KR102217319B1 - Device for extracting of shape features of porcelain fragments based on rotational axis information and virtual matching system of porcelain fragments using the same - Google Patents

Abstract

Provided are a device for extracting shape features of porcelain fragments based on rotational axis information and a virtual matching system for porcelain fragments using the same. A virtual matching system for porcelain fragments comprises: a porcelain fragment shape data generating device which scans a plurality of porcelain fragments to generate three-dimensional shape data for each porcelain fragment; a point cloud generation unit which generates point cloud data showing the appearance of the porcelain fragment fragments from the three-dimensional shape data of each of the ceramic fragments; a rotational axis estimating unit for estimating a rotational axis of each of the porcelain fragments; an apparatus for extracting shape features of porcelain fragments which aligns the point cloud data of each porcelain fragment based on the estimated rotational axis to define the coordinates of the point cloud data, and defines shape features based on the rotational axis for each porcelain fragment; and a porcelain fragment virtual matching device which performs virtual matching of the porcelain fragments by using the rotational axis-based shape feature.

Description

Device for extracting of shape features of porcelain fragments based on rotational axis information and virtual matching system of porcelain fragments using the same}

It relates to a device for extracting shape features of ceramic fragments and a virtual matching system for ceramic fragments using the same. Specifically, the shape features of ceramic fragments are extracted based on rotation axis information, and the ceramic fragments are matched using the shape features of the extracted ceramic fragments. It relates to a virtual matching system.

[Explanation of nationally supported R&D]

This research is under the management of the Korea Advanced Institute of Science and Technology and is supported by the Ministry of Culture, Sports and Tourism's Cultural Technology Research and Development Project's "Simulation Solution Technology Development for the Original Restoration of Damaged Porcelain" (Task Serial Number: 1375026917, Detailed Task Number: R2018020101-0002). It was made by.

Inorganic ceramics are less damaged than organic ones, so the amount of discharge and excavation is vast, so the time, manpower, equipment, and costs required for restoration and preservation are relatively high. Since excavated pottery is damaged and excavated by various damage factors, the original restoration has been carried out over a long period of time. Roughly, about 10 pieces are excavated per pottery, mainly in the form of a mixture of several pieces.

Joining, which is the process of restoring the original shape of damaged ceramics, uses a method to find the original by matching each fragment at random by an expert.Therefore, there is a risk of damage due to the buildup of fatigue on the joint surface.The more complex the shape or the larger the number of fragments, the more the restoration process is. This lengthens and the original recovery rate decreases.

In order to compensate for this, a method of scanning fragments and realizing them as virtual objects, matching the virtual fragments and restoring them into virtual ceramics has been proposed. However, in order to realize each fragment as a virtual object, the work of attaching a marker, scanning from various fields, and matching the fragments required more labor and time for humans to directly fit the fragments. In addition, local geometric features are required to find the junction section in the porcelain fragment matching process. Fracture lines of ceramic fragments are mainly straight or similar curves, and it is very difficult to extract local features with high discrimination power. That is, there is a need for a matching method and a matching system for more efficient matching of large-capacity ceramic fragments.

The present invention was conceived to solve the above-described problem, and specifically, a virtual matching system that extracts shape features of ceramic fragments based on rotation axis information, and matches ceramic fragments using the shape features of the extracted ceramic fragments. to provide.

The apparatus for extracting shape features of pottery fragments according to an embodiment of the present invention receives 3D shape data for each pottery fragment generated by scanning a plurality of pottery fragments, and the 3D shape data of the pottery fragments A point cloud generation unit that generates point cloud data showing an appearance; A rotation axis estimating unit for estimating a rotation axis of each of the ceramic fragments; And a shape feature extraction unit for defining coordinates of the point cloud data by aligning the point cloud data of each ceramic fragment based on the estimated rotation axis, and defining shape features based on the rotation axis for the 3D shape data of each ceramic fragment. .

The virtual matching system of pottery fragments according to an embodiment of the present invention includes: a ceramic fragment shape data generating device for generating 3D shape data for each ceramic fragment by scanning a plurality of ceramic fragments; A point cloud generation unit for generating point cloud data showing the appearance of the ceramic shards from the three-dimensional shape data of the pottery shards; A rotation axis estimating unit for estimating a rotation axis of each of the ceramic fragments; And a shape feature extraction unit defining the coordinates of the point cloud data by aligning the point cloud data of each ceramic fragment based on the estimated rotation axis, and defining shape features based on the rotation axis for each ceramic fragment. Extraction device; And a porcelain fragment virtual matching device that performs virtual matching of the porcelain fragments by using the shape features based on the rotation axis.

According to an embodiment of the present invention, a method for virtually matching ceramic fragments includes the steps of: generating 3D shape data for each ceramic fragment by scanning a plurality of ceramic fragments; Generating point cloud data showing the appearance of the pottery shards from the three-dimensional shape data of the pottery shards; Estimating a rotation axis of each of the ceramic fragments; Defining coordinates of the point cloud data by aligning the point cloud data of each pottery fragment based on the estimated rotation axis, and defining shape characteristics based on the rotation axis for each pottery fragment; And performing virtual matching of the ceramic shards by using the shape features based on the rotation axis.

The computer program according to the embodiment of the present invention is combined with hardware and stored in a medium to execute the virtual matching method of the ceramic shards.

The virtual matching system of ceramic fragments according to the present embodiment extracts shape features of ceramic fragments based on a rotation axis, and performs virtual matching of large-capacity ceramic fragments based on the shape features of the ceramic fragments. That is, since samples of ceramic fragments to be matched are grouped in consideration of the shape characteristics of ceramic fragments, and matching between samples of ceramic fragments within a specific group, more efficient and accurate registration of ceramic fragments can be performed. .

1 is a block diagram showing the configuration of a virtual matching system of pottery fragments according to an embodiment of the present invention.
2 is an exemplary three-dimensional shape data of a piece of porcelain.
3 is a block diagram of an apparatus for extracting shape features of pottery fragments according to an embodiment of the present invention.
FIG. 4 shows point cloud data extracted from the 3D shape data of the ceramic fragment of FIG. 3.
5 is an exemplary view for explaining the height (H) of a point defined in a state in which point cloud data is aligned based on an estimated rotation axis.
6 is an exemplary view for explaining the distance (D) and angle (T) of a point defined in a state in which point cloud data is aligned based on an estimated rotation axis.
7 is a graph exemplarily showing a value obtained by calculating a shape feature based on a rotation axis.
8 is a flowchart of a method for virtually matching ceramic fragments according to another embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The detailed description to be disclosed hereinafter together with the accompanying drawings is intended to describe exemplary embodiments of the present invention, and is not intended to represent the only embodiments in which the present invention may be practiced. The following detailed description includes specific details to provide a thorough understanding of the present invention. However, one of ordinary skill in the art can appreciate that the invention may be practiced without these specific details. Specific terms used in the following description are provided to aid understanding of the present invention, and the use of these specific terms may be changed in other forms without departing from the technical spirit of the present invention.

1 is a block diagram showing the configuration of a virtual matching system of pottery fragments according to an embodiment of the present invention. 2 is an exemplary three-dimensional shape data of a piece of porcelain.

1 and 2, the virtual matching system 10 of pottery fragments according to an embodiment of the present invention includes a pottery fragment shape data generating device 100, a pottery fragment shape feature extraction device 110, and a pottery fragment A virtual matching device 120 is included.

The virtual matching system of porcelain fragments according to the embodiments may have an aspect that is entirely hardware, or partially hardware and partially software. For example, the virtual matching system of pottery fragments and each device included therein may collectively refer to a device for sending and receiving data of a specific format and content through an electronic communication method and software related thereto. In this specification, terms such as "unit", "module", "server", "system", "device" or "terminal" refer to a combination of hardware and software driven by the hardware. Is intended to be. For example, the hardware here may be a data processing device including a CPU or other processor. In addition, software driven by hardware may refer to an executing process, an object, an executable file, a thread of execution, a program, and the like.

Further, each device constituting the virtual matching system of porcelain fragments is not intended to necessarily refer to separate components that are physically distinct. The porcelain fragment shape data generating device 100, the porcelain fragment shape feature extraction device 110, and the porcelain fragment virtual matching device 120 are shown as separate blocks, which constitute a virtual matching system of porcelain fragments. It is just a functional separation of the configurations. Accordingly, according to an embodiment, the porcelain fragment shape data generating device 100, the porcelain fragment shape feature extraction device 110, and the porcelain fragment virtual matching device 120 may be partially or entirely integrated in the same device. , One or more may be implemented as separate devices that are physically separated from other units, and may be components that are communicatively connected to each other in a distributed computing environment.

The porcelain fragment shape data generating apparatus 100 generates 3D shape data by scanning ceramic fragments positioned on a plane portion. The planar portion may be a transparent material or an opaque material. For accurate image acquisition, the flat portion may be a monochromatic opaque material, for example, black. In addition, in order to obtain information on the upper and lower portions of the ceramic shards, the planar portion may be a transparent material. The porcelain fragment shape data generating apparatus 100 may prepare 3D shape data for a plurality of ceramic fragments. The porcelain fragment shape data generating apparatus 100 may scan the ceramic fragments one by one to generate 3D shape data of each ceramic fragment, but is not limited thereto, and after scanning a plurality of ceramic fragments at once, the generated 3-dimensional The process of separating the shape data can be performed. That is, there may be a plurality of pieces of porcelain fragments positioned on the flat portion, and may be a randomly selected sample, but the present invention is not limited thereto. Porcelain fragments may be multiple fragments excavated from similar locations and selected. A plurality of pottery fragments may be spaced apart from each other by a certain distance or more so as not to overlap and be positioned on a flat surface.

In one embodiment, the porcelain fragment shape data generating apparatus 100 may include a 3D scanner. The porcelain fragment shape data generating apparatus 100 may sense a plurality of ceramic fragments located on a plane portion at various angles and collect the sensed information to generate 3D shape data for the plurality of ceramic fragments. In addition, the porcelain fragment shape data generating apparatus 100 may separate each of the ceramic fragments from the 3D shape data of the plurality of ceramic fragments generated. 2 is an exemplary diagram illustrating 3D shape data generated by the apparatus 100 for generating ceramic fragment shape data.

In another embodiment, the porcelain fragment shape data generating apparatus 100 may include a depth (RGB-D) camera and a color (RGB) camera. The porcelain fragment shape data generating apparatus 100 may acquire an RGB image and a depth image of the fragment image by capturing the ceramic fragment located on the flat surface. The porcelain fragment shape data generating apparatus 100 may generate 3D shape data for a plurality of ceramic fragments based on an RGB image and a depth image.

The 3D shape data of the ceramic fragments generated by the ceramic fragment shape data generating apparatus 100 is provided to the shape feature extraction apparatus 110 of the ceramic fragments. The apparatus 110 for extracting shape features of ceramic fragments may define shape features based on a rotation axis with respect to 3D shape data of ceramic fragments. The porcelain fragment virtual matching device 120 may perform virtual matching of the ceramic fragment by using the defined shape features. That is, the order of joining can be effectively determined based on the shape feature based on the symmetry axis, so that time and cost for joining can be reduced.

Hereinafter, a process of extracting a shape feature based on a symmetrical axis in the shape feature extraction apparatus 110 of pottery fragments and a configuration constituting the shape feature extraction apparatus 110 of pottery fragments will be described in more detail.

3 is a block diagram of an apparatus for extracting shape features of pottery fragments according to an embodiment of the present invention. FIG. 4 shows point cloud data extracted from the 3D shape data of the ceramic fragment of FIG. 3. 5 is an exemplary diagram for explaining a height H defined in a state in which point cloud data is aligned based on an estimated rotation axis. 6 is an exemplary diagram for explaining a distance (D) and an angle (T) from an axis defined in a state in which point cloud data is aligned based on an estimated rotation axis. 7 is a graph exemplarily showing a value obtained by calculating a shape feature based on a rotation axis.

Referring to FIG. 3, the apparatus 110 for extracting shape features of ceramic fragments includes a point cloud generating unit 111, a rotation axis estimating unit 112, and a shape feature extracting unit 113.

The point cloud generator 111 may receive 3D shape data for each ceramic fragment generated by scanning a plurality of ceramic fragments from the ceramic fragment shape data generating apparatus 100. The point cloud generation unit 111 may generate point cloud data showing the appearance of the pottery shard from the 3D shape data of each pottery shard. FIG. 4 exemplarily shows point cloud data generated according to the appearance of the 3D shape data of the ceramic fragment of FIG. 2. That is, the point cloud data may be composed of a plurality of points defining the appearance of the 3D shape data of the pottery fragment. In addition, the point cloud generator 111 may exclude a portion of the defined point cloud data that is not a broken line. In other words, the lip part of the original pottery (Rim, the rim or edge where the human lip touches) or the base part (base, the base that supports the pottery), etc., of the pottery fragment's appearance, not the rupture line caused by the breakage of the pottery, match. This does not correspond to the appearance being performed. Accordingly, the point cloud generator 111 may exclude the above-described lip portion and/or base portion from the defined point cloud data from the point cloud data. That is, only the appearance corresponding to the break line can be defined as point cloud data.

The rotation axis estimating unit 112 estimates the rotation axis of each ceramic shard. Porcelain is manufactured based on one rotating shaft, and each piece of porcelain may also include information related to the rotating shaft of the porcelain before being damaged. The rotation axis estimating unit 112 may fit a shape obtained by transforming each piece of ceramic fragment into a very thin shape to a virtual circle and estimate a rotation axis based on the center of the virtual circle. In addition, the rotation axis estimating unit 112 may estimate the rotation axis of each ceramic piece by further considering at least one of a change amount of a normal vector and a change amount of curvature information of the ceramic piece.

The shape feature extraction unit 113 may define coordinates of the point cloud data by arranging the point cloud data of each ceramic fragment based on the estimated rotation axis. In addition, the shape feature extraction unit 113 may define shape features based on a rotation axis with respect to 3D shape data of each ceramic fragment. That is, the shape feature extraction unit 113 may locate the point cloud data of each ceramic fragment on the same three-dimensional coordinates (x, y, z), and extract the shape features of the ceramic fragment based on the three-dimensional coordinates. have. The three-dimensional coordinates are defined as the z-axis perpendicular to the x-y plane.

Specifically, the shape feature extraction unit 113 arranges the point cloud data so that the estimated rotation axis of each ceramic fragment coincides with the z-axis of the 3D coordinates. As shown in FIGS. 5 and 6, 3D shape data and point cloud data of each ceramic fragment may be displayed together, and the estimated rotation axis may be aligned on 3D coordinates so as to coincide with the z axis.

The coordinates of each point p of the point cloud data aligned so that the rotation axis coincides with the z-axis in the three-dimensional space is defined as in Equation 1 below.

[Equation 1]

In addition, the shape feature extraction unit 113 obtains a feature value representing the feature of each point p (height of the point (H(p)), distance (D(p)), angle (T(p)))), It is defined as in Equation 2 below.

[Equation 2]

Here, the height of the point is the vertical height along the z-axis as shown in FIG. 5, the distance of the point is the shortest distance from the coordinate center with respect to the xy plane as shown in FIG. 6, and the angle of the point is shown in FIG. It refers to an angle formed by a straight line according to the shortest distance from the x-axis in the xy plane as described above.

를 따라 미분한 데이터를 도자기 파편의 형상 특징으로 추출할 수 있다. 도자기 파편의 형상 특징(

)는 하기 수학식 3과 같이 정의된다.The shape feature extraction unit 113 may extract shape features of the pottery fragment by using the vertical height, distance, and angle of each point defined above. Specifically, the point cloud data

The differential data can be extracted as the shape feature of the ceramic fragment. Shape features of pottery fragments (

) Is defined as in Equation 3 below.

[Equation 3]

)은 각 포인트의 높이(H(p))를

를 따라 미분한 데이터, 제2 형상 특징(

)는 각 포인트의 거리(D(p))를

를 따라 미분한 데이터, 제3 형상 특징(

)은 각 포인트의 각도(T(p))를

를 따라 미분한 데이터에 각 포인트의 거리(D(p))를 곱한 값으로 정의될 수 있다. 제3 형상 특징의 연산에 각 포인트의 거리(D(p))가 고려됨에 따라 제1, 제2 형상 특징과의 단위가 통일되어 추후 형상 특징을 이용한 가상 정합 시 발생될 수 있는 불확실성이 보다 보완될 수 있다.First shape feature (

) Is the height of each point (H(p))

Data differentiated along the second shape feature (

) Is the distance of each point (D(p))

Data differentiated along the third shape feature (

) Is the angle of each point (T(p))

It may be defined as a value obtained by multiplying the differential data along with the distance D(p) of each point. As the distance (D(p)) of each point is considered in the calculation of the third shape feature, the units of the first and second shape features are unified, thereby further supplementing the uncertainty that may occur during virtual matching using the shape feature. Can be.

)의 생성시 발생되는 수치 미분 노이즈를 제거하기 위한 필터링을 더 수행할 수 있다. 예시적으로 가우시안 필터링이 수행되어 수치 미분의 노이즈를 제거할 수 있다. The shape feature extraction unit 113 is the shape feature (

Filtering to remove numerical differential noise generated when) is generated may be further performed. As an example, Gaussian filtering may be performed to remove noise of numerical differential.

7 is an exemplary view showing results of calculating a shape feature based on a rotation axis according to each point (before filtering: black line, after filtering: red line). It can be seen that the shape feature values for the points of the ceramic fragments constituting the ceramic fragment based on one ceramic fragment are calculated as shown in the graph of FIG. 7. In addition, it can be seen that the calculated first shape feature, the second shape feature, and the third shape feature are filtered to remove noise due to numerical differentiation.

)은 고정된 회전축의 기반으로 생성되어 방사상 거리 추정의 오류가 보완된 상태로 각 포인트의 특징을 나타내는 미분한 데이터일 수 있다. 즉, 본 발명의 일 실시예에 따른 도자기 파편의 형상 특징(

)는 도자기 파편의 외형의 기하학적 변화를 표현하는 특징 값으로, 직관적이고 신속한 가상 정합이 가능하게 한다. 각 도자기 파편에 대해 추출된 도자기 파편의 형상 특징(

)은 도자기 파편 가상 정합 장치(120)로 제공될 수 있다. Such, shape characteristics of the pottery fragment according to an embodiment of the present invention (

) May be differentiated data representing characteristics of each point in a state in which errors in radial distance estimation are compensated for by being generated based on a fixed rotation axis. That is, the shape characteristics of the pottery fragments according to an embodiment of the present invention (

) Is a feature value that expresses the geometric change of the outer shape of the ceramic shard, enabling intuitive and rapid virtual registration. Shape characteristics of the extracted pottery shards for each pottery shard (

) May be provided as a porcelain fragment virtual matching device 120.

The porcelain fragment virtual matching device 120 performs virtual matching of the porcelain fragments using the shape features of the porcelain fragments based on a rotation axis.

)도자기 파편의 외형의 기하학적 변화를 표현하는 특징 값이다. 따라서, 서로 다른 도자기 파편의 형상 특징 값이 유사 또는 동일하다는 것은 상기 도자기 파편들이 하나의 도자기에서 유래된 파편이라는 것뿐만 아니라, 정합이 가능한 상태라는 것도 의미할 수 있다. The porcelain fragment virtual matching device 120 may select a matching candidate for each porcelain fragment in consideration of shape characteristics. As described above, the shape features of the ceramic shards according to this embodiment (

) It is a characteristic value that expresses the geometric change of the outer shape of a ceramic fragment. Accordingly, the fact that the shape characteristic values of the different ceramic fragments are similar or the same may mean that the ceramic fragments are not only fragments derived from one ceramic, but also that they are in a state in which matching is possible.

)이 유사한 도자기 파편을 먼저 선별하여 그룹화할 수 있다. 예를 들어, 도자기 파편 가상 정합 장치(120)는 LCS(Longest common substring) 알고리즘을 이용하여 도자기 파편의 형상 특징(

)의 비교하여 매칭되는 도자기 파편을 선별하여 그룹화할 수 있다. 다음으로, 도자기 파편 가상 정합 장치(120)는 그룹화된 도자기 파편에서 실제 정합이 가능한 도자기 파편들을 선별하는 클러스터링 알고리즘을 수행할 수 있다. 예를 들어, 도자기 파편 가상 정합 장치(120)는 도자기 파편과 회전축 과의 거리를 더 고려하여 정합이 가능한 도자기 파편들을 선별하고, 정합이 불가능한 도자기 파편을 제외하는 과정을 더 수행할 수 있다. 도자기 파편 가상 정합 장치(120)는 상술한 과정을 통해 선별된 도자기 파편들 간에 가상 정합을 수행하게 된다.Therefore, the porcelain fragment virtual matching device 120 has the shape characteristics of the porcelain fragment based on the rotation axis (

Porcelain fragments with similar) can be selected and grouped first. For example, the porcelain fragment virtual matching device 120 uses the LCS (Longest common substring) algorithm to feature the shape characteristics of the porcelain fragment (

), you can select and group matching porcelain fragments. Next, the porcelain fragment virtual matching device 120 may perform a clustering algorithm for selecting ceramic fragments that can be actually matched from the grouped ceramic fragments. For example, the porcelain fragment virtual matching device 120 may further perform a process of selecting ceramic fragments that can be matched by further considering the distance between the ceramic fragments and the rotating shaft, and excluding the ceramic fragments that cannot be matched. The porcelain fragment virtual matching device 120 performs virtual matching between the porcelain fragments selected through the above-described process.

)을 추출하고, 이러한 도자기 파편의 형상 특징(

)을 기초로 대용량의 도자기 파편의 가상 정합을 수행한다. 즉, 정합을 시도하여야 하는 도자기 파편의 샘플을 도자기 파편의 형상 특징(

)을 고려하여 그룹화하고, 특정 그룹 내에서 도자기 파편의 샘플 간의 정합을 수행하기에 보다 효율적이고 정확한 도자기 파편의 정합이 수행될 수 있다.The virtual matching system of ceramic fragments according to an embodiment of the present invention includes the shape characteristics of ceramic fragments based on a rotation axis (

), and the shape features of these pottery fragments (

) To perform virtual matching of large-capacity ceramic shards. In other words, the sample of the pottery shards that should be matched is determined by

) In consideration of the grouping, and more efficient and accurate matching of porcelain debris can be performed to perform matching between samples of porcelain debris within a specific group.

Hereinafter, a method for virtually matching ceramic fragments according to another embodiment of the present invention will be described.

8 is a flowchart of a method for virtually matching ceramic fragments according to another embodiment of the present invention.

The virtual matching method of pottery fragments according to another embodiment of the present invention is a method performed by the above-described virtual matching system 10 of pottery fragments. Referring to FIG. 8, the virtual matching method of pottery fragments according to the present embodiment includes the steps of generating 3D shape data for each pottery fragment by scanning a plurality of pottery fragments (S100); Generating point cloud data showing the appearance of the pottery shards from the three-dimensional shape data of the pottery shards (S110); Estimating the rotation axis of each of the ceramic fragments (S120); Defining coordinates of the point cloud data by arranging the point cloud data of each ceramic fragment based on the estimated rotation axis, and defining shape characteristics based on the rotation axis for each ceramic fragment (S130); And performing virtual matching of the ceramic fragments by using the shape feature based on the rotation axis (S140).

First, a plurality of pottery fragments are scanned to generate 3D shape data for each pottery fragment (S100).

Three-dimensional shape data is generated by scanning ceramic fragments located on the plane. Three-dimensional shape data for each of the plurality of ceramic fragments may be prepared. The step (S100) is performed in the porcelain fragment shape data generating apparatus 100 of the virtual matching system 10 of the porcelain fragments. The porcelain fragment shape data generating apparatus 100 may scan the ceramic fragments one by one to generate 3D shape data of each ceramic fragment, but is not limited thereto, and after scanning a plurality of ceramic fragments at once, the generated 3-dimensional The process of separating the shape data can be performed.

Next, point cloud data showing the appearance of the pottery shard is generated from the three-dimensional shape data of each pottery shard (S110).

The step (S110) is performed by the apparatus 110 for extracting shape features of the ceramic fragments of the virtual matching system 10 of the ceramic fragments. The apparatus 110 for extracting shape features of pottery fragments may generate point cloud data showing the appearance of pottery fragments from 3D shape data of each pottery fragment. The point cloud data may be composed of a plurality of points defining the appearance of the three-dimensional shape data of the ceramic shard. In the point cloud data, parts that are not broken lines can be excluded. For example, in the defined point cloud data, the aforementioned lip portion and/or base portion, etc. may be excluded from the point cloud data. That is, only the appearance corresponding to the break line can be defined as point cloud data.

Next, the rotation axis of each of the ceramic fragments is estimated (S120).

The step (S120) is performed in the apparatus 110 for extracting shape features of pottery fragments. The apparatus 110 for extracting shape features of pottery fragments may fit a shape obtained by transforming each fragment of pottery into a very thin shape to a virtual circle and estimate a rotation axis based on the center of the virtual circle. In addition, the apparatus 110 for extracting shape features of ceramic fragments may estimate a rotation axis of each ceramic fragment by further considering at least one of a change amount of a normal vector and a change amount of curvature information of the ceramic fragment.

Subsequently, the coordinates of the point cloud data are defined by aligning the point cloud data of each pottery fragment based on the estimated rotation axis, and shape characteristics based on the rotation axis are defined for each pottery fragment (S130).

The step (S130) is performed in the apparatus 110 for extracting shape features of pottery fragments. The shape feature extraction device 110 of the ceramic fragment locates the point cloud data of each ceramic fragment on the same three-dimensional coordinates (x, y, z), and extracts the shape features of the ceramic fragment based on the three-dimensional coordinates. I can. The three-dimensional coordinates are defined as the z-axis perpendicular to the x-y plane. The apparatus 110 for extracting shape features of pottery fragments aligns the point cloud data so that the estimated rotation axis of each pottery fragment coincides with the z-axis of the 3D coordinates. The coordinates of each point p of the point cloud data aligned so that the rotation axis coincides with the z-axis in the three-dimensional space is defined as in Equation 1 below.

[Equation 1]

In addition, the shape feature extraction device 110 of the pottery fragment is a feature value representing the feature of each point (p) (height of the point (H(p)), distance (D(p))), angle (T(p)) ] Is defined as in Equation 2 below.

[Equation 2]

를 따라 미분한 데이터를 도자기 파편의 형상 특징으로 추출할 수 있다. 도자기 파편의 형상 특징(

)는 하기 수학식 3과 같이 정의된다.The shape feature extraction device 110 of the ceramic fragment may extract the shape features of the ceramic fragment by using the vertical height, distance, and angle of each point defined above. Specifically, the point cloud data

The differential data can be extracted as the shape feature of the ceramic fragment. Shape features of pottery fragments (

) Is defined as in Equation 3 below.

[Equation 3]

)은 각 포인트의 높이(H(p))를

를 따라 미분한 데이터, 제2 형상 특징(

)는 각 포인트의 거리(D(p))를

를 따라 미분한 데이터, 제3 형상 특징(

)은 각 포인트의 각도(T(p))를

를 따라 미분한 데이터에 각 포인트의 거리(D(p))를 곱한 값으로 정의될 수 있다. 제3 형상 특징의 연산에 각 포인트의 거리(D(p))가 고려됨에 따라 제1, 제2 형상 특징과의 단위가 통일되어 추후 형상 특징을 이용한 가상 정합 시 발생될 수 있는 불확실성이 보다 보완될 수 있다. 또한, 도자기 파편의 형상 특징 추출 장치(110)는 도자기 파편의 형상 특징(

)의 생성시 발생되는 수치 미분 노이즈를 제거하기 위한 필터링을 더 수행할 수 있다. 예시적으로 가우시안 필터링이 수행되어 수치 미분의 노이즈를 제거할 수 있다.First shape feature (

) Is the height of each point (H(p))

Data differentiated along the second shape feature (

) Is the distance of each point (D(p))

Data differentiated along the third shape feature (

) Is the angle of each point (T(p))

It may be defined as a value obtained by multiplying the differential data along with the distance D(p) of each point. As the distance (D(p)) of each point is considered in the calculation of the third shape feature, the units of the first and second shape features are unified, thereby further supplementing the uncertainty that may occur during virtual matching using the shape feature. Can be. In addition, the shape feature extraction device 110 of the pottery fragment is the shape feature of the pottery fragment (

Filtering to remove numerical differential noise generated when) is generated may be further performed. As an example, Gaussian filtering may be performed to remove noise of numerical differential.

Virtual matching of the ceramic shards is performed using a shape feature based on a rotation axis (S140).

)이 유사한 도자기 파편을 먼저 선별하여 그룹화할 수 있다. 예를 들어, 도자기 파편 가상 정합 장치(120)는 LCS(Longest common substring) 알고리즘을 이용하여 도자기 파편의 형상 특징(

)의 비교하여 매칭되는 도자기 파편을 선별하여 그룹화할 수 있다. 다음으로, 도자기 파편 가상 정합 장치(120)는 그룹화된 도자기 파편에서 실제 정합이 가능한 도자기 파편들을 선별하는 클러스터링 알고리즘을 수행할 수 있다. 예를 들어, 도자기 파편 가상 정합 장치(120)는 도자기 파편과 회전축 과의 거리를 더 고려하여 정합이 가능한 도자기 파편들을 선별하고, 정합이 불가능한 도자기 파편을 제외하는 과정을 더 수행할 수 있다. 도자기 파편 가상 정합 장치(120)는 상술한 과정을 통해 선별된 도자기 파편들 간에 가상 정합을 수행하게 된다.The step (S140) is performed in the porcelain fragment virtual matching device 120. The porcelain fragment virtual matching device 120 has the shape characteristics of the porcelain fragment based on the rotation axis (

Porcelain fragments with similar) can be selected and grouped first. For example, the porcelain fragment virtual matching device 120 uses the LCS (Longest common substring) algorithm to feature the shape characteristics of the porcelain fragment (

), you can select and group matching porcelain fragments. Next, the porcelain fragment virtual matching device 120 may perform a clustering algorithm for selecting ceramic fragments that can be actually matched from the grouped ceramic fragments. For example, the porcelain fragment virtual matching device 120 may further perform a process of selecting ceramic fragments that can be matched by further considering the distance between the ceramic fragments and the rotating shaft, and excluding the ceramic fragments that cannot be matched. The porcelain fragment virtual matching device 120 performs virtual matching between the porcelain fragments selected through the above-described process.

)을 추출하고, 이러한 도자기 파편의 형상 특징(

)을 기초로 대용량의 도자기 파편의 가상 정합을 수행한다. 즉, 정합을 시도하여야 하는 도자기 파편의 샘플을 도자기 파편의 형상 특징(

)을 고려하여 그룹화하고, 특정 그룹 내에서 도자기 파편의 샘플 간의 정합을 수행하기에 보다 효율적이고 정확한 도자기 파편의 정합이 수행될 수 있다.The virtual matching method of pottery fragments according to the present embodiment includes the shape characteristics of pottery fragments based on a rotation axis (

), and the shape features of these pottery fragments (

) To perform virtual matching of large-capacity ceramic shards. In other words, the sample of the pottery shards that should be matched is determined by

) In consideration of the grouping, and more efficient and accurate matching of porcelain debris can be performed to perform matching between samples of porcelain debris within a specific group.

The operation by the virtual matching method of pottery fragments according to the above-described embodiments may be implemented at least partially as a computer program and recorded on a computer-readable recording medium. A computer-readable recording medium in which a program for implementing an operation by the virtual matching method of pottery fragments according to the embodiments is recorded and includes all types of recording devices storing data that can be read by a computer. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tapes, floppy disks, and optical data storage devices. In addition, the computer-readable recording medium may be distributed over a computer system connected through a network, and computer-readable codes may be stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for implementing this embodiment may be easily understood by those skilled in the art to which this embodiment belongs.

Although the above has been described with reference to the embodiments, the present invention should not be construed as being limited by these embodiments or drawings, and those skilled in the art will have the spirit and scope of the present invention described in the following claims. It will be understood that various modifications and changes can be made to the present invention within the range not departing from.

10: Virtual Matching System of Porcelain Fragments
100: porcelain fragment shape data generation device
111: point cloud generation unit
112: rotation axis estimation unit
113: shape feature extraction unit
110: Apparatus for extracting shape features of ceramic fragments
120: porcelain fragment virtual matching device

Claims (10)

A point cloud generator configured to receive 3D shape data for each pottery fragment generated by scanning a plurality of pottery fragments, and to generate point cloud data showing the appearance of the pottery fragment from the 3D shape data of the pottery fragment;
A rotation axis estimating unit for estimating a rotation axis of each of the ceramic fragments; And
It includes a shape feature extraction unit for defining the coordinates of the point cloud data by aligning the point cloud data of each ceramic fragment based on the estimated rotation axis, and defining shape features based on the rotation axis for the three-dimensional shape data of each ceramic fragment,
The shape feature extraction unit aligns the point cloud data so that the estimated rotation axis of each porcelain fragment coincides with the z axis of the three-dimensional space (x, y, z),
The coordinates in the three-dimensional space of the point cloud data are defined as in Equation 1 below, and the height (H(p)), distance (D(p)), and angle (T(p) of each point of the point cloud data) )) is defined as in Equation 2 below,

[Equation 1]

[Equation 2]

The shape feature extraction unit is characterized in that the shape feature is defined as shown in Equation 3 below by using the height (H(p)), distance (D(p)), and angle (T(p)) of each point. Apparatus for extracting shape features of ceramic fragments.

[Equation 3]

delete delete The method of claim 1,
The shape feature extraction unit further performs filtering to remove the numerical differential noise generated when the shape feature is generated. A porcelain fragment shape data generating device that scans a plurality of ceramic fragments and generates 3D shape data for each ceramic fragment;
A point cloud generation unit for generating point cloud data showing the appearance of the ceramic shards from the three-dimensional shape data of the pottery shards; A rotation axis estimating unit for estimating a rotation axis of each of the ceramic fragments; And a shape feature extraction unit defining the coordinates of the point cloud data by aligning the point cloud data of each ceramic fragment based on the estimated rotation axis, and defining shape features based on the rotation axis for each ceramic fragment. Extraction device; And
A ceramic shard virtual matching device for performing virtual registration of pottery shards using the shape features based on the rotation axis,
The shape feature extraction unit aligns the point cloud data so that the estimated rotation axis of each porcelain fragment coincides with the z axis of the three-dimensional space (x, y, z),
The coordinates in the three-dimensional space of the point cloud data are defined as in Equation 1 below, and the height (H(p)), distance (D(p)), and angle (T(p) of each point of the point cloud data) )) is defined as in Equation 2 below,

[Equation 1]

[Equation 2]

The shape feature extraction unit is characterized in that the shape feature is defined as shown in Equation 3 below by using the height (H(p)), distance (D(p)), and angle (T(p)) of each point. Virtual matching system of pottery fragments.

[Equation 3]

The method of claim 5,
The porcelain fragment shape data generating apparatus comprises a 3D scanner to sense the plurality of ceramic fragments at various angles, and collects the sensed information to generate 3D shape data for each of the ceramic fragments. Fragment virtual matching system. The method of claim 5,
The porcelain fragment virtual matching device,
Porcelain fragments having similar shape characteristics based on the rotating shaft are first selected and grouped,
A virtual matching system for porcelain debris, characterized in that to perform a clustering algorithm to further select porcelain debris that can be matched from the grouped porcelain debris. The method of claim 7,
The porcelain debris virtual matching device is a porcelain debris virtual matching system, characterized in that by comparing the shape characteristics of the porcelain debris using an LCS (Longest common substring) algorithm to select and group matching porcelain debris. Generating 3D shape data for each of the ceramic fragments by scanning the plurality of ceramic fragments;
Generating point cloud data showing the appearance of the pottery shards from the three-dimensional shape data of the pottery shards;
Estimating a rotation axis of each of the ceramic fragments;
Defining coordinates of the point cloud data by aligning the point cloud data of each pottery fragment based on the estimated rotation axis, and defining shape characteristics based on the rotation axis for each pottery fragment; And
Including the step of performing virtual matching of the ceramic fragments by using the shape feature based on the rotation axis,
The defining of the shape characteristics includes arranging the point cloud data so that the estimated rotation axis of each porcelain fragment coincides with the z axis of the 3D space (x, y, z), and the 3D space of the point cloud data The coordinates within are defined as in Equation 1 below, and the height (H(p)), distance (D(p)), and angle (T(p)) of each point of the point cloud data are as shown in Equation 2 below. Define,

[Equation 1]

[Equation 2]

The step of defining the shape feature includes defining the shape feature as shown in Equation 3 below using the height (H(p)), distance (D(p)), and angle (T(p)) of each point. How to virtually match ceramic fragments.

[Equation 3]

A computer program combined with hardware and stored in a medium to execute the method of virtually matching ceramic shards according to claim 9.

Images