Device and method for carrying out an intraoral scan

Abstract

The invention relates to a method and a device for creating 3-dimensional images with the aid of an intraoral scanning device. In the method according to the invention, a color pattern (111, 112, 113) is projected onto a surface to be captured – in particular the teeth (40). During projection, a plurality of images are recorded with the aid of at least 3 cameras (13) which are spaced apart from one another and arranged in a plane on the scanning device. By way of the displacement of corresponding pixels, a depth map can be directly calculated. In a subsequent step, the calculated depth maps are merged and, finally, a 3D model is calculated. The essence of the invention is that the projected color pattern generates intensity profiles that are as continuous as possible in the individual color channels and matching pixels or patterns are detected by way of the evaluation of the vectorial differences in the color space. The color pattern is projected intermittently. Novel solution approaches needed to be invented in order to realize micro-projection with high light intensity during flash operation.

Description

[0001] Title:

[0002] Device and method for performing an intraoral scan

[0003] Description:

[0004] The present invention relates to a method for acquiring depth data in the oral cavity of a patient, as well as a device for performing these scans and for generating a 3D model of the acquired area.

[0005] Such devices are also referred to as intraoral scanners.

[0006] By creating a 3D model of the scanned area in the patient's oral cavity, it is possible to visualize teeth, dentures, crowns, and bridges, and to produce dental products without the need for conventional impressions using an impression material. Previously necessary intermediate steps, such as the production of a plaster model and the use of an extraoral scanning system, can be eliminated with an intraoral scan.

[0007] The 3D models obtained with the help of an intraoral scan can then be used in the computer-aided production of dental prostheses—such as crowns or bridges—or for computer-aided dental diagnostics. This is a growing market.

[0008] A variety of different methods and devices for performing intraoral scanning are available on the market. Each method has its advantages and disadvantages.

[0009] One way to collect distance data is through triangulation.

[0010] This makes it possible to measure distance using just one camera and a projected pattern. A single-camera approach requires calibrating the correspondence between the patterns projected by the light source, e.g., fringes, and the fringes observed in the image.

[0011] In general, the patterns captured by the camera differ from the projected pattern due to the different optical properties of the objects or materials being scanned. This is especially true for teeth. For example, sharp edges may be present in the pattern, making accurate triangulation and matching to the pattern more difficult.

[0012] In order to calculate as much depth information as possible, as many lines as possible must be displayed and evaluated. The problem here is that, on the one hand, the number of evaluable lines is limited by the optical properties of the teeth, and on the other hand, the assignment to the projected line can no longer be established. To ensure assignment to the projected line, several images with stripes of different widths are usually taken and evaluated. The disadvantage is that the intraoral scanner moves between the images and this must be taken into account in the evaluation. Known alternative methods, which do not appear suitable due to the properties of the teeth, would be to encode the stripes or to use colored stripes.

[0013] Another approach is stereovision, which uses two cameras to enable spatial vision. However, it should be noted that there are also triangulation methods that use two cameras.

[0014] WO201 9085402A1 describes an intraoral scanner that requires at least one camera. Therefore, it is essentially a scanner based on the triangulation method.

[0015] The basic principle of stereo vision is that the same points on an object to be scanned are recorded by at least two cameras from different viewing directions. The disparity detected in this way can be used to generate a depth map and ultimately a 3D image of the object to be scanned. As with triangulation, there are many different versions of stereo vision.

[0016] For example, there are active and passive stereo vision. Unlike active stereo vision, passive stereo vision does not require light projection onto the object being scanned.

[0017] In active stereo vision, a pattern is projected onto the object being scanned using light projection, allowing disparity to be determined more clearly and accurately. Therefore, random patterns are often used here.

[0018] Especially with teeth that are still moist in the patient's mouth, many optical effects, combined with partial translucency, occur. Finely structured, random patterns can no longer be clearly recognized. Therefore, patterns are required that are as clear as possible on the teeth and allow the assignment of image sections between the left and right images.

[0019] One option is to use patterns, colors, and gradients in addition to black and white. This makes the colors very easy to identify and allows for accurate matching. Using color stripes allows for maximum contrast.

[0020] The arrangement of the cameras, projectors and overall geometry, particularly of the distal section of the scanning device, which is inserted into the oral cavity of a patient, is often designed differently in the known intraoral scanners and adapted to the scanning method used in each case.

[0021] However, compared to other 3D scanning methods commonly used in the field, it is important to consider that the geometric and optical properties of a patient's oral cavity pose significant technical challenges, and the distal part of the intraoral scanner must be very small. Special solutions have been developed for this purpose in the state of the art for intraoral scanners.

[0022] EP2166303A1 describes an intraoral scanner with a large number of cameras arranged alternately with light projectors on the device.

[0023] WO2016142917A1 describes a scanning device that partially covers the teeth during scanning.

[0024] The object of the present invention is to propose an improved scanner and an improved device of the type initially described, which enables fast and accurate scanning. Due to the arrangement of the cameras and projector in the distal region of the intraoral scanner, these are limited in size. New approaches would have to be found to implement the microprojector.

[0025] According to the invention, this object is achieved by a scanning method according to claim 1 and a scanning device according to claim 11. The subclaims represent preferred embodiments of the invention. The scanning device is used in particular for scanning one or more teeth.

[0026] In a first step of the method according to the invention, a color pattern or color stripe pattern is projected onto the surface to be scanned, preferably with the aid of at least one projector. While the color pattern is projected onto the surface of the teeth, a group of images or groups of images are simultaneously recorded continuously or at short intervals using at least three cameras arranged in a plane and laterally spaced from each other.

[0027] This plurality of images accordingly comprises a plurality of image groups, with the images of the respective image group being taken simultaneously. This plurality of images is then or continuously forwarded to a processing unit. If necessary, only certain information from these images can be forwarded to the processing unit. This variant is preferably included in the term "forwarding the images to the processing unit."

[0028] In the processing unit, at least two of the three images captured simultaneously are then compared. These two compared images of a group of images are referred to below as an image pair. When comparing the images, an algorithm is used that is designed to recognize corresponding patterns in the image pair(s).

[0029] Corresponding patterns refer to individual image pixels or regions of the image pairs that are depicted in the image pairs with a corresponding offset from each other. From the offset (disparity) of the patterns or image pixels in the image pair, the depth can be directly calculated in the form of a depth map. By aligning the camera, the search space for corresponding pixels can be reduced to one row.

[0030] In a final step, the depth maps created from the images of the image groups are combined to create the 3D model.

[0031] The present method differs from the state of the art in particular in that the generated color pattern, projected onto the tooth and captured by the cameras, exhibits largely continuous intensity gradients in the individual color channels. Matching pixels or patterns are then detected by color space analysis. Matching pixels are determined by comparing the directions and lengths of the color vectors in the color space.

[0032] This consistency can reduce noise in the image. "Largely consistent" means that, in the present invention, jumps can occur due to shadows in the projection pattern.

[0033] Preferably, at least two pairs of images are created from each group of images and examined for corresponding patterns. Thus, at least two depth maps can be generated through disparity detection. Due to the arrangement of the cameras, the depth maps overlap. A depth map is calculated from the at least two depth maps, with the quality of the points in the overlapping area being evaluated and improved.

[0034] The color patterns can be generated in different ways. According to a preferred variant of the invention, the color pattern is generated using three laser diodes and at least one diffractive optical element. According to one embodiment, the three laser diodes, which are located, for example, in a proximal section of a handpiece of the device according to the invention, are connected to the diffractive optical element via optical fibers. The diffractive optical element (DOE) is preferably arranged in the distal section of the handpiece or in the headpiece of the device and is positioned in the oral cavity of a patient during the scan. The diffractive optical element preferably has three regions and ensures that red, green, and blue stripes are projected onto the teeth.

[0035] According to a second variant of the invention, the color pattern or color stripe pattern is generated by colored LEDs (preferably red, green, blue) and apertures. In this embodiment, the light elements—here, LEDs—are preferably arranged in the handpiece of the device and connected via optical fibers to apertures located in the distal section of the headpiece of the device.

[0036] To avoid motion blur in the image, the projection pattern is flashed for 1 ms. This requires a high pulsed light output in the range of 300 mW for the light source, not taking into account losses due to filters or coupling into the fiber optic cable.

[0037] According to the invention, the color patterns are projected intermittently onto the surface to be detected and the images of at least one group of images are detected within this projection time.

[0038] Offset from the intermittent projection of the color patterns, it is preferably provided that the light from a white LED illuminates the surface to be scanned. This makes the surface to be scanned more visible to the human eye and / or to a camera for recording and playback on a screen, and facilitates the positioning and movement of the distal section of the handpiece of the intraoral scanner in the patient's mouth during the scanning process.

[0039] The light from the white LED, or the light reflected from the tooth surface from the white LED, can preferably also be used to determine the color of the teeth or the surface to be scanned and roughly the scanned material, such as teeth, gums, or metals. For this purpose, it can be provided that an image is captured with the aid of one of the cameras while the surface to be scanned is illuminated with the white LED and passed on to a computing unit. The computing unit preferably textures the generated 3D model based on this data. Analogously, according to another embodiment, the color and / or material information can also be stored in the 3D model. According to yet another embodiment of the invention, using color and / or material information, the stereo matching can be optimized, and / or the luminosity of the structured light can be regulated, and / or the projection pattern can be adapted to the material or color at runtime.In this context, the scanner may have or be connected to a control unit that is designed to carry out these steps.

[0040] Storing the color and material information in the 3D model makes it possible to segment the dentition during the scanning process, identify gums, palate, braces, metal implants, or obstructive data such as the cheek or tongue. Using all this information, it is possible to estimate the accuracy of the scanning process and provide the user with indications of insufficient scan quality or to highlight them in the 3D representation.

[0041] The scanning device according to the invention preferably has an acceleration sensor. If the detected acceleration or movement exceeds a predetermined threshold, the scanning process is interrupted or terminated according to one embodiment. This interrupts the scanning process if the movement is too violent and can be resumed if the movement is gentler.

[0042] According to a further embodiment of the method and device according to the invention, the movement changes detected by the acceleration sensor are compared by a processor with predefined movement change data stored in a memory unit. This allows gesture control by the user.

[0043] In contrast to triangulation methods, the method according to the invention essentially does not require constant calibration. Only factory calibration of the camera optics is provided.

[0044] A scanning device according to the invention is proposed for creating 3-dimensional images, in particular of a patient's teeth, using the method described above.

[0045] The scanning device according to the invention comprises a handpiece having a proximal section and a distal section for insertion into the oral cavity of a patient. Furthermore, the scanning device comprises a processing unit and a control unit. The handpiece is connected to a control unit and a processing unit for the exchange of information. The control unit and / or the processing unit can also be arranged in the handpiece. According to one embodiment, the handpiece is connected to the control unit and / or processing unit via a cable. According to another embodiment, however, a system for the wireless transmission of information between the handpiece and the processing unit and / or the control unit can also be provided.

[0046] On the distal section of the handpiece, there are at least two, typically three, cameras used to perform a scan. The three cameras are arranged in one plane. In addition to the three-camera version, versions with four or more cameras are also possible.

[0047] The distal section also contains a system for generating a color stripe pattern, or at least components thereof. The color stripe pattern is preferably generated using a diffractive optical element (DOE) together with at least three laser diodes. The at least three laser diodes are preferably arranged in the proximal section of the handpiece and connected to the DOE on the distal section of the handpiece via optical fibers.

[0048] The DOE preferably has three different regions, so that each of the three laser diodes is optically coupled into a region of the DOE. The three laser diodes are preferably red, green, and blue.

[0049] According to a particularly preferred embodiment, the DOE is designed to project red, green and blue color stripes with a width of 700 to 1800 micrometers at a distance of 5-15 mm onto the surface to be scanned, in particular the teeth.

[0050] According to an alternative embodiment, the color pattern or color stripe pattern is generated with at least three colored LEDs (red, green, and blue) and apertures. One embodiment also provides for the use of fiber optics to save space and prevent heat problems. For example, stripe or triangular patterns or tapered stripes can be generated by the apertures. Surprisingly, it has been shown that apertures with a tapered cross-section (from the light entrance to the light exit) optimize the light intensity within the projected stripes or within the projected triangle to such an extent that a triangular intensity curve is created, instead of the more trapezoidal intensity curve otherwise typical for apertures.

[0051] The control unit of the scanning device is designed to control the cameras arranged on the distal section in order to simultaneously capture a series of images during the scanning process and forward them to a computing unit. According to the invention, the computing unit is designed to evaluate the captured images using a stereo matching method. This makes it possible to generate depth maps of the images, which are then converted into a 3D model by the computing unit of the device or by an external computing unit.

[0052] The core of the invention is that the DOEs, together with the laser diodes, or the LEDs, together with the apertures, create a pattern with continuous intensity transitions. The pattern can be implemented using colored stripes, a triangular pattern, or a combination of white light with colored elements such as lines or stripes.

[0053] According to the invention, the computing unit is designed to use changes in the color space to detect matching patterns in the image pairs.

[0054] In addition, the surface to be scanned is illuminated with a white LED, offset from the intermittent projection of the color stripe pattern, as described above. According to one embodiment, the scanning device in the handpiece has a rotary vibration mechanism. This vibration mechanism is designed to set the DOE located on the distal section in motion. The color stripes generated by the DOE and projected onto the surface to be scanned are thereby blurred, which is equivalent to the generation of continuous intensity transitions of the color stripe pattern. Preferably, the intensity transition is sinusoidal or approximately sinusoidal. Another important effect here is the reduction of speckles of any kind caused by the laser.

[0055] In addition to the possibility of vibrating around a single axis, vibration can also be performed around two axes. This allows patterns to be generated with and without a DOE. Without a DOE—i.e., only with the laser point—straight lines or Lissajou patterns can be generated. With a DOE, multiple stripes and Lissajou patterns can be generated. The amplitude and frequency of the vibrations can be adjusted during the scanning process; this allows the pattern, especially its width, to be adapted to the material being scanned. Material detection is possible using intermittent white illumination.

[0056] A vibration mechanism can also be designed in a 90° deflection in the form of a prism or mirror.

[0057] In both embodiments, the laser diodes or LEDs are operated with a pulse length of between 0.1 and 2.5 ms. Particularly preferably, approximately 1 ms. This time is sufficient to capture at least one group of images, i.e., three images taken simultaneously. The short illumination time reduces blur caused by motion and also enables the use of rolling shutter camera sensors.

[0058] The cameras used in the scanning device are preferably so-called rolling shutter sensors due to their size

[0059] According to a further variant of the invention, the control unit is designed to vary the light intensity of the laser diodes or LEDs during the scanning process. According to a further embodiment of the invention, the control unit adjusts the light intensity of the laser diodes or LEDs depending on the intensity values ​​of the colors recorded by at least one camera.

[0060] One advantage of the scanning device according to the invention is the short scanning distance. To further reduce this distance, one embodiment of the invention provides a prism arranged in front of the cameras in the optical direction.

[0061] A further advantage of the invention is that when using at least 3 cameras - especially when the scanning process is carried out over the inner and outer edges of the teeth - the distal section of the handpiece only needs to be moved once over the outer edge and once over the inner edge of the teeth (for the upper and lower jaw respectively).

[0062] In order to better enable the distal section of the handpiece to be guided over the edges of the teeth, one embodiment of the invention provides “wings” to act as a guide to the sides of the cameras of the scanning device. The wings are designed and aligned in such a way that when the distal section is guided over the edges of the teeth, said edge finds space between the wings. The inner sides of the wings touch either the front and top side or the back of the teeth and the top side of the teeth (in particular the molars). The wings are therefore designed to be supported on said sides and to enable scanning over the edge. The term “scanning over the edge” in the context of this invention means that the distal section of the scanning device is moved once along the front edge of the teeth and once along the back edge of the teeth.For this purpose, the wings are preferably spaced between 1 and 2 mm apart and preferably extend at an angle of approximately 35 degrees relative to the surface in which the cameras are arranged. They preferably have a height of between 2 and 4 mm.

[0063] To prevent fogging of the camera optics and other optically active components on the distal section of the scanning device's headpiece, the distal section of the handpiece or headpiece preferably has a heating element and a temperature sensor. The heating element, together with the temperature sensor and the control unit, ensure that the optically functional components remain above 32°C. This prevents fogging of the outer surfaces of these components. Maintaining a constant temperature for the camera and the optical elements also improves measurement accuracy.

[0064] After the scanning process in a patient's mouth, at least the distal section of the scanning device's handpiece must be autoclaved and is designed accordingly. However, according to a particularly efficient embodiment, the scanning device can also have a removable protective cover that can be disposed of or individually autoclaved after each use of the device.

[0065] Alternatively, the distal section of the handpiece can be designed to be attachable to the proximal part of the handpiece and can be autoclavable as a whole.

[0066] According to a further embodiment, in addition to the scanning function for the near range, the device can also have a mode for recording a long-range range, so that an intraoral overview scan of the dental arch or an image of the face or a facial area outside the oral cavity can also be taken. The invention is described in more detail below with reference to drawings. The drawings represent only preferred embodiments and are not to be construed as limiting.

[0067] They show:

[0068] Fig. 1 is a schematic representation of the intraoral scanner in a version with a diffractive optical element and laser diodes;

[0069] Figure 2 shows an illustration of the intraoral scanner with LEDs and apertures;

[0070] Figure 3a shows the head with light guide, DOE and 3 cameras;

[0071] Figure 3b shows the head with light guide, deflection, aperture body and 3 cameras;

[0072] Figure 3c shows a cross section through a diaphragm body according to one embodiment;

[0073] Figure 4 shows the application of the intraoral scanner on teeth;

[0074] Figures 5a), b) show two preferred projection patterns projected onto a flat white surface;

[0075] Figures 5c), d) the intensity curves of an image line recorded by the cameras when using a projection pattern according to Figure 5b);

[0076] Figure 6: Group of images with a pattern projected onto a tooth; Figure 7: Color gradient of row 50 in the left and middle images with the intensities of red, green, and blue;

[0077] Figure 8 shows the detected disparity of the left and middle cameras, as well as the middle and right cameras, and a composite disparity map in the middle.

[0078] List of reference symbols:

[0079] 10 headpiece

[0080] 12 Deflection with prism or mirror

[0081] 13 Camera

[0082] 14 aperture(s)

[0083] 15 Projection

[0084] 16 Diffractive optical element

[0085] 17 Focusing / Lens

[0086] 18 wings

[0087] 19 aperture body

[0088] 20 neck piece

[0089] 21 light guides

[0090] 22 Inlet aperture

[0091] 23 Exit aperture

[0092] 30 handpiece

[0093] 31 current controllers

[0094] 32a LEDs

[0095] 32b laser diodes

[0096] 33 Coupling into optical fiber

[0097] 40 tooth surface / model

[0098] 50 camera lines

[0099] 51 left camera image

[0100] 52 middle camera image

[0101] 53 right camera image 111 color filter stripe red

[0102] 112 color filter strips green

[0103] 113 color filter strips blue

[0104] 121 Color intensity red

[0105] 122 Color intensity green

[0106] 123 Color intensity blue

[0107] 211 Maxima red

[0108] 212 Maxima green

[0109] 213 Maxima blue

[0110] 214 same intensity red / green

[0111] 215 same intensity red / blue

[0112] 216 same intensity green / blue

[0113] Figure 1 shows an embodiment of the intraoral scanner according to the invention. The intraoral scanner is divided into a handpiece 30, a neck piece 20, and a head piece 10. According to this embodiment, a current controller 31 is housed in the handpiece 30, which supplies power to the laser diodes 32b, also housed in the handpiece 30, so that they emit pulsed light with a pulse length of approximately 10 s.

[0114] 1 ms. The laser diodes 32b are connected via a coupling 33 to the optical fibers 21, which run through the neck piece 20 of the scanner. The optical fibers 21 are optically coupled to a diffractive optical element 16 via a focus or lens 17 housed in the head piece 10, a 90° deflection by means of a prism 12. The lens is preferably a GRIN lens. The diffractive optical element has three regions (RGB) and ultimately generates a color pattern for projection onto a patient's teeth. The cameras of the intraoral scanner are not visible in this illustration. Figure 2 shows an embodiment of the intraoral scanner which, instead of laser diodes, has at least three colored LEDs (red, green, blue). As in Figure 1, these are supplied with a pulsed current of approximately 1 ms to generate corresponding light flashes.In this embodiment, the LEDs 32a are preferably arranged in the handpiece 30 and are optically connected to the optical fiber 21 arranged in the neck piece 20. One optical fiber is provided for each aperture. A 90° deflection 12 (prism) and apertures 14 are arranged in the headpiece 10 of the intraoral scanner. The apertures are arranged in an aperture body 19, as shown in Figure 3c. Ultimately, up to 100 mW of power per color is emitted at the projection 15 during the light pulse. The cameras of the intraoral scanner are not visible in this illustration.

[0115] Figure 3a is a representation of the headpiece 10 of the intraoral scanner with three cameras 13. Cylindrical lenses 17 for focusing are attached to the optical fibers 21 (one optical fiber is provided for each color). The laser beams are thereby sent to the diffractive optical element (DOE) 16 via a prism 12. The color pattern is created in the DOE 16 and projected onto the surface to be scanned—here, the teeth 40 of a patient. The light reflected by the teeth is then captured by the three cameras 13. The received signals are then passed on in whole—or at least in part—to a processing unit for determining the disparities of the individual pixels.

[0116] Figure 3b shows a representation of the components in the headpiece 10 according to another embodiment of the intraoral scanner with LEDs. The version with LEDs comprises two aperture bodies 19 in which a plurality of apertures 14 (through-openings) are arranged. The apertures 14 or through-openings preferably have a rectangular, trapezoidal, or triangular cross-sectional area. It has been shown that the color patterns, in particular a color stripe pattern or a triangular pattern, can be formed particularly advantageously with two aperture bodies, wherein the respective corresponding apertures 14 of the two aperture bodies 19 are aligned with one another in the case of the color stripe pattern, i.e., continue the line, or are arranged offset from one another in the case of a triangular pattern. Each aperture 14 is connected (not shown here for clarity) via an optical fiber 21 to preferably a red, blue, or green LED.

[0117] The use of at least two aperture bodies 19, which comprise a plurality of apertures 14 that together generate the projection, is particularly advantageous in intraoral scanners because this allows a small geometric size of the head piece 10.

[0118] The light emitted by the apertures and reflected by the teeth during treatment is captured by the three cameras 13.

[0119] Figure 3c shows a cross-section through the aperture body 19 in the plane of the apertures 14 / through-openings. Surprisingly, it has been shown that a particularly advantageous, continuous intensity distribution of the projection can be achieved when the apertures have a cross-sectional reduction from the inlet opening 22 compared to the outlet opening 23, i.e., the cross-section of the through-opening tapers. This allows a preferred triangular intensity distribution to be achieved. This triangular intensity distribution has much better optical properties for performing a scan than previously known apertures or apertures used in this field, so that such an aperture can be considered an independently inventive aspect, particularly in connection with intraoral scanners or other, preferably medical, applications where the smallest possible dimensions are to be achieved.

[0120] Figure 4 shows the intraoral scanner in use on a tooth model 40. This illustration clearly shows that the scanner is used to scan along the edge of the teeth. In order to guide the scanner accordingly, the headpiece has two wings 18 which laterally limit the plane of the cameras. During scanning, one edge of the teeth - i.e. the inner or outer edge of the teeth - is positioned between the wings 18. The wings are preferably spaced between 1 and 2 mm from each other and are at an angle of 35 degrees to the plane of the cameras. Figure 5a shows a projection 15 of the stripe pattern on a white background. According to the embodiment shown here, the projection 15 has red, green and blue colored stripes 111, 112, 113. The projection distance is preferably between 5 and 15 mm. The strips are preferably formed by corresponding and mutually aligned apertures 14 in two aperture bodies 19.

[0121] Figure 5b shows a projection 15 of a triangular pattern on a white background. The projection distance is preferably between 5 and 15 mm. This variant is created using LEDs and corresponding apertures 14 with a triangular cross-section, which are arranged laterally offset from each other on two aperture bodies 19, preferably by half the width of the base of the triangle.

[0122] Overall, this offset triangular projection allows for more color transitions with fewer light sources than a stripe projection, in which the stripes are generated by two corresponding and aligned apertures 14 on two aperture bodies 19. Compared to the stripes, color changes also occur in the transverse direction—that is, from the base of the triangle toward the tip.

[0123] Figure 5c shows the intensity curve of the middle line of the middle camera image 52 (at optimal aperture) when projecting the triangular color pattern of Figure 5b onto a white background. The aperture, or rather the light source, used to create a triangle is located in the wide part of the triangle relative to the camera center (see aperture arrangement in Figure 3b). Therefore, the intensity decreases noticeably toward the apex as the distance increases.

[0124] Figure 5d shows the intensity curve of a lower line of the middle camera image 52 (at optimal aperture) when projecting the triangular color pattern from Figure 5b onto a white background. Compared to Figure 5c, the intensity peak is also smaller because the distance to the aperture increases. The differences between the intensity peaks can be reduced by optimizing the aperture. Ideally, the peaks should have the same intensity. Figure 6 shows three camera images 51, 52, and 53 on which the projection according to Figure 5a was recorded on the surface of a tooth model 40. The individual color stripes red 111, green 112, and blue 113 are additionally smoothed by the optical properties of the tooth and thus blend more closely into one another. Due to the uneven surface of the teeth, the individual stripes in the projection are no longer purely parallel to one another.The straight line 50 represents a camera line, which is analyzed in Figure 7 with regard to the color gradient.

[0125] Figure 7 shows the intensity distribution of the individual colors along the camera line 50 from Figure 6 for two cameras. It can be seen that the color stripes 111, 112, and 113 in a camera line 50 produce minima and maxima. Comparing the intensity distribution of the individual colors, one sees similar distributions and corresponding maxima (211, 212, 213) or intersection points (214, 215, 216). Additional values ​​in between can also be assigned. The displacement of the individual image pixels (disparity) within the line can thus be determined point by point. The stereo matching algorithm is more complex, but is based on this assignment. It analyzes not only point by point, but entire areas.

[0126] Figure 8 shows three disparity maps - one (left map) from the left camera to the middle camera and one (right map) from the middle to the right camera and one (middle map) a combined disparity map. The disparity was calculated for each of the middle cameras. The depth can be determined from the disparity. Dark dots mean that the disparity cannot be determined or cannot be determined with sufficient accuracy. Light values ​​indicate high disparity or proximity to the camera. The left disparity map is limited on the right side, the right on the left side, because the disparity can only be determined for areas that the corresponding cameras "see".

[0127] The two disparity maps have an overlapping area where two disparity values ​​can be obtained per image pixel. This allows the middle map to be calculated, which has disparity or depth values ​​with greater accuracy in the overlapping area.

Claims

Claims: 1 . A method for creating 3-dimensional images using a scanning device comprising the steps: - projecting a color pattern (111, 112, 113) onto a surface to be detected (40), - recording a plurality of images with the aid of at least three cameras (13) spaced apart from one another and arranged in a plane, wherein the plurality of images comprises a plurality of image groups, each of which contains images recorded simultaneously with the aid of the at least three cameras (13), - Forwarding the majority of images to a computing unit, - comparing at least 2 of the images (image pair) taken simultaneously by at least 3 cameras (13), - detecting corresponding patterns in the pair of images, each corresponding to the same part of the surface (40) to be detected, - Calculating the displacement of individual image pixels of the matching patterns or the image pixels of the image pair, - Creating at least one depth map with depth information from at least one pair of images, - combining the depth maps generated from the images of the image groups, wherein the color pattern (111, 112, 113) projected onto a surface (40) and captured by the cameras (13) generates largely continuous intensity profiles in the individual color channels, and wherein matching pixels or patterns in the image pairs are recognized by comparing the directions and lengths of the color vectors in the color space, characterized in that the color patterns (111, 112, 113) are projected intermittently onto the surface (40) to be captured and the images of at least one image group are captured within the projection duration. Method according to claim 1, characterized in that a white LED illuminates the surface to be scanned (40) intermittently and offset from the projection of the color pattern (111, 112, 113). Method according to claim 2, characterized in that the reflection of the light emitted by the white LED is captured and used to determine the color and / or material of the surface to be scanned (40). Method according to claim 2, characterized in that the reflection of the light emitted by the white LED is captured and additionally used to create a depth map. Method according to one of the preceding claims, characterized in that at least two image pairs are formed from each image group and are each examined for corresponding patterns.Method according to one of the preceding claims, characterized in that the depth information of at least two image pairs of an image group is compared with each other to achieve higher information quality and reliability. Method according to one of the preceding claims, characterized in that the color pattern is generated with the aid of three laser diodes (32b) and at least one diffractive optical element (16) or by colored LEDs (32a) and (tapering) apertures. Method according to one of the preceding claims, characterized in that the scanning process is interrupted if the detected movement of the cameras (13) is above a specified limit and is automatically continued if the movement of the cameras (13) detected with the aid of an acceleration sensor is below a predetermined threshold. Method according to one of the preceding claims, characterized in that the scanning process is controlled by comparing recorded acceleration patterns with acceleration patterns stored in a memory (gestures). Method according to one of the preceding claims, characterized in that, in contrast to triangulation methods, no calibration is required before or during the scan.Scanning device for carrying out an intraoral scan for dental purposes, comprising: a control unit, a computing unit located in the handpiece (30) of the intraoral scanner, a handpiece (30) connected to the computing unit for exchanging information, with a distal section (10) provided for insertion into the oral cavity of a patient, wherein at least three cameras (13) arranged in a plane and at least one diffractive optical element (16) or at least one diaphragm (14) are arranged on the distal section, wherein the diffractive optical element(s) (16) together with at least one laser diode (32b) or the diaphragms (14) together with at least three LEDs (32a) are designed to... Process to project a color pattern (111, 112, 113) onto the surface (40) to be scanned, and wherein the computing unit is designed to generate depth maps with the aid of the images recorded during the scanning process using the stereomatching method and to pass them on to a method for composing the depth maps, characterized in that the color pattern (111, 112, 113) projected onto a surface (40) and captured by the cameras (13) generates largely continuous intensity profiles in the individual color channels, and the computing unit is designed to use changes in the color space to recognize matching patterns in the image pairs. Scanning device according to claim 11, wherein, for generating the color patterns (111, 112, 113), a diffractive optical element (16) having three regions is optically connected via optical fibers to three laser diodes (32b) (red, green, blue) arranged in a proximal section of the handpiece (30).Scanning device according to claim 11, wherein a vibration mechanism is provided to generate continuous intensity transitions of the color pattern (111, 112, 113), which causes the diffractive optical element or a 90° deflection (MEMS-based micromirror) to oscillate, thereby "blurring" the color patterns and reducing speckle. Scanning device according to claim 11, characterized in that the laser beams are coupled directly and without optical fibers (21) into the diffractive optical element(s) (16). Scanning device according to claim 11, wherein the LEDs are each connected via optical fibers to at least one aperture (14) per color. Scanning device according to claim 11, characterized in that on the head piece (10) of the scanning device, laterally to the cameras (13) of the scanning device, “wings” are arranged as a guide aid (18), which in particular enables guidance of the head piece (10) along the inner edge and along the outer edge of the molars.

Images