EP-4742149-A1 - A COMPUTER-IMPLEMENTED METHOD AND AN INTRAORAL SCANNING SYSTEM FOR REMOVING GLARE DEFECTS IN IMAGE SCAN DATA

EP4742149A1EP 4742149 A1EP4742149 A1EP 4742149A1EP-4742149-A1

Abstract

The present disclosure relates to a computer-implemented method and an intraoral scanning system. The computer-implemented method for removing a glare defect in a processed image scan data, comprising: receiving a plurality of image scan data of a dental object produced by an intraoral scanner, processing one or more image scan data of the plurality of image scan data into a processed image scan data, wherein the processed image scan data includes a plurality of pixels, determining a plurality of defect pixels in the processed image scan data by a glare detection algorithm, wherein the glare detection algorithm may be configured to determine the plurality of defect pixels, when a pixel value of each of the plurality of defect pixels fulfils one or more pixel value criteria, and, when a geometry of an arrangement of the plurality of defect pixels fulfils one or more geometry criteria, determining, by a glare-removal algorithm, non-glare image scan data by analysing pixels of the plurality of pixels that are in vicinity to the plurality of defect pixels, and replacing the processed image scan data in the plurality of defect pixels with the non-glare image scan data.

Inventors

MIHOV, Filip
HUSEINI, Admir
BAZAOU, Slimane

Assignees

3Shape A/S

Dates

Publication Date: 20260513
Application Date: 20251105

Claims (15)

A computer-implemented method for removing a glare defect from diagnostic image scan data of a dental object obtained by an intraoral scanner, the method comprising: • receiving a plurality of image scan data of a dental object produced by an intraoral scanner, wherein the plurality of image scan data includes image scan data with three-dimensional image scan data and diagnostic image scan data, • processing the diagnostic image scan data into a processed image scan data, wherein the processed image scan data includes a plurality of pixels, • determining a plurality of defect pixels in the processed image scan data by a glare detection algorithm, wherein the glare detection algorithm is configured to determine the plurality of defect pixels, when a pixel value of each of the plurality of defect pixels fulfils one or more pixel value criteria, and, when a geometry of an arrangement of the plurality of defect pixels fulfils one or more geometry criteria, • determining, by a glare-removal algorithm, non-glare image scan data by analysing pixels of the plurality of pixels that are in vicinity to the plurality of defect pixels, • replacing the processed image scan data in the plurality of defect pixels with the non-glare image scan data, • reconstructing and displaying a three-dimensional model of the dental object based on the image scan data with three-dimensional image scan data, and • displaying the diagnostic image scan data with the non-glare image scan data.
The computer-implemented method according to claim 1, wherein the one or more pixel value criteria includes a change in pixel value between the pixel values of the plurality of defect pixels and the pixels of the plurality of pixels that are in vicinity to the plurality of defect pixels, wherein the change in pixel value is above a contrast threshold.
The computer-implemented method according to any of the previous claims, wherein the one or more pixel value criteria includes a pixel value threshold that the pixel values of the plurality of defect pixels are above.
The computer-implemented method according to any of the previous claims, wherein the plurality of image scan data includes a first image scan data that includes a visible wavelength and at least a second image scan data that includes a non-visible wavelength, wherein the processed image scan data includes a composition of the first image scan data and the at least second image scan data.
The computer-implemented method according to claim 4, comprising determining a composition of the first image scan data and the processed image scan data, wherein the processed image scan data includes the at least second image scan data, and the processed image scan data in the plurality of defect pixels are replaced by the non-glare image scan data.
The computer-implemented method according to any of the previous claims, wherein the arrangement of the plurality of defect pixels includes contiguous pixels.
The computer-implemented method according to any previous claims, wherein the one or more geometry criteria includes: • a size criterion assessing whether the geometry has an area that is within an area range, • a width criterion assessing whether the geometry has a width along a dimension of the plurality of defect pixels that is within a width range, and • a shape criterion assessing whether the geometry has a shape that matches a plurality of glare shape templates.
The computer-implemented method according to any of the previous claims, wherein the glare detection algorithm includes a top-hat transformation, wherein the top-hat transformation includes one or more of following transformations: - white top-hat transformation (WTH) which includes an opening operation; and - black top-hat transformation (BTH) which includes a closing operation.
The computer-implemented method according to claim 8, wherein the white top-hat transformation includes following equation: WTH f = f − f ∘ b , where f is the processed image scan data, b is a structuring glare defect element which includes a size or width of glare defects to be detected by the white top-hat transformation, and ∘ is an opening operation.
The computer-implemented method according to claim 8, wherein the black top-hat transformation includes following equation: BTH f = f • b − f , where f is the processed image scan data, b is a structuring glare defect element which includes a size or width of glare defects to be detected by the black top-hat transformation, and • is a closing operation.
The computer-implemented method according to any of the previous claims, wherein the glare detection algorithm includes following steps: • processing the diagnostic image scan data into a second processed image scan data, • determining, based on a pixel value of the plurality of pixels of the processed image scan data, a first group of defect pixels, • determining, based on a pixel value of a plurality of pixels of the second processed image scan data, a second group of defect pixels, • determining, by comparing the first and second processed image scan data, a defect pixel of the plurality of defect pixels of the first processed image scan data when a location of a defect pixel of the first and the second group of defect pixels relative to the dental object matches.
The computer-implemented method according to claim 11, wherein the first processed image scan data includes diagnostic image scan data that is captured after the diagnostic image scan data of the second processed image scan data.
An intraoral scanning system that is configured to remove a glare defect from diagnostic image scan data of a dental object obtained by an intraoral scanner, wherein the system comprises: • an intraoral scanner configured to produce a plurality of image scan data of a dental object, wherein the plurality of image scan data includes image scan data with three-dimensional image scan data and diagnostic image scan data; • one or more processors configured to: ∘ process the diagnostic image scan data into a processed image scan data, wherein the processed image scan data includes a plurality of pixels, ∘ determine a plurality of defect pixels in the processed image scan data by a glare detection algorithm, wherein the glare detection algorithm is configured to determine the plurality of defect pixels when a geometry of an arrangement of the plurality of pixels which has a pixel value that is above a pixel value threshold fulfils one or more geometry criteria, ∘ determine, by a glare-removal algorithm, non-glare image scan data by analysing pixels that are in vicinity to the plurality of defect pixels, and ∘ replace the plurality of defect pixels in the processed image scan data with the non-glare image scan data, ∘ reconstruct and display a three-dimensional model of the dental object based on the image scan data with three-dimensional image scan data, and ∘ display the diagnostic image scan data with the non-glare image scan data.
The intraoral scanning system according to claim 13, wherein the intraoral scanner includes a first light source configured to emit visible wavelengths and a second light source configured to emit a non-visible wavelength, and wherein the plurality of image scan data includes a first image scan data that includes the visible wavelength and at least a second image scan data that includes the non-visible wavelength, and wherein the processed image scan data includes a composition of the first image scan data and the at least second image scan data.
The intraoral scanning system according to claim 14, wherein the intraoral scanner includes a third light source configured to emit another non-visible wavelength, and wherein the plurality of image scan data includes a third image scan data that includes the another non-visible wavelength, and wherein the processed image scan data includes a composition of the first image scan data, the at least second image scan data and the third image scan data.

Description

FIELD The disclosure relates to a computer-implemented method and an intraoral scanning system that is configured to remove glared defects in image scan data. More specifically, the glare defect is removed based on image processing. BACKGROUND Scanning a patient's mouth may involves many artifacts caused, for example, by stripe loss due to the scene geometry, shadows in the scene, tooth defects and/or other factors affecting the stripe shape, continuousness and/or alignment; stripe reconstruction failure and/or false detection of stripes may be caused for example by the different materials in the scene (e.g. teeth and gingiva), specular reflections from fluids in the scene (e.g. blood, saliva) and natural and/or cast shadows in the scene. Especially, specular reflections are not a trivial artifact to overcome. One way to overcome specular reflections is by cross polarize the illumination and the light source, but to have an effective cross-polarization, in the sense of removing glares, across a broad-spectrum of wavelength is difficult. For example, in an intraoral scanning system that emits wavelengths from ultraviolet (UV) wavelengths to infrared (NIR/IR) wavelengths would impose a risk of glare defects in images with UV and/or NIR/IR wavelengths. It is known that NIR/IR light can be used for assessing internal structure of a tooth, internal features (such as cracks, caries) and tooth surface in the form of transillumination of teeth or light reflection and backscattering from teeth. Unfortunately, glare defects would for some dentists be difficult to distinguish from for example caries which would results in false positives, i.e. false caries. Therefore, there is a need for a solution to the problem of removing glare defects when emitting light that includes wavelengths between UV and NIR/IR. SUMMARY It is an aspect of the present disclosure to overcome the above-mentioned problem on how to remove glare defects in image scan data when the emitted light includes wavelengths between UV and NIR/IR. According to the aspect, a computer-implemented method for removing a glare defect in a processed image scan data is disclosed. The method includes receiving a plurality of image scan data of a dental object produced by an intraoral scanner. The dental object may be a tooth, a series of teeth and/or gingival. The dental object may be a single jaw, or a combination of upper and lower jaw being separated or in a clinched state. The plurality of image scan data may include image scan data with three-dimensional image scan data and/or two-dimensional image scan data, such as diagnostic image scan data. The three-dimensional image scan data may be stitched together to produce a three-dimensional model of the dental object, wherein the two-dimensional image scan data may be used for diagnosing the dental object either manually by a user (via a display unit configured to display the two-dimensional image scan data and the three-dimensional model) or automatically by artificial intelligence. The intraoral scanner may employ a scanning principle such as triangulation-based scanning, confocal scanning, or focus scanning. The method may further include processing one or more image scan data of the plurality of image scan data or the two-dimensional image scan data into a processed image scan data, wherein the processed image scan data includes a plurality of pixels. The one or more image scan data may be similar to the two-dimensional image scan data and diagnostic image scan data. A processed image scan data may include a captured image scan data which have been processed for improving the quality of the captured image scan data. For example, the improvements may involve enhanced detail and contrast and reduced noise and distortion. The processed image scan data may include a composition of multiple image scan data which includes wavelengths within UV wavelength range, visible (VIS) wavelength range, and infrared (NIR/IR) wavelength range. The composition of multiple image scan may include a weighted combination of image scan data with different wavelengths. The advantages of the composition of multiple image scan data are that the contrast and detail of internal features and structures of a tooth improves significantly more in comparison to processed image scan data that solely includes wavelengths within NIR/IR range. The method may further include determining, by a glare detection algorithm, a plurality of defect pixels in the processed image scan data by analysing a pixel value of each of the plurality of pixels. The processed image scan data includes a plurality of pixels that may correspond to the pixels of an image sensor of the intraoral scanner, and some of these pixels of the processed image scan data may include a glare defect that is seen as specular reflections with high intensities. The plurality of defect pixels includes these pixels that comprises the glare defect. The glare detection algorithm analysis eac