EP-4071709-B1 - IMAGE FILTERING METHOD

EP4071709B1EP 4071709 B1EP4071709 B1EP 4071709B1EP-4071709-B1

Inventors

LEE, DONG HOON
LEE, SEUNG JIN

Dates

Publication Date: 20260506
Application Date: 20210108

Claims (9)

An image filtering method comprising: an image acquiring operation of acquiring (S1) a two-dimensional image of a scan target having a valid data part including a tooth inside an oral cavity and a noise data part including an object to be filtered through a scanner; a color acquiring operation of acquiring (S2) color information from all data of the two-dimensional image acquired from the image acquiring operation; a filtering operation of determining at least some data as data to be deleted having a color to be deleted and deleting (S5) the at least some data when the color information of the at least some data is included within a reference color range that is a color range of an object to be filtered distinguished from an inside of the oral cavity in the color acquiring operation; and a three-dimensional calculating operation of converting (S6) two-dimensional data having only the valid data part by deleting the data to be deleted through the filtering operation into three-dimensional volume data, characterized in that the at least some data corresponds to the noise data within the two-dimensional image and is deleted before the two-dimensional image being converted into three-dimensional volume data.
The image filtering method of claim 1, further comprising: a reference color determining operation of determining (S4) whether the color information of the at least some data acquired in the color acquiring operation is included within the reference color range, wherein the filtering operation determines the at least some data determined as the reference color range from the reference color determining operation as the data to be deleted to delete the at least some data within the two-dimensional image data.
The image filtering method of claim 2, further comprising: a reference color setting operation of setting (S3) the color to be deleted, wherein the color to be deleted in the reference color setting operation is designated through a user interface.
The image filtering method of claim 3, wherein the size of the reference color range is adjustable through the user interface with respect to the color to be deleted.
The image filtering method of any one of claims 1 to 4, wherein the color information is information expressed by using an RGB additive color mixture method.
An image filtering method comprising: an image acquiring operation of acquiring (S1) a two-dimensional image of a scan target having a valid data part including a tooth inside an oral cavity and a noise data part including an object to be filtered through a scanner; a color acquiring operation of acquiring (S2) color information from all data of the two-dimensional image acquired from the image acquiring operation; a reference color range defining operation of defining (S32) a color range of an object to be filtered as a reference color range based on an image of the object to be filtered distinguished from the inside of the oral cavity; a reference color determining operation of determining (S4) whether color information of the at least some data acquired in the color acquiring operation is included within the reference color range defined through learning; a filtering operation of determining the at least some data as data to be deleted having a color to be deleted and deleting (S5) the at least some data within the two-dimensional image data when the color information of the at least some data is included within the reference color range in the color determining operation; and a three-dimensional calculating operation of converting (S6) two-dimensional data having only the valid data part by deleting the data to be deleted through the filtering operation into three-dimensional volume data, characterized in that the at least some data corresponds to the noise data within the two-dimensional image and is deleted before the two-dimensional image being converted into three-dimensional volume data.
The image filtering method of claim 6, wherein the reference color range defining operation includes: a reference color acquiring operation of repeatedly acquiring (S32a) at least one image of the object to be filtered including the color to be deleted; and a reference color range learning operation of determining (S32b) the reference color range from the image of the object to be filtered acquired from the reference image acquiring operation.
The image filtering method of claim 7, wherein the reference color range learning operation learns an overlapping color from at least one image acquired through the reference image acquiring operation as the reference color range.
The image filtering method of any one of claims 6 to 8, wherein the color information is information expressed by using an RGB additive color mixture method.

Description

[Technical Field] The present disclosure relates to an image filtering method, and more specifically, to a method of excluding a part unnecessary for a user from an acquired image. [Background Art] A three-dimensional scanning technology is currently being widely used regardless of industrial fields, and its practicality continues to attract attention, particularly in a dental treatment field such as the production of dental prosthetic treatment products. Meanwhile, when the patient's affected area, that is, the inside of the oral cavity (meaning teeth, gums, or the like) is captured by a three-dimensional scanner, the three-dimensional scanner converts the image of the captured area into three-dimensional volume data to eventually acquire one entire oral cavity model data. At this time, when the inside of the patient's oral cavity is captured, there is a possibility that foreign substances present in the oral cavity, the hand of a user of the three-dimensional scanner (usually may be a dentist), or the like may be captured together. As described above, the foreign substance or the user's hand is a factor that hinders the acquisition of the patient's entire oral model data, and as a result, needs to be deleted in a three-dimensional volume data conversion process or a post-correction process. EP 1 607 041 A2 discloses a method for providing data useful in procedures associated with the oral cavity, in which at least one numerical entity representative of the three-dimensional surface geometry and colour of at least part of the intra-oral cavity is provided and then manipulated to provide desired data therefrom. WO 2018/219800 A1 discloses a method and apparatus for generating a 3D representation of an intraoral scene by acquiring 3D point cloud data and a corresponding color image, labeling undesired regions in the image, transforming the image to point cloud coordinate space, and filtering out 3D data points corresponding to the labeled regions before generating the final 3D representation. WO 2019/158442 A1 discloses a method for generating a digital 3D representation of an intraoral cavity by recording multiple views using an intraoral scanner, determining weights for each surface point based on scores reflecting the belief of the point representing a particular surface type, and executing a weighted stitching algorithm to generate the 3D representation from the weighted surface points. Until now, the user of the three-dimensional scanner manually deletes the part that does not correspond to the inside of the actual oral cavity from the converted three-dimensional volume data after performing the scan, so that there is a problem in that the post-correction operation takes a lot of time. [Summary of Invention] [Technical Problem] To solve the above problem, the invention is set out in the appended set of claims. An object of an image filtering method according to the present disclosure is to provide a method of not including pixel data determined as not corresponding to oral cavity information in a three-dimensional volume data conversion process by acquiring color information on each pixel of two-dimensional image data from the acquired two-dimensional image data, comparing the acquired color information with reference color information, and deleting the pixel data determined as not corresponding to the oral cavity information. In addition, another object of the present disclosure is to provide a method of not including pixel data determined as not corresponding to oral cavity information in a three-dimensional volume data conversion process by defining a reference color a reference color range through learning of the color information to delete the pixel data determined as not corresponding to the oral cavity information from the learned reference color or reference color range. The objects of the present disclosure are not limited to the above-described technical objects, and other objects not mentioned will be clearly understood to those skilled in the art from the following descriptions. [Solution to Problem] An image filtering method according to the present disclosure may include an image acquiring operation of acquiring a two-dimensional image of a scan target having a valid data part including a tooth inside an oral cavity through a scanner, a color acquiring operation of acquiring color information from at least some data of the two-dimensional image acquired from the image acquiring operation, a filtering operation of determining the at least some data as data to be deleted having a color to be deleted and deleting the data to be deleted within the two-dimensional image data when the color information of the at least some data is included within a reference color range that is a color range of an object to be filtered distinguished from an inside of the oral cavity in the color acquiring operation, and a three-dimensional calculating operation of converting two-dimensional data having only the