame is lower than that of the first stripe frame in the first laser diode group or the second laser diode group.

[0011] As a preferred technical solution, continuous pixel regions in the uniform reference frame whose gray values ​​reach the camera saturation threshold are recorded as specular reflection candidate regions, continuous pixel regions whose gray values ​​are lower than the dark area threshold and are located inside the tooth boundary are recorded as occlusion candidate regions, and continuous pixel regions whose chromaticity values ​​fall within the gingival chromaticity range are recorded as gingival candidate regions. The stripe continuity in the low-brightness stripe frame is then combined to obtain each mask.

[0012] As a preferred technical solution, in step S4, the outer expansion area of ​​the specular reflection mask is obtained by extending the boundary of the specular reflection mask outward by 3 to 15 pixels, and the edge area of ​​the occlusion mask is obtained by extending the boundary of the occlusion mask outward by 2 to 10 pixels.

[0013] As a preferred technical solution, the first stripe frame uses a first pulse peak current and a first pulse width, and the second stripe frame uses a second pulse peak current and a second pulse width. The second pulse peak current is 20% to 70% of the first pulse peak current, and the second pulse width is 80% to 160% of the first pulse width.

[0014] As a preferred technical solution, in step S5, stripe center extraction includes grayscale normalization, stripe direction filtering, sub-pixel center positioning, and center line numbering. Phase unfolding involves determining the phase order between adjacent stripes based on the centerline number, and converting the phase order into spatial depth values ​​using a calibration parameter table; When the same imaging coordinates simultaneously contain the spatial depth values ​​of the first fringe frame and the second fringe frame, the saturation pixel ratio, fringe center residual, and neighborhood depth dispersion of the corresponding pixels are compared, and the spatial depth value with the smaller value of two or three of them is retained.

[0015] As a preferred technical solution, in step S6, the region source marking includes ordinary tooth marking, specular reflection edge marking, and occlusion edge marking; When registering point cloud fragments on the tooth surface, a first registration weight is set for the three-dimensional points corresponding to ordinary tooth markings, a second registration weight is set for the three-dimensional points corresponding to the mirror reflection edge markings, and a third registration weight is set for the three-dimensional points corresponding to the occlusion edge markings. The second and third registration weights are both less than the first registration weight. After registration, the three-dimensional points corresponding to the gingival mask, isolated three-dimensional points, and three-dimensional points that exceed the dental arch boundary are deleted.

[0016] The present invention also provides a dental three-dimensional scanning system with a high-efficiency laser light source for implementing the method described above, including an intraoral scanning head, comprising a housing, an intraoral scanning window, an imaging camera, a pose sensor, and a laser stripe light source, wherein the laser stripe light source comprises a first laser diode group, a second laser diode group, a collimating lens group, an optical fiber coupler, an integrating homogenizing rod, a movable scattering sheet, and a diffraction stripe element; The light source driving module is connected to two sets of laser diodes and a movable scattering sheet, respectively. The optical calibration module generates a calibration parameter table; The region recognition module outputs a specular reflection mask, an occlusion mask, a gingival mask, and a tooth structure mask. The partition acquisition module controls the acquisition of the first stripe frame, the second stripe frame, and the auxiliary reference frame according to each mask; The 3D calculation module generates tooth surface point cloud fragments with region source markers; The model fusion module generates a three-dimensional surface model based on the scanning head pose coordinates, tooth cusp edge features, adjacent tooth contact area features, and region source markers.

[0017] As a preferred technical solution, the partition acquisition module includes: a general tooth acquisition unit, a reflective compensation acquisition unit, and an auxiliary reference acquisition unit; The ordinary tooth acquisition unit is connected to the data terminal of the tooth mask, the reflective compensation acquisition unit is connected to the data terminals of the specular reflection mask and the occlusion mask respectively, and the auxiliary reference acquisition unit is connected to the control terminal of the corresponding non-striped light transmission area in the light source driving module.

[0018] As a preferred technical solution, the three-dimensional calculation module includes a stripe center extraction unit, a phase unrolling unit, a triangulation unit, and a point cloud discarding unit; the model fusion module includes a pose coarse registration unit, a feature fine registration unit, a non-dental point deletion unit, and a mesh connection unit. The mesh connection unit receives the deleted tooth surface point cloud fragments and outputs a three-dimensional surface model.

[0019] Beneficial effects In this invention, a uniform reference frame and a low-brightness stripe frame are first acquired during the scanning process. Different masks are generated based on saturated pixels, dark area pixels, gingival color pixels, tooth boundary pixels, and stripe continuity. This allows the high-reflectivity areas, occluded areas, gingival areas, and tooth areas of the tooth surface to be distinguished and processed before stripe acquisition, avoiding the use of the same exposure conditions for the entire image and reducing the interference of local saturation or dark areas on the depth resolution of the tooth surface.

[0020] This invention acquires a first stripe frame corresponding to the tooth mask area, and a second stripe frame with low peak current for the outer expansion area of ​​the specular reflection mask and the edge area of ​​the mask. It also combines an auxiliary reference frame to participate in the judgment, so that the reflective edge of the enamel, the contact area between adjacent teeth and the pit and fissure shadow area can obtain compensation stripe data that can be used for phase unfolding, reducing the loss of depth values ​​caused by reflective saturation, local occlusion and stripe breakage.

[0021] In the three-dimensional point generation and dental arch model fusion stages, this invention introduces the saturation pixel ratio, stripe center residual, and neighborhood depth discretization to select the spatial depth value under the same imaging coordinates. It also combines ordinary tooth body markers, specular reflection edge markers, and occlusion edge markers to set registration weights, so that tooth surface point cloud fragments from different sources participate in registration and deletion according to unified rules, thereby improving the boundary continuity and local morphological consistency of the three-dimensional surface model of the dental arch. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0023] To enhance understanding of the present invention, the present invention will be further described in detail below with reference to embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0024] Example 1 This embodiment provides a dental 3D scanning method using a high-efficiency laser light source, used to perform 3D scanning of a single tooth, a portion of the dentition, the maxillary dentition, or the mandibular dentition within the patient's oral cavity. The method obtains a 3D surface model including cusps, pits and fissures, adjacent tooth contact areas, gingival margins, tooth boundaries, and the margins of local prepared structures. This embodiment uses a handheld intraoral scanning device as an example. During scanning, the operator inserts the intraoral scanning head into the oral cavity, aligning the intraoral scanning window with the area to be scanned, and slowly moves the intraoral scanning head along the dental arch direction. The process sequentially completes the establishment of light source stripes, optical calibration, tooth surface area identification, sectional stripe acquisition, 3D point generation, and dentition model fusion. (Reference) Figure 1 As shown, the method includes the following steps: S1. Establishment of light source stripes: A laser stripe light source is set inside the intraoral scanning head. The laser stripe light source includes a first laser diode group, a second laser diode group, a collimating lens group, an optical fiber coupler, an integrating uniform rod, a movable scattering sheet, and diffraction stripe elements.

[0025] The first laser diode group outputs visible laser light in the first band, preferably with an output wavelength of 450nm to 470nm; the second laser diode group outputs visible laser light in the second band, preferably with an output wavelength of 510nm to 540nm. The first and second laser diode groups are mounted on the same heat-conducting base and correspond to the collimating lenses in the collimating lens group, respectively, so that the two visible laser beams form collimated beams with the same direction before entering the fiber coupler.

[0026] The fiber coupler has two input ends and one output end. First-band visible laser and second-band visible laser enter the fiber coupler from the two input ends respectively, and a mixed beam is output from the same output end. The mixed beam enters an integrating homogenizer, whose input end is coaxially aligned with the output end of the fiber coupler, and whose output end is coaxially aligned with the incident surface of the diffraction fringe element. This ensures that the two visible laser beams are homogenized before entering the diffraction fringe element.

[0027] The diffraction fringe element includes a fringe diffraction region and a non-fringe light transmission region. When the mixed beam enters the fringe diffraction region, it forms structured light fringes on the tooth surface. When the mixed beam enters the non-fringe light transmission region, it forms uniform illumination on the tooth surface without fringe modulation.

[0028] A movable diffuser is positioned between the integrating homogenizing rod and the diffraction fringe element, or between the diffraction fringe element and the intra-aperture scanning window. The movable diffuser reciprocates within a single frame exposure time along a direction perpendicular to the fringe extension direction, with a preferred displacement amplitude of 0.05 mm to 0.8 mm and a preferred movement frequency of 80 Hz to 800 Hz. For example, in this embodiment, the displacement amplitude of the movable diffuser is set to 0.2 mm, and the movement frequency is set to 300 Hz. When acquiring the first and second fringe frames, the displacement phase of the movable diffuser is written into the frame header data of the corresponding fringe frame, and the frame header data is bound to the imaging timestamp of the corresponding fringe frame. This ensures that subsequent fringe center extraction and spatial depth calculations correspond to the actual light source state of each frame image.

[0029] S2. Optical Calibration Establishment: Before the formal oral scan, the intraoral scanning head is optically calibrated. In this embodiment, stripe images are sequentially acquired from a planar calibration plate with a planar grid pattern, a stepped calibration plate with at least three known height steps, and a cylindrical arc surface calibration plate with a known radius of curvature.

[0030] The planar calibration plate is used to determine the mapping relationship between imaging coordinates and planar grid coordinates; the stepped calibration plate has a known height difference between adjacent steps, which is used to determine the correspondence between phase values ​​and spatial depth values; the cylindrical arc surface calibration plate is used to record the distortion parameters of the field of view at the edge of the intraoral scanning window.

[0031] Specifically, the planar calibration plate is first placed within the working distance range of the intraoral scanning head. The laser stripe light source is controlled to project structured light stripes through the stripe diffraction region, and the stripe image of the planar calibration plate is acquired by the imaging camera. Based on the positions of the grid intersections or grid lines in the planar grid pattern, the mapping relationship between the imaging coordinates and the planar grid coordinates is determined, and the camera intrinsic parameters and projection extrinsic parameters are recorded.

[0032] Subsequently, fringe images of a stepped calibration plate are acquired. The stepped calibration plate can be set to three to six height steps, with the height difference between adjacent steps being a known value, such as 0.2 mm, 0.5 mm, or 1.0 mm. By comparing the fringe phase values ​​corresponding to different step regions, the conversion relationship between phase values ​​and spatial depth values ​​is obtained, and phase-depth conversion coefficients are generated. Next, fringe images of a cylindrical arc surface calibration plate are acquired. The cylindrical arc surface calibration plate has a known radius of curvature, and its arc surface region covers the central and edge fields of view of the intraoral scanning window. By measuring the deviation between the fringe positions in the arc surface image and the theoretical arc surface positions, the distortion parameters of the edge field of view of the intraoral scanning window are recorded.

[0033] After completing the above calibration, a calibration parameter table is established. This table includes the correspondence between projection coordinates, imaging coordinates, scanning head pose coordinates, phase values, and spatial depth values. It also includes fringe period, fringe tilt angle, camera intrinsic parameters, projection extrinsic parameters, phase-depth conversion coefficient, edge field-of-view distortion parameters, pulse current, pulse width, on / off sequence, and movable scatterer displacement. For different light source operating states, such as the first laser diode group operating alone, the second laser diode group operating alone, or both groups operating alternately or simultaneously, corresponding data items can be set in the calibration parameter table, allowing subsequent fringe frames to call matching calibration parameters based on their frame header data.

[0034] S3. Tooth surface area recognition: Align the intraoral scanning head with the tooth row to be scanned, first collect a uniform reference frame in the non-striped light-transmitting area of ​​the laser passing through the diffraction stripe element, and then collect a low-brightness stripe frame under the condition of a pulse peak current lower than that of the first stripe frame.

[0035] The uniform reference frame is used to reflect the brightness and chromaticity distribution of the tooth surface area under stripe-free modulated illumination, while the low-brightness stripe frame is used to determine the continuity of stripes on the tooth surface under low light conditions.

[0036] In a uniform reference frame, pixel grayscale and chromaticity values ​​are analyzed. Continuous pixel regions with grayscale values ​​reaching the camera saturation threshold are designated as specular reflection candidate regions; continuous pixel regions with grayscale values ​​below the dark area threshold and located inside the tooth boundary are designated as occlusion candidate regions; continuous pixel regions with chromaticity values ​​falling within the gingival chromaticity range are designated as gingival candidate regions; and regions within the tooth contour that do not belong to the gingival candidate regions are designated as tooth candidate regions.

[0037] The candidate regions are then corrected based on the continuity of the stripes in the low-brightness stripe frame: if the center of the stripe is missing, the stripe width is abnormally increased, or the stripe numbering is interrupted in the specular reflection candidate region, the region is identified as a specular reflection mask; if the stripe contrast is low and the adjacent stripe numbers are not continuous in the occlusion candidate region, the region is identified as an occlusion mask; if the gingival candidate region is continuous with the outer region of the tooth boundary, it is identified as a gingival mask; and the region outside the gingival mask where continuous stripes can be detected is identified as a tooth mask.

[0038] This step divides the dentition to be scanned into four regions: the ordinary tooth structure region, the specular reflection region, the occluded region, and the gingival region. The ordinary tooth structure region is mainly used for routine stripe acquisition and 3D reconstruction; the specular reflection region mainly corresponds to the enamel highlight areas, moist tooth surfaces, or reflective areas of restorations; the occluded region mainly corresponds to the contact areas between adjacent teeth, the cusp lateral walls, and the pit and fissure shadow areas; the gingival region is used for subsequent deletion of non-dental points. This division process is completed before stripe acquisition, allowing subsequent acquisition parameters to be set separately for different regions.

[0039] S4. Segmented Stripe Acquisition: Segmented acquisition is performed based on the tooth structure mask, specular reflection mask, and occlusion mask. The first stripe frame is acquired for the corresponding area of ​​the tooth structure mask, and the first stripe frame uses the first pulse peak current and the first pulse width.

[0040] For example, the first pulse peak current is set to 80% of the rated pulse peak current of the laser diode, and the first pulse width is set to 120 μs. The first fringe frame is mainly used to obtain the fringe center position and phase value of the ordinary tooth area.

[0041] Extend the boundary of the specular reflection mask outward by 3 to 15 pixels to form the outer extension region of the specular reflection mask; extend the boundary of the occlusion mask outward by 2 to 10 pixels to form the edge region of the occlusion mask. The outer extension region of the specular reflection mask and the edge region of the occlusion mask are used to cover the transition boundary between the reflective or dark area and the ordinary tooth area. A second stripe frame is acquired for the above areas, and the second stripe frame uses the second pulse peak current and the second pulse width.

[0042] The peak current of the second pulse is less than the peak current of the first pulse, preferably 20% to 70% of the peak current of the first pulse; the width of the second pulse is 80% to 160% of the width of the first pulse. For example, the peak current of the second pulse is set to 50% of the peak current of the first pulse, and the width of the second pulse is set to 120% of the width of the first pulse, in order to reduce the saturation pixels in the specular reflection area and retain the stripe information in the occluded edge area.

[0043] After acquiring the first and second fringe frames, the laser is passed through the non-fringe light-transmitting area of ​​the diffraction fringe element to acquire the auxiliary reference frame.

[0044] The auxiliary reference frame is used to record the brightness, chromaticity, and boundary state of the current tooth surface region in a stripe-free state, and is written into the same acquisition group along with the first and second stripe frames. Each acquisition group records the acquisition order of the first, second, and auxiliary reference frames, the imaging timestamp, the peak pulse current, the pulse width, the displacement phase of the movable diffuser, and the change in scanning head pose. If the change in scanning head pose exceeds a preset range, the acquisition group is discarded or the tooth surface region is re-acquired to reduce the impact of positional deviations between image frames on spatial depth calculation during handheld scanning.

[0045] S5. 3D point generation: Extract the fringe center, unwrap the phase, and perform triangulation on the first and second fringe frames respectively.

[0046] During stripe center extraction, the stripe image is first normalized to reduce local brightness differences; then, stripe direction filtering is performed along the stripe extension direction to suppress noise inconsistent with the stripe direction; subsequently, sub-pixel center localization is performed for each stripe, and center lines are numbered according to the stripe arrangement order in the imaging coordinates. For stripe breakage regions, the numbering relationship of adjacent continuous stripes is used as a constraint, and the break point is not directly used as an effective center line.

[0047] During phase unwrapping, the phase order between adjacent stripes is determined based on the centerline number, and the phase-depth conversion coefficient in the calibration parameter table is read to convert the phase value into a spatial depth value. For regions where only the spatial depth value of the first stripe frame exists at the same imaging coordinate, and these regions do not belong to the gingival mask, the spatial depth value of the first stripe frame is written into the tooth surface point cloud fragment.

[0048] For regions where both the first and second fringe frame spatial depth values ​​exist at the same imaging coordinates, the saturation pixel ratio, fringe center residual, and neighborhood depth discrepancy of the corresponding pixels are compared. The saturation pixel ratio represents the proportion of saturated pixels within the corresponding region; the fringe center residual represents the deviation between the fringe center localization result and the neighborhood fitted fringe center; and the neighborhood depth discrepancy represents the degree of variation between the spatial depth value and its neighboring spatial depth values. Spatial depth values ​​with low saturation pixel ratios, small fringe center residuals, and small neighborhood depth discrepancies are retained and converted into three-dimensional points.

[0049] For pixels in the auxiliary reference frame that represent the gingival region, no tooth point cloud is written. For 3D points from the ordinary region of the tooth mask, ordinary tooth markers are written; for 3D points from the outer region of the specular reflection mask (which are selected and retained), specular reflection edge markers are written; for 3D points from the edge region of the occluding mask (which are selected and retained), occluding edge markers are written. This forms a tooth surface point cloud fragment with region source markers. This tooth surface point cloud fragment contains not only 3D coordinates but also the corresponding image frame source, mask source, fringe center residual, and neighborhood depth discretization, facilitating differential processing during subsequent registration and fusion.

[0050] S6, Tooth Arch Model Fusion: While the operator moves the intraoral scanning head along the direction of the dental arch, S3 to S5 are repeated to obtain multiple continuously acquired tooth surface point cloud fragments.

[0051] When registering adjacent tooth surface point cloud segments, the scanning head pose coordinates output by the pose sensor are first read to perform coarse registration of adjacent tooth surface point cloud segments, ensuring they are within the same initial coordinate range. Subsequently, cusp edge features, adjacent tooth contact area features, and tooth body boundary points are extracted to perform fine registration of adjacent tooth surface point cloud segments, ensuring that cusps, pits, fissures, proximal surface boundaries, or prepared body edges in adjacent segments correspond to the same dental row coordinate system.

[0052] During the fine registration process, different registration weights are set according to the region source markers. A first registration weight is set for the three-dimensional points corresponding to ordinary tooth markings, a second registration weight is set for the three-dimensional points corresponding to the specular reflection edge markings, and a third registration weight is set for the three-dimensional points corresponding to the occluded edge markings. The second and third registration weights are both less than the first registration weight.

[0053] Ordinary tooth points are used as the primary registration basis, while mirror reflection edge points and occlusion edge points are used for boundary supplementation but are not used as the primary registration basis. After registration is completed, non-tooth points, isolated 3D points, and 3D points that exceed the dental arch boundary corresponding to the gingival mask are deleted. For tooth boundary points in adjacent tooth surface point cloud segments, triangular meshes are connected according to the boundary point spacing and normal direction to obtain a continuous 3D surface mesh.

[0054] The final generated 3D surface model includes the crown shape, cusp margins, occlusal fissures, contact areas with adjacent teeth, tooth boundaries, and the tooth contour near the gingival margin. For local tooth preparation scenarios, this 3D surface model may also include the shoulder line, axial plane, and proximal surface boundaries. This 3D surface model can be output in the data format required for dental digital design, for subsequent restoration design, orthodontic planning, implant guide design, or occlusal relationship analysis.

[0055] Example 2 This embodiment provides a dental 3D scanning system with a high-efficiency laser light source, used to perform the dental 3D scanning method described in Embodiment 1. This system can be integrated into a handheld intraoral scanning device to scan a single tooth, a portion of the dentition, the maxillary dentition, or the mandibular dentition, and generate a 3D surface model of the dentition to be scanned.

[0056] refer to Figure 2 As shown, the system includes an intraoral scanning head, a light source driving module, an optical calibration module, a region recognition module, a zone acquisition module, a 3D calculation module, and a model fusion module. Data is transmitted between the modules in the order of light source control, image acquisition, region recognition, depth calculation, and model fusion.

[0057] The intraoral scanning head includes a housing, an intraoral scanning window, an imaging camera, a pose sensor, and a laser stripe light source. The housing supports the intraoral scanning window, the imaging camera, the pose sensor, and the laser stripe light source, and forms a front end structure suitable for insertion into the oral cavity.

[0058] An intraoral scanning window is located at the front of the housing, allowing the laser beam from the laser stripe light source to enter the oral cavity and for reflected light from the tooth surface to return to the imaging camera. The imaging camera is located behind the intraoral scanning window and is used to acquire uniform reference frames, low-brightness stripe frames, the first stripe frame, the second stripe frame, and auxiliary reference frames. A pose sensor is located inside the housing, used to output the pose coordinates of the intraoral scanning head during continuous scanning and to send these coordinates to the model fusion module.

[0059] The laser stripe light source includes a first laser diode group, a second laser diode group, a collimating lens group, a fiber coupler, an integrating homogenizing rod, a movable scattering plate, and diffraction stripe elements. The first and second laser diode groups are mounted on the same heat-conducting base. The first laser diode group outputs visible laser light in a first wavelength band, and the second laser diode group outputs visible laser light in a second wavelength band. The collimating lens group is positioned corresponding to both the first and second laser diode groups to collimate the two visible laser beams into a collimated beam. The fiber coupler has two input ends and one output end. The two input ends receive the first and second wavelength visible laser lights, respectively, and the output end outputs a coaxial mixed beam.

[0060] The input end of the integrating homogenizing rod is coaxially aligned with the output end of the fiber coupler, and the output end of the integrating homogenizing rod faces the diffraction fringe element, used for homogenizing the mixed beam. A movable scattering plate is positioned between the integrating homogenizing rod and the diffraction fringe element, or between the diffraction fringe element and the intra-aperture scanning window, and reciprocates within a single exposure frame along a direction perpendicular to the fringe extension direction. The diffraction fringe element includes a fringe diffraction region and a non-fringe light-transmitting region. The fringe diffraction region is used to form structured light fringes, and the non-fringe light-transmitting region is used to form uniform illumination without fringe modulation.

[0061] The light source driving module is connected to the first laser diode group, the second laser diode group, the movable scattering sheet, and the diffraction fringe element. The light source driving module includes a pulse current control unit, a pulse width control unit, a band switching control unit, a scattering sheet motion control unit, and a light transmission zone switching control unit. The pulse current control unit outputs pulse currents with different peak values ​​to the first and second laser diode groups; the pulse width control unit sets the pulse width corresponding to the first and second fringe frames; the band switching control unit controls the on / off sequence of the first and second laser diode groups; the scattering sheet motion control unit controls the displacement amplitude, frequency, and phase of the movable scattering sheet; and the light transmission zone switching control unit controls whether the light beam passes through the fringe diffraction zone or the non-fringe light transmission zone of the diffraction fringe element. The light source driving module sends the pulse current, pulse width, on / off sequence, light transmission zone status, movable scattering sheet displacement, and movable scattering sheet displacement phase to the optical calibration module, the partition acquisition module, and the three-dimensional calculation module.

[0062] The optical calibration module is connected to the imaging camera, the light source driving module, and the 3D calculation module. The optical calibration module receives fringe images from a planar calibration plate, a stepped calibration plate, and a cylindrical arc surface calibration plate, and generates a calibration parameter table. The planar calibration plate has a planar grid pattern, used to determine the mapping relationship between imaging coordinates and planar grid coordinates; the stepped calibration plate includes at least three steps of known height, with adjacent steps having a known height difference, used to determine the correspondence between phase values ​​and spatial depth values; the cylindrical arc surface calibration plate has a known radius of curvature, used to record the distortion parameters of the field of view at the edge of the intraoral scanning window.

[0063] The calibration parameter table generated by the optical calibration module includes the correspondence between projection coordinates, imaging coordinates, scanning head pose coordinates, phase values, and spatial depth values. It also includes fringe period, fringe tilt angle, camera intrinsic parameters, projection extrinsic parameters, phase-depth conversion coefficients, edge field-of-view distortion parameters, and pulse current, pulse width, and movable scatterer displacement phase data under the corresponding light source conditions. The 3D calculation module reads this calibration parameter table when subsequently calculating the spatial depth value.

[0064] The region identification module is connected to the imaging camera, the light source driving module, and the partition acquisition module. The region identification module includes a reference frame reading unit, a pixel classification unit, a stripe continuity judgment unit, and a mask generation unit. The reference frame reading unit reads the uniform reference frame and low-brightness stripe frame acquired by the imaging camera. The pixel classification unit identifies saturated pixels, dark area pixels, gingival color pixels, and tooth boundary pixels based on the grayscale and chromaticity values ​​in the uniform reference frame.

[0065] The stripe continuity determination unit is used to determine whether the stripe center is continuous, whether the stripe numbering is interrupted, and whether the stripe contrast is lower than a preset threshold in a low-brightness stripe frame. The mask generation unit generates specular reflection masks, occlusion masks, gingival masks, and tooth masks based on the pixel classification results and stripe continuity determination results, and sends the above masks to the partition acquisition module and the 3D calculation module.

[0066] The partitioned acquisition module is connected to the region recognition module, the light source driving module, and the imaging camera. The partitioned acquisition module includes a standard tooth acquisition unit, a reflection compensation acquisition unit, an occlusion edge acquisition unit, and an auxiliary reference acquisition unit. The standard tooth acquisition unit reads the tooth mask and sends first acquisition control information to the light source driving module, causing the light source driving module to control the laser stripe light source to output structured light stripes through the stripe diffraction region according to the first pulse peak current and the first pulse width, and the imaging camera acquires the first stripe frame. The reflection compensation acquisition unit reads the specular reflection mask and extends the boundary of the specular reflection mask outward to form an extended region of the specular reflection mask; the occlusion edge acquisition unit reads the occlusion mask and extends the boundary of the occlusion mask outward to form an edge region of the occlusion mask.

[0067] The reflection compensation acquisition unit and the occlusion edge acquisition unit jointly send second acquisition control information to the light source driving module, causing the light source driving module to acquire the second fringe frame according to the second pulse peak current and the second pulse width, wherein the second pulse peak current is less than the first pulse peak current. The auxiliary reference acquisition unit sends non-fringe acquisition control information to the light source driving module, causing the laser to pass through the non-fringe light-transmitting area of ​​the diffraction fringe element, and the imaging camera acquires the auxiliary reference frame. The partition acquisition module sends the first fringe frame, the second fringe frame, the auxiliary reference frame, the acquisition order, the imaging timestamp, and the corresponding mask to the three-dimensional calculation module.

[0068] The 3D calculation module is connected to the partition acquisition module, optical calibration module, and region recognition module. The 3D calculation module includes a fringe center extraction unit, a phase unwrapping unit, a triangulation unit, a point cloud discarding unit, and a point cloud marking unit. The fringe center extraction unit performs grayscale normalization, fringe direction filtering, sub-pixel center localization, and centerline numbering on the first and second fringe frames, respectively. The phase unwrapping unit determines the phase order between adjacent fringes based on the centerline number and reads the phase-depth conversion coefficients from the calibration parameter table. The triangulation unit calculates the spatial depth value based on the phase value, projection coordinates, imaging coordinates, and the calibration parameter table, and converts the spatial depth value into 3D points.

[0069] The point cloud discarding unit reads the auxiliary reference frame, specular reflection mask, occlusion mask, gingival mask, and tooth structure mask. When the same imaging coordinate simultaneously contains spatial depth values ​​from the first fringe frame and the second fringe frame, the unit determines the retained spatial depth value based on the saturation pixel ratio, fringe center residual, and neighborhood depth discrepancy. The point cloud marking unit writes region source markers onto the retained 3D points. These region source markers include ordinary tooth structure markers, specular reflection edge markers, and occlusion edge markers, thereby generating a tooth surface point cloud fragment with region source markers.

[0070] The model fusion module is connected to the 3D calculation module and the pose sensor. The model fusion module includes a coarse pose registration unit, a fine feature registration unit, a non-tooth point deletion unit, and a mesh connection unit. The coarse pose registration unit reads the scanning head pose coordinates output by the pose sensor and converts the continuously acquired tooth surface point cloud fragments to the same initial coordinate system based on the scanning head pose coordinates.

[0071] The feature-based fine registration unit extracts cusp edge features, adjacent tooth contact area features, and tooth boundary points from tooth surface point cloud fragments, and performs fine registration on adjacent tooth surface point cloud fragments. During the fine registration process, the feature-based fine registration unit reads the region source markers, sets a first registration weight for the 3D points corresponding to ordinary tooth body markers, sets a second registration weight for the 3D points corresponding to specular reflection edge markers, and sets a third registration weight for the 3D points corresponding to occluded edge markers. Both the second and third registration weights are less than the first registration weight.

[0072] The non-dental point deletion unit reads the gingival mask and dental arch boundary data, and deletes the 3D points corresponding to the gingival mask, isolated 3D points, and 3D points that extend beyond the dental arch boundary. The mesh connection unit receives the deleted tooth surface point cloud fragments, performs triangular mesh connection according to the distance and normal direction between adjacent tooth boundary points, and outputs a 3D surface model of the dentition to be scanned.

[0073] In this embodiment, a corresponding data flow is formed between the intraoral scanning head, the light source driving module, the optical calibration module, the region identification module, the partition acquisition module, the 3D calculation module, and the model fusion module. The light source driving module provides traceable light source status, the optical calibration module provides the basis for conversion between phase values ​​and spatial depth values, the region identification module distinguishes different tooth surface regions before acquisition, the partition acquisition module sets different stripe acquisition conditions according to each mask, the 3D calculation module generates tooth surface point cloud fragments according to frame source and mask source, and the model fusion module completes registration, deletion, and mesh connection according to the region source markers.

[0074] With the cooperation of the above modules, the system can generate a three-dimensional surface model that includes the tooth cusps, pits and fissures, contact areas between adjacent teeth, tooth boundaries, and tooth contours near the gingival margin. This three-dimensional surface model can be exported to dental restoration design, orthodontic modeling, implant guide design, or occlusal relationship analysis software.

[0075] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A dental three-dimensional scanning method using a high-efficiency laser light source, characterized in that, It comprises the following steps: S1, light source stripe establishment: a laser stripe light source is arranged in the intraoral scanning head, the laser stripe light source comprises a first laser diode group, a second laser diode group, a collimating lens group, a fiber coupler, an integrating light uniformizing rod, a movable scattering sheet and a diffraction stripe element, the diffraction stripe element comprises a stripe diffraction area and a non-stripe light transmission area, the first laser diode group and the second laser diode group output visible lasers of different wavebands, the two visible lasers enter the stripe diffraction area after collimation, coupling and integrating light uniformization, and form a structured light stripe projected to a tooth surface area; S2, optical calibration establishment: stripe images of a planar calibration plate with a planar grid pattern, a step calibration plate with at least three known height steps and a cylindrical arc surface calibration plate with a known curvature radius are collected, a calibration parameter table between projection coordinates, imaging coordinates, scanning head pose coordinates, phase values and spatial depth values is established, and pulse current, pulse width, on-off sequence and movable scattering sheet displacement amount are written into the calibration parameter table; S3, tooth surface area identification: a uniform reference frame and a low-brightness stripe frame of a tooth row to be scanned are collected, a mirror surface reflection mask, an occlusion mask, a gum mask and a tooth mask are generated according to saturated pixels, dark area pixels, gum chroma pixels and tooth boundary pixels in the uniform reference frame and in combination with stripe continuity in the low-brightness stripe frame; S4, subarea stripe collection: a first stripe frame is collected for a region corresponding to the tooth mask, a second stripe frame is collected for an extended region outside the mirror surface reflection mask and an edge region of the occlusion mask, and an auxiliary reference frame is collected in a state that the laser passes through the non-stripe light transmission area, the peak value current of the laser pulse of the second stripe frame is smaller than that of the first stripe frame; S5, three-dimensional point generation: stripe center extraction, phase unwrapping and triangulation are respectively performed on the first stripe frame and the second stripe frame, spatial depth values under the same imaging coordinates are selected according to the auxiliary reference frame and the masks, and tooth surface point cloud segments with region source marks are generated; S6, tooth row model fusion: the continuously collected tooth surface point cloud segments are registered according to the scanning head pose coordinates, tooth tip edge features, adjacent tooth contact area features and region source marks, non-tooth points corresponding to the gum mask are deleted, tooth boundary points in adjacent tooth surface point cloud segments are connected, and a three-dimensional surface model of the tooth row to be scanned is obtained.

2. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein, In the S1 step, the output wavelength of the first laser diode group is 450nm-470nm, and the output wavelength of the second laser diode group is 510nm-540nm; The first laser diode group and the second laser diode group are arranged on the same heat-conducting base, the collimating lens group is arranged corresponding to the two groups of laser diodes, the fiber coupler has two light input ends and one light output end, the light input end of the integrating light uniformizing rod is coaxially arranged with the light output end of the fiber coupler, and the light output end of the integrating light uniformizing rod is coaxially arranged with the light input surface of the diffraction stripe element.

3. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein, In step S1, the movable scattering sheet is placed between the integrating homogenizing rod and the diffraction fringe element; the movable scattering sheet moves back and forth in a direction perpendicular to the fringe extension direction during a single frame exposure time, with a displacement amplitude of 0.05mm to 0.8mm and a moving frequency of 80Hz to 800Hz. When acquiring the first and second stripe frames, the displacement phase of the movable scatterer is written into the frame header data of the corresponding stripe frame.

4. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein In step S2, the planar calibration plate is used to determine the mapping relationship between imaging coordinates and planar grid coordinates; the adjacent steps in the stepped calibration plate have a known height difference, which is used to determine the correspondence between phase values ​​and spatial depth values. The cylindrical arc surface calibration plate is used to record the distortion parameters of the field of view at the edge of the intraoral scanning window; The calibration parameter table includes fringe period, fringe tilt angle, camera intrinsic parameters, projection extrinsic parameters, phase-depth conversion coefficient, edge field-of-view distortion parameters, and movable scatterer displacement phase data.

5. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein, In step S3, the uniform reference frame is acquired when the laser passes through the non-striped light-passing area of ​​the diffraction stripe element, and the low-brightness stripe frame is acquired under the condition that the pulse peak current of the first stripe frame is lower than that of the first stripe frame in the first laser diode group or the second laser diode group. In a uniform reference frame, a continuous pixel region whose grayscale value reaches the camera saturation threshold is recorded as a specular reflection candidate region. A continuous pixel region whose grayscale value is lower than the dark area threshold and is located inside the tooth boundary is recorded as an occlusion candidate region. A continuous pixel region whose chromaticity value falls within the gingival chromaticity range is recorded as a gingival candidate region. The stripes continuity in the low-brightness stripe frame is then combined to obtain each mask.

6. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein, In step S4, the outer expansion area of ​​the specular reflection mask is obtained by extending the boundary of the specular reflection mask outward by 3 to 15 pixels, and the edge area of ​​the occlusion mask is obtained by extending the boundary of the occlusion mask outward by 2 to 10 pixels. The first stripe frame uses the first pulse peak current and the first pulse width, and the second stripe frame uses the second pulse peak current and the second pulse width. The second pulse peak current is 20% to 70% of the first pulse peak current, and the second pulse width is 80% to 160% of the first pulse width.

7. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein, In step S5, stripe center extraction includes grayscale normalization, stripe direction filtering, sub-pixel center localization, and center line numbering. Phase unfolding involves determining the phase order between adjacent stripes based on the centerline number, and converting the phase order into spatial depth values ​​using a calibration parameter table; When the same imaging coordinates simultaneously contain the spatial depth values ​​of the first fringe frame and the second fringe frame, the saturation pixel ratio, fringe center residual, and neighborhood depth dispersion of the corresponding pixels are compared, and the spatial depth value with the smaller value of two or three of them is retained.

8. The high-light-efficiency laser light source-based dental three-dimensional scanning method according to claim 1, wherein, In step S6, the region source marking includes ordinary tooth marking, specular reflection edge marking, and occlusion edge marking; When registering point cloud fragments on the tooth surface, a first registration weight is set for the three-dimensional points corresponding to ordinary tooth body markers, a second registration weight is set for the three-dimensional points corresponding to mirror reflection edge markers, and a third registration weight is set for the three-dimensional points corresponding to occlusion edge markers. The second and third registration weights are both less than the first registration weight. After registration, delete the 3D points corresponding to the gingival mask, isolated 3D points, and 3D points that extend beyond the dental arch boundary.

9. A dental three-dimensional scanning system with a high-efficiency laser light source, used to implement the method as described in any one of claims 1-8, characterized in that, The device includes an intraoral scanning head, comprising a housing, an intraoral scanning window, an imaging camera, a pose sensor, and a laser stripe light source. The laser stripe light source includes a first laser diode group, a second laser diode group, a collimating lens group, an optical fiber coupler, an integrating homogenizing rod, a movable scattering sheet, and diffraction stripe elements. The light source driving module is connected to two sets of laser diodes and a movable scattering sheet, respectively. The optical calibration module generates a calibration parameter table; The region recognition module outputs a specular reflection mask, an occlusion mask, a gingival mask, and a tooth structure mask. The partition acquisition module controls the acquisition of the first stripe frame, the second stripe frame, and the auxiliary reference frame according to each mask. The 3D calculation module generates tooth surface point cloud fragments with region source markers; The model fusion module generates a three-dimensional surface model based on the scanning head pose coordinates, tooth cusp edge features, adjacent tooth contact area features, and region source markers.

10. The high-efficiency laser light source for a dental three-dimensional scanning system according to claim 9, wherein, The partition acquisition module includes: a general tooth acquisition unit, a reflective compensation acquisition unit, and an auxiliary reference acquisition unit; the general tooth acquisition unit is connected to the data terminal of the tooth mask, the reflective compensation acquisition unit is connected to the data terminals of the specular reflection mask and the occlusion mask respectively, and the auxiliary reference acquisition unit is connected to the control terminal of the corresponding non-striped light transmission area in the light source driving module. The three-dimensional calculation module includes a stripe center extraction unit, a phase unrolling unit, a triangulation unit, and a point cloud discarding unit; the model fusion module includes a pose coarse registration unit, a feature fine registration unit, a non-dental point deletion unit, and a mesh connection unit. The mesh connection unit receives the deleted tooth surface point cloud fragments and outputs a three-dimensional surface model.

Images