KR101827847B1 - Three dimensional image restoring device and three dimensional image restoring method - Google Patents

Abstract

A three-dimensional image restoration apparatus includes a device for restoring a three-dimensional image using a plurality of first three-dimensional point coordinates acquired by photographing a target object at a plurality of points of view using a stereo scanning system including a first point- For each of the plurality of viewpoints using the plurality of first three-dimensional point coordinates, a first unwrapped phase corresponding to the first viewpoint device and a second unfolded phase corresponding to the second viewpoint device And a second three-dimensional point obtaining unit that obtains a plurality of second three-dimensional point coordinates that minimize a first function that is a difference between the first expanded phase and the second expanded phase with respect to the plurality of view points, And a coordinate acquiring unit.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a three-dimensional image restoration apparatus and a three-

The present invention relates to a three-dimensional image restoration apparatus and a three-dimensional image restoration method.

3D graphics, reverse engineering, and so on, various high-quality 3D model acquisition methods are being studied.

Some existing methods (non-patent documents 1 and 2) focus on achieving high visual quality rather than metric accuracy. The human visual system is not sensitive to small scale errors and therefore visualization applications in these graphics fields do not require high precision.

However, many applications require micro-accuracy, such as reverse engineering, dental scanners for implants, and flaw detection systems. Especially, in case of precise replication using a 3D printer or a milling machine, a very precise 3D model is needed to alleviate accumulation errors in scanning and printing.

1. T. Beeler, B. Bickel, P. Beardsley, B. Sumner, and M. Gross. High-quality single-shot capture of facial geometry. ACM Transactions on Graphics (SIGGRAPH), 2010. 2. J. Sturm, E. Bylow, F. Kahl, and D. Cremers. CopyMe3D: Scanning and printing persons in 3D. In German Conference on Pattern Recognition (GCPR), 2013.

A technical problem to be solved is to provide a three-dimensional image restoration device and a three-dimensional image restoration method capable of restoring a three-dimensional structure whose accuracy has been increased by software without changing existing hardware.

A three-dimensional image restoration apparatus according to an exemplary embodiment of the present invention includes a plurality of first three-dimensional point coordinates obtained by photographing a target object at a plurality of points of view using a stereo scanning system including a first point- Can be restored. Wherein the three-dimensional image restoration device is configured to perform, for each of the plurality of viewpoints using the plurality of first three-dimensional point coordinates, a first unwrapped phase corresponding to the first viewpoint device and a second unopphased phase corresponding to the second viewpoint device And acquiring a plurality of second three-dimensional point coordinates for making the first function, which is the difference between the first expanded phase and the second expanded phase, minimum for the plurality of points of view Dimensional point coordinate obtaining unit.

Wherein the three-dimensional image reconstructing apparatus comprises an intensity vector obtaining unit for obtaining a plurality of intensity vectors for each of the plurality of second three-dimensional point coordinates, and an intensity-vector obtaining unit for obtaining a plurality of incident light direction matrices for each of the plurality of second three- A matrix normalizing unit for obtaining a plurality of surface normal vectors for each of the plurality of second three-dimensional point coordinates using the plurality of intensity vectors and the corresponding plurality of incident light direction matrices; Dimensional mesh structure obtaining unit for obtaining a three-dimensional mesh structure for the target object by using the second three-dimensional point coordinates of the target object and the corresponding plurality of surface normal vectors.

The first viewpoint device may be a first camera and the second viewpoint device may be a second camera or a projector.

Wherein the expanded phase obtaining unit obtains the first expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the first point-of-view device to the plurality of first three-dimensional point coordinates for each point in time, And obtain the second expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the second viewpoint device to the plurality of first three-dimensional point coordinates.

The second three-dimensional point coordinate obtaining unit may obtain the plurality of second three-dimensional point coordinates by obtaining a displacement value of the three-dimensional coordinates that minimizes a second function obtained by linearizing the first function.

The intensity vector obtaining unit may obtain an intensity vector by interpolating the intensity of corresponding two-dimensional point coordinates at the plurality of points with respect to each of the plurality of second three-dimensional point coordinates.

Each of the plurality of incident light direction matrices may include a plurality of incident light direction vectors from a light source center of each of the plurality of viewpoints to a corresponding second three-dimensional point coordinate.

The surface normal vector obtaining unit may obtain a surface normal vector that minimizes a third function which is a difference between an intensity vector, a surface normal vector and an incident light direction matrix matrix multiplication for each of the plurality of second three-dimensional point coordinates.

The three-dimensional mesh structure obtaining unit may obtain the three-dimensional mesh structure by using a plurality of oriented points that directly assign a surface normal vector corresponding to the plurality of second three-dimensional point coordinates.

A three-dimensional image restoration method according to an embodiment of the present invention uses a plurality of first three-dimensional point coordinates obtained by photographing a target object at a plurality of points of time using a stereo scanning system including a first point-of-view device and a second point-and-point device.

The three-dimensional image restoration method may further include a step of, for each of the plurality of viewpoints using the plurality of first three-dimensional point coordinates, a first unwrapped phase corresponding to the first viewpoint device and a second unwrapped phase corresponding to the second viewpoint device And obtaining a plurality of second three-dimensional point coordinates that minimize the first function, which is the difference between the first expanded phase and the second expanded phase, for the plurality of points of view .

Wherein the 3D reconstruction method further comprises: obtaining a plurality of intensity vectors for each of the plurality of second three-dimensional point coordinates; obtaining a plurality of incident light direction matrices for each of the plurality of second three-dimensional point coordinates; Obtaining a plurality of surface normal vectors for each of the plurality of second three-dimensional point coordinates using the vector and the corresponding plurality of incident light direction matrices, and obtaining the plurality of second three-dimensional point coordinates and the corresponding plurality And acquiring a three-dimensional mesh structure for the target object using the surface normal vector of the target object.

Wherein the obtaining of the first and second expanded phases comprises: obtaining the first expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the first point-of-view device to the plurality of first three- And obtaining the second expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the second point-in-time device to the plurality of first three-dimensional point coordinates for each point in time .

The obtaining of the plurality of second three-dimensional point coordinates may include obtaining the plurality of second three-dimensional point coordinates by obtaining a displacement value of the three-dimensional coordinate that minimizes a second function that linearizes the first function have.

The obtaining of the plurality of intensity vectors may include acquiring an intensity vector by interpolating the intensity of the corresponding two-dimensional point coordinates at the plurality of points for each of the plurality of second three-dimensional point coordinates .

Each of the plurality of incident light direction matrices may include a plurality of incident light direction vectors from a light source center of each of the plurality of viewpoints to a corresponding second three-dimensional point coordinate.

The obtaining of the plurality of surface normal vectors may include obtaining a surface normal vector that minimizes a third function that is a difference between a matrix vector of an intensity vector, a surface normal vector, and an incident light direction matrix, for each of the plurality of second three- And a step of acquiring.

The acquiring of the three-dimensional mesh structure may acquire the three-dimensional mesh structure using a plurality of directivity points to which surface normal vectors corresponding to the plurality of second three-dimensional point coordinates are directly allocated.

The three-dimensional image restoration apparatus and the three-dimensional image restoration method according to the embodiments can restore the three-dimensional structure in which the precision is increased without changing the existing hardware.

1 is a view for explaining a stereo scanning system according to an embodiment of the present invention.
2 is a view for explaining a stereo scanning system according to another embodiment of the present invention.
3 is a view for explaining a three-dimensional image restoration apparatus according to an embodiment of the present invention.
4 is a diagram for explaining a plurality of first three-dimensional point coordinates for a target object.
5 is a view for explaining a plurality of second three-dimensional point coordinates according to an embodiment of the present invention.
6 is a view for explaining a plurality of points of interest according to an embodiment of the present invention.
7 is a view for explaining a three-dimensional mesh structure according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. The embodiments may be implemented in various different forms and are not limited to the embodiments described herein.

In order to clearly illustrate the embodiments, parts not related to the description are omitted, and the same reference numerals are used for the same or similar components throughout the specification. Therefore, reference numerals of the components used in the previous drawings can be used in the following drawings.

1 is a view for explaining a stereo scanning system according to an embodiment of the present invention.

Referring to FIG. 1, a stereo scanning system according to an exemplary embodiment of the present invention includes a first camera C11, a projector P1, and a second camera C12. In the embodiment of FIG. 1, the first camera C11 is a first viewpoint device and the second camera C12 is a second viewpoint device.

The first camera C11, the projector P1, and the second camera C12 may be fixed to a rigid body with a predetermined gap therebetween. The stereo scanning system may further include a rotary table which rotates the target object 21 or rotates the first camera C11, the projector P1, and the second camera C12, To rotate around the target object (21). Therefore, the stereo scanning system according to the present embodiment can irradiate structured light to the target object 21 at a plurality of viewpoints (viewpoint 11, viewpoint 12, viewpoint 13) and photograph it. The plurality of viewpoints is not limited to three, and may have a larger number of viewpoints.

The projector P1 irradiates a structured light to the object 21. Such structured light may have various patterns. For example, the structured light pattern may be composed of 16 horizontal patterns including eight gray-code patterns and nine phase-shifting patterns. When using such structured light patterns, dense corresponding points can be estimated based on a standard phase unwrapping method (J. Salvi, S. Fernandez, T. Pribanic, and X. Llado. A state of the art in structured light patterns for surface profilometry, Pattern Recognition, 2010.).

The first camera C11 and the second camera C12 photograph a target object 21 irradiated with the structured light at a plurality of viewpoints (viewpoint 11, viewpoint 12, viewpoint 13), reconstruct a three- To the device (10).

The 3D image reconstruction apparatus 10 can obtain a plurality of unwrapped phase images for a plurality of viewpoints (viewpoint 11, viewpoint 12, viewpoint 13) using the captured image information. The 3D image reconstruction apparatus 10 reconstructs a plurality of initial depth images for a plurality of viewpoints (viewpoint 11, viewpoint 12, viewpoint 13) based on the plurality of expanded phase images and calibration data, Lt; / RTI >

The 3D image reconstruction apparatus 10 merges the plurality of initial depth images using an ICP (Iterative Closest Point) method or the like to thereby generate a plurality of first three-dimensional point coordinates 400 corresponding to the target object 21 (See FIG. 4).

Hereinafter, a plurality of second three-dimensional point coordinates are obtained using a plurality of first three-dimensional point coordinates 400, a plurality of expanded phase images, and calibration data obtained according to the above description, and the plurality of second three- A method of acquiring a three-dimensional mesh structure will be described.

2 is a view for explaining a stereo scanning system according to another embodiment of the present invention.

The stereo scanning system of Fig. 2 includes a camera C21 and a projector P2. In the embodiment of FIG. 2, the camera C21 is the first viewpoint device and the projector P2 is the second viewpoint device.

In the embodiment of FIG. 2, since only one camera C21 is provided, the following method can be applied assuming that the projector P2 is the second camera. At this time, the projector P2 can be modeled as a pin-hole camera.

3 is a view for explaining a three-dimensional image restoration apparatus according to an embodiment of the present invention. The following description will be made on the basis of the embodiment of FIG.

3, the three-dimensional image reconstruction apparatus 10 according to an embodiment of the present invention includes an expanded phase obtaining unit 11, a second three-dimensional point coordinate obtaining unit 12, an intensity vector obtaining unit 13, Direction matrix obtaining unit 14, a surface normal vector obtaining unit 15, and a three-dimensional mesh structure obtaining unit 16.

The 3D image restoration device 10 may be a computing device capable of computing a program such as a desktop, a notebook, and a server. The 3D image reconstruction apparatus 10 may include at least one processor and a memory. The 3D image reconstruction method according to the embodiment of the present invention may be pre-programmed in the memory. The processor can drive such a program. On the other hand, such a program may be imported remotely from an external server or the like. On the other hand, the distinction between the processor and the memory may be ambiguous in the three-dimensional image restoration apparatus 10, and it is possible for a person skilled in the art to determine the proportion of software and hardware for implementing the three-dimensional image restoration method.

The three-dimensional image restoration apparatus 10 may implement the three-dimensional image restoration method of the present invention with one processor and one memory. However, the three-dimensional image restoration apparatus 10 may use a plurality of processors, a plurality of memories, Can be implemented. 3, the three-dimensional image restoration apparatus 10 includes a plurality of function units 11, 12, 13, 14, 15, and 16 logically divided according to a three-dimensional image restoration method, . It is optional for those skilled in the art to integrate or separate each functional unit into one or a specific number and implement it in hardware and software.

The expanded phase acquiring unit 11 acquires the first expanded coordinates corresponding to the first camera C11 for each of a plurality of viewpoints (viewpoint 11, viewpoint 12, viewpoint 13) using the plurality of first three- An unwrapped phase and a second expanded phase corresponding to the second camera C12. At this time, the expanded phase obtaining unit 11 obtains the first expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the first camera C11 to the plurality of first three-dimensional point coordinates 400 for each viewpoint And obtain the second expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the second camera C12 to the plurality of first three-dimensional point coordinates 400 for each viewpoint.

라 하고, 시점 i에 대한 외부(extrinsic) 카메라 파라미터를

라 하고, 시점 i에 대한 내부(intrinsic) 카메라 파라미터를

라고 하는 경우,

에 대응하는 시점 i의 이차원 점 좌표는

라고 할 수 있다. 이러한 관계는 아래 수학식 1과 같이 표현될 수 있다.Each of the plurality of first three-dimensional point coordinates (400)

, And extrinsic camera parameters for the time i

, And the intrinsic camera parameters for the time i

If so,

The two-dimensional point coordinates at the time point i corresponding to

. This relationship can be expressed as Equation 1 below.

[Equation 1]

및 내부 파라미터

는 전술한 캘리브레이션 데이터로부터 획득할 수 있다.Camera external parameters

And internal parameters

Can be obtained from the above-described calibration data.

에 대한 펼친 위상을 삼차원 점 좌표로 표현하면

라고 할 수 있다.These two-dimensional point coordinates

Expressed in terms of three-dimensional point coordinates

.

에 대해 두 개의 펼친 위상을 획득할 수 있다. 이때, 제1 카메라(C11)에 대응하는 펼친 위상을 제1 펼친 위상

이라고 하고, 제2 카메라(C12)에 대응하는 펼친 위상을 제2 펼친 위상

이라고 할 수 있다.The stereo scanning system according to the embodiment of the present invention includes a first camera C11 and a second camera C12 which are different from each other and the camera parameters of the first camera C11 and the second camera C12 are Respectively. Therefore, in accordance with the above-described formula (1), when one three-dimensional point coordinate

Lt; / RTI > can obtain two < RTI ID = 0.0 > At this time, the expanded phase corresponding to the first camera C11 is referred to as a first extended phase

And the expanded phase corresponding to the second camera C12 is referred to as a second extended phase

.

The second three-dimensional point coordinate acquiring section 12 acquires a plurality of points (point 11, point 12, point 13) for causing the first function, which is the difference between the first and second expanded phases, Two three-dimensional point coordinates 500 can be obtained (see FIG. 5).

및 도출된 제2 삼차원 점 좌표

의 관계는 아래 수학식 2와 같이 표현될 수 있다.The exemplary first function

And the derived second three-dimensional point coordinates

Can be expressed by Equation (2) below.

&Quot; (2) "

은 이진 시각성 함수(binary visibility function)로서 제1 삼차원 점 좌표

가 i 시점의 이미지에서 보이는 경우 1이고, 보이지 않는 경우 0이다.

Is a binary visibility function, which is a first three-dimensional point coordinate

Is 1 when viewed from the image at the i-th point, and 0 when it is not visible.

Equation (2) corresponds to a non-convex optimization problem due to the non-linearity of the phase mapping. Equation (2) finds solutions through various techniques, but in the present embodiment solves this with an exemplary linearization technique described below.

를 최소화하는 삼차원 좌표의 변위 값

을 획득함으로써 복수의 제2 삼차원 점 좌표(500)를 획득할 수 있다. 이는 아래 수학식 3으로 표현될 수 있다.The second three-dimensional point coordinate acquiring unit 12 acquires a second function, which is a linear function of the first function,

The displacement value of the three-dimensional coordinates

The second three-dimensional point coordinates 500 can be obtained. This can be expressed by the following equation (3).

&Quot; (3) "

는 초기 추측(initial guess) 좌표이고,

는 수학식

에 기초하고,

는

에 대한 자코비안 표현(Jacobian representation)이다.

Is an initial guess coordinate,

Is expressed by the following equation

Lt; / RTI >

The

Is a Jacobian representation of.

를 삼차원 좌표 축(3D coordinate axis)을 따라 수치적으로(numerically) 계산하고, 각 단계 별

를 최소 자승법 방식(least squares manner)으로 계산한다.

은 각 단계별로

를 통해 업데이트된다.The second three-dimensional point coordinate obtaining unit 12 obtains

Is calculated numerically along the 3D coordinate axis, and is calculated for each step

In a least squares manner.

By each step

Lt; / RTI >

4. Compared to the first three-dimensional point coordinates 400 of FIG. 4 that follow the average shape, the plurality of second three-dimensional point coordinates 500 of FIG. 5 have the same shape as the microstructures 400 through the geometric refinement And it can be confirmed that it is sharply corrected.

Based on the plurality of second three-dimensional point coordinates 500, a three-dimensional mesh structure is obtainable through an existing geometric method (e.g., deriving a normal vector using the coordinates of surrounding pixels of the target pixel) .

However, applying these existing geometric methods results in averaging artifacts that make it difficult to determine the proper radius to define the surrounding pixels and smooth the geometric details.

Therefore, in this embodiment, further, a process of acquiring a fine three-dimensional mesh structure 700 through a photometric method will be described below (see FIG. 7).

The intensity vector obtaining unit 13 obtains a plurality of intensity vectors for each of the plurality of second three-dimensional point coordinates 500. At this time, the intensity vector obtaining unit 13 interpolates the intensity of corresponding two-dimensional point coordinates at a plurality of points of time (point 11, point 12, point 13) with respect to each of the plurality of second three-dimensional point coordinates 500, The intensity vector can be obtained.

에 대해서 대응하는 이차원 점 좌표들의 픽셀 강도를 보간함으로써 강도 벡터

를 획득할 수 있다.

는 해당 제2 삼차원 점 좌표

가 보이는 시점의 개수를 의미한다.The intensity vector acquiring unit 13 acquires the intensity vector of the arbitrary second three-

By interpolating the pixel intensity of the corresponding two-dimensional point coordinates with respect to the intensity vector

Can be obtained.

The second three-dimensional point coordinates

Quot; is the number of viewpoints that are visible.

은 복수의 시점 각각의 광원(light source) 중심으로부터 대응하는 제2 삼차원 점 좌표로 향하는 복수의 입사광 방향 벡터를 포함한다. 이때 광원의 중심이란, 도 1의 실시예에서 프로젝터(P1)의 중심을 의미한다. 따라서 본 실시예에 따르면 외부 광원이 불필요하여 비용이 절감될 수 있다.The incident light direction matrix obtaining unit 14 obtains a plurality of incident light direction matrixes for each of the plurality of second three-dimensional point coordinates. At this time, each of the plurality of incident light directional matrices

Includes a plurality of incident light direction vectors from the light source center of each of the plurality of viewpoints to the corresponding second three-dimensional point coordinates. Here, the center of the light source means the center of the projector P1 in the embodiment of FIG. Therefore, according to the present embodiment, an external light source is unnecessary and the cost can be reduced.

The surface normal vector obtaining unit 15 obtains a plurality of surface normal vectors for each of the corresponding second plurality of three-dimensional point coordinates using the plurality of intensity vectors and the corresponding plurality of incident light direction matrices . At this time, the surface normal vector obtaining unit 15 obtains a surface normal vector for minimizing the third function, which is a difference between the intensity vector, the surface normal vector, and the incident light direction matrix, for each of the plurality of second three-dimensional point coordinates . Details will be described later.

, 입사광 방향 행렬

, 및 표면 법선 벡터

에 대해서 아래 수학식 4의 관계가 도출될 수 있다.Under the Lambertian shading assumption, the obtained intensity vector < RTI ID = 0.0 >

, An incident light direction matrix

, And surface normal vectors

The following relation (4) can be derived.

&Quot; (4) "

를 최소화하는 표면 법선 벡터

를 획득할 수 있다. 이때, 표면 법선 벡터를 그림자(shadow) 또는 포화(saturation)에 대해 강건하게(robust) 추정하기 위하여, 가중치

를 도입할 수 있다. 이때 가중치

는

일 때

를 만족할 수 있다. 이러한 가중치를 도입한 수정된 제3 함수는

로 표현될 수 있다.The surface normal vector obtaining unit 15 can obtain the surface normal vector by the least squares method. For example, the third function

Surface normal vector that minimizes

Can be obtained. At this time, in order to robustly estimate the surface normal vector with respect to shadow or saturation,

Can be introduced. At this time,

The

when

Can be satisfied. The modified third function incorporating such a weight

. ≪ / RTI >

Finally, the surface normal vector obtaining unit 15 can obtain the surface normal vector in a closed-form according to Equation (5) below.

&Quot; (5) "

은 해당 좌표에 대한 표면 알베도(surface albedo)를 나타낸다.At this time

Represents the surface albedo for that coordinate.

Next, the three-dimensional mesh structure obtaining unit 16 may obtain the three-dimensional mesh structure for the target object 21 using the plurality of second three-dimensional point coordinates and the corresponding plurality of surface normal vectors. At this time, the three-dimensional mesh structure obtaining unit 16 uses the plurality of oriented points 600 that directly assign the surface normal vectors corresponding to the plurality of second three-dimensional point coordinates to the three-dimensional mesh structure 700 ) (See Figures 6 and 7).

The 3D mesh structure acquisition unit 16 can acquire a three-dimensional mesh structure using a PSR (Poisson Surface Reconstruction) method for a plurality of orientation points (M. Kazhdan, M. Bolitho, and H. Hoppe. In Proceedings of the fourth Eurographics symposium on Geometry processing, 2006.).

It is to be understood that both the foregoing general description and the following detailed description of the present invention are illustrative and explanatory only and are intended to be illustrative of the invention and are not to be construed as limiting the scope of the invention as defined by the appended claims. It is not. Therefore, those skilled in the art will appreciate that various modifications and equivalent embodiments are possible without departing from the scope of the present invention. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

10: 3D image restoration device
11: Expanded phase acquiring unit
12: second three-dimensional point coordinate acquiring unit
13: Intensity vector acquiring unit
14: Incident light direction matrix acquisition unit
15: Surface normal vector obtaining unit
16: Three-dimensional mesh structure acquiring unit
21: Target object
P1: Projector
C11, C12: Camera

Claims (17)

An apparatus for reconstructing a three-dimensional image using a plurality of first three-dimensional point coordinates obtained by photographing a target object at a plurality of points of view using a stereo scanning system including a first point-of-view apparatus and a second point-
Acquiring, for each of the plurality of viewpoints, a first unwrapped phase corresponding to the first viewpoint device and a second expanded phase corresponding to the second viewpoint device using the plurality of first three-dimensional point coordinates An expanded phase acquisition unit
A second three-dimensional point coordinate obtaining unit that obtains a plurality of second three-dimensional point coordinates that minimize a first function that is a difference between the first expanded phase and the second expanded phase with respect to the plurality of view points;
An intensity vector obtaining unit for obtaining a plurality of intensity vectors for each of the plurality of second three-dimensional point coordinates;
An incident light direction matrix obtaining unit for obtaining a plurality of incident light direction matrices for each of the plurality of second three-dimensional point coordinates;
A surface normal vector obtaining unit that obtains a plurality of surface normal vectors for each of the plurality of second three-dimensional point coordinates corresponding to the plurality of intensity vectors and the corresponding plurality of incident light direction matrices; And
And a three-dimensional mesh structure obtaining unit for obtaining a three-dimensional mesh structure for the target object using the plurality of second three-dimensional point coordinates and the corresponding plurality of surface normal vectors
3D image restoration device. delete The method according to claim 1,
Wherein the first viewpoint device is a first camera,
Wherein the second viewpoint device is a second camera or projector,
3D image restoration device. The method according to claim 1,
The expanded phase acquiring unit
Acquiring the first expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the first point-of-view device to the plurality of first three-dimensional point coordinates for each point in time,
Acquiring the second expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the second point-of-view device to the plurality of first three-dimensional point coordinates for each point in time,
3D image restoration device. The method according to claim 1,
The second three-dimensional point coordinate obtaining unit
Obtaining a plurality of second three-dimensional point coordinates by obtaining a displacement value of a three-dimensional coordinate that minimizes a second function obtained by linearizing the first function;
3D image restoration device. The method according to claim 1,
Wherein the intensity vector obtaining unit obtains the intensity vector of each of the plurality of second three-
Acquiring an intensity vector by interpolating intensity of corresponding two-dimensional point coordinates of the plurality of points of view;
3D image restoration device. The method according to claim 1,
Each of the plurality of incident light direction matrices
And a plurality of incident light direction vectors from a light source center of each of the plurality of viewpoints to a corresponding second three-
3D image restoration device. The method according to claim 1,
Wherein the surface normal vector obtaining unit obtains, for each of the plurality of second three-dimensional point coordinates,
Obtaining a surface normal vector that minimizes a third function which is a difference between a matrix vector of an intensity vector and a surface normal vector and an incident light direction matrix,
3D image restoration device. The method according to claim 1,
The three-dimensional mesh structure obtaining unit
Acquiring the three-dimensional mesh structure using a plurality of oriented points directly assigning surface normal vectors corresponding to the plurality of second three-dimensional point coordinates,
3D image restoration device. A method for reconstructing a three-dimensional image using a plurality of first three-dimensional point coordinates obtained by photographing a target object at a plurality of points of view using a stereo scanning system including a first point-of-view device and a second point-
Acquiring, for each of the plurality of viewpoints, a first unwrapped phase corresponding to the first viewpoint device and a second expanded phase corresponding to the second viewpoint device using the plurality of first three-dimensional point coordinates ;
Obtaining a plurality of second three-dimensional point coordinates such that a first function which is a difference between the first expanded phase and the second expanded phase is minimized for the plurality of time points;
Obtaining a plurality of intensity vectors for each of the plurality of second three-dimensional point coordinates;
Obtaining a plurality of incident light direction matrices for each of the plurality of second three-dimensional point coordinates;
Obtaining a plurality of surface normal vectors for each corresponding second plurality of three-dimensional point coordinates using the plurality of intensity vectors and the corresponding plurality of incident light direction matrices; And
Acquiring a three-dimensional mesh structure for the target object using the plurality of second three-dimensional point coordinates and the corresponding plurality of surface normal vectors
3D image reconstruction method. delete 11. The method of claim 10,
In obtaining the first and second expanded phases,
Acquiring the first expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the first point-of-view device to the plurality of first three-dimensional point coordinates for each point in time,
Acquiring the second expanded phase from the two-dimensional point coordinates obtained by applying the camera parameters of the second point-of-view device to the plurality of first three-dimensional point coordinates for each point in time,
3D image reconstruction method. 11. The method of claim 10,
In obtaining the plurality of second three-dimensional point coordinates,
Obtaining a plurality of second three-dimensional point coordinates by obtaining a displacement value of a three-dimensional coordinate that minimizes a second function that linearizes the first function,
3D image reconstruction method. 11. The method of claim 10,
Wherein the obtaining of the plurality of intensity vectors comprises: for each of the plurality of second three-dimensional point coordinates,
Obtaining an intensity vector by interpolating intensity of corresponding two-dimensional point coordinates of the plurality of points of view,
3D image reconstruction method. 11. The method of claim 10,
Each of the plurality of incident light direction matrices
And a plurality of incident light direction vectors from a light source center of each of the plurality of viewpoints to a corresponding second three-
3D image reconstruction method. 11. The method of claim 10,
Wherein the obtaining of the plurality of surface normal vectors comprises: for each of the plurality of second three-dimensional point coordinates,
Obtaining a surface normal vector that minimizes a third function, which is a difference of a matrix multiplication of an intensity vector, a surface normal vector and an incident light direction matrix,
3D image reconstruction method. 11. The method of claim 10,
In obtaining the three-dimensional mesh structure,
Acquiring the three-dimensional mesh structure using a plurality of directivity points to which surface normal vectors corresponding to the plurality of second three-
3D image reconstruction method.

Images