KR101980864B1 - Watermark inserting apparatus, watermark detecting apparatus and the method thereof - Google Patents

Abstract

A watermark inserting apparatus, a watermark detecting apparatus and a method thereof are disclosed. The watermark embedding method includes aligning the 3D model in the printing direction based on the stacking direction of the stacked 3D model and setting the watermark to be inserted in the 3D model aligned so as not to be related to the printing direction in a direction orthogonal to the printing direction And inserting a watermark of the pattern.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a watermark embedding apparatus, a watermark embedding apparatus, a watermark embedding apparatus, a watermark embedding apparatus,

The present invention relates to a watermark inserting apparatus, a watermark detecting apparatus and a method thereof, and more particularly to a watermark inserting apparatus, a watermark detecting apparatus and a watermark embedding apparatus relating to a blind mesh watermarking technique robust to three-dimensional (3D) It is about the method.

As the technology of layered manufacturing, which is also called 3D printing, has been developed, the manufacturing technology has been applied in various industrial fields. As the demand for 3D models increases, various sharing services appear on the Web. However, as shown in the precedent in the MP3 and video market, the increase in the model sharing environment will lead to copyright problems Lt; / RTI >

The 3D model in the 3D printing environment has a special characteristic that it is converted not only into the digital domain but also into the analog form through the printing and scanning process. Therefore, existing copyright protection technologies such as encryption and DRM can not be applied to 3D models actually printed, and copyright protection technology suited for such special situations is required.

3D models obtained with 3D scanners can be redistributed through the Web or replicated to offline models through 3D printers. 3D watermarking is the only protection technique that can display the copyright by displaying distorted shapes on the surface of the model.

Since existing 3D mesh watermarking is designed for various applications on-line, there is a problem that it can not operate in response to the distortion generated in the 3D print-scan process. Distortion in 3D printing environment can be roughly divided into three types. First, as shown in FIG. 1, during the lamination process, stepwise lamination distortion occurs, digital information including the file and the coordinate system information is lost in the printing process, and finally, various cutting and surface damage occur in the scanning process do.

The non-blind method of existing 3D mesh watermarking detects the watermark by referring to the information of the original model in the watermark detection, so that the watermark detection performance is very excellent. However, in order to detect the watermark, It is a very inefficient technology because it requires building a database of copyright protection models.

Therefore, there is a need for a technique that can detect a watermark without reference to the original model.

The present disclosure is directed to solving the above problems and an object of the present disclosure is to provide a three-dimensional print-scan robust blind mesh watermarking blind mesh watermarking method capable of detecting a watermark without reference to an original model using a print axis estimation technique And an apparatus.

According to another aspect of the present invention, there is provided a method of inserting a watermark, the method comprising: aligning the 3D model in a printing direction based on a stacking direction of a 3D model formed by stacking; And inserting a predetermined pattern of watermarks in the aligned 3D model in a direction orthogonal to the printing direction so as not to be associated with the printing direction.

The inserting step may convert the coordinate system for the aligned 3D model into a cylindrical coordinate system, and insert the watermark of the predetermined pattern into the radial component of the 3D model converted into the cylindrical coordinate system.

The inserting step may convert the watermark of the predetermined pattern into a spread spectrum signal and insert the watermark of the predetermined pattern into a sinusoidal signal formed by the radial component.

Meanwhile, the watermark embedding method may further include performing uniform remeshing on the 3D model to uniformize the vertex distribution of the 3D model.

According to another aspect of the present invention, there is provided a method for detecting a watermark according to an embodiment of the present invention. The watermark detection method comprises: 3D scanning a 3D model formed by stacking to obtain a digital pattern; Detecting the periodicity of the laminated noise pattern based on the set method, and identifying the printing direction based on the periodicity of the detected laminated noise pattern.

The preset method generates a plurality of projected images rotated in units of the predetermined angle, detects a projected image having a periodicity of the laminated noise pattern among the generated plurality of projected images, and determines the periodicity of the laminated noise pattern A method of detecting a periodicity of a laminated noise pattern by generating a plurality of surface images rotated by the predetermined angle unit or by detecting a surface image having a periodicity of a laminated noise pattern among the plurality of generated surface images, Lt; / RTI >

The step of detecting the periodicity may detect periodicity when a peak value larger than a preset threshold value is detected at a predetermined cycle through frequency analysis.

On the other hand, the watermark detection method may further include aligning the digital version in the identified printing direction and detecting watermarks embedded in the 3D model based on the aligned digital version.

The step of detecting the watermark may convert the coordinate system for the aligned digital bone into a cylindrical coordinate system and detect the inserted watermark from the radius component of the digital bone converted into the cylindrical coordinate system.

The watermark detection method may further include resampling the vertex coordinates of the 3D model at regular intervals.

The step of detecting the watermark may detect the inserted watermark using a frequency analysis algorithm and a correlation technique for the aligned digital pattern when a spread spectrum watermark is inserted into the 3D model .

According to another aspect of the present invention, there is provided an apparatus for inserting a watermark according to an embodiment of the present invention. The watermark embedding apparatus includes an alignment unit for aligning the 3D model in a printing direction based on a stacking direction of a 3D model, And inserting a predetermined pattern of watermarks into the aligned 3D model in a direction orthogonal to the printing direction so as not to be related to the direction.

The processor may convert the coordinate system for the aligned 3D model into a cylindrical coordinate system, and insert the watermark of the predetermined pattern into the radius component of the 3D model converted into the cylindrical coordinate system.

The processor may convert the watermark of the predetermined pattern into a spread spectrum signal and insert the watermark of the predetermined pattern into the sinusoidal signal formed by the radial component.

In addition, the processor may perform uniform remeshing on the 3D model to uniformize the vertex distribution of the 3D model.

According to an aspect of the present invention, there is provided a watermark detection apparatus comprising: a 3D scanner for scanning a 3D model of a stacked 3D model to obtain a digital pattern; and a controller for rotating the obtained digital pattern by a predetermined angle unit And a processor for detecting the periodicity of the laminated noise pattern based on the predetermined method, and the processor identifies the printing direction based on the periodicity of the detected laminated noise pattern.

The predetermined method may include generating a plurality of projection images rotated in units of the predetermined angle and detecting a periodic feature of the laminated noise pattern by detecting a projection image having a periodicity of the laminated noise pattern among the plurality of generated projection images Or a method of detecting a periodicity of a laminated noise pattern by detecting a surface image having a periodicity of the laminated noise pattern among the plurality of generated surface images by generating a plurality of surface images rotated by the predetermined angle unit have.

The processor can detect that the peak value is greater than a preset threshold value at a certain period of time through the frequency analysis to have periodicity.

The processor may also align the digital bones in the identified orientation and detect watermarks embedded in the 3D model based on the aligned digital bones.

The processor may convert the coordinate system for the aligned digital bone into a cylindrical coordinate system and detect the inserted watermark from the radial component of the digital bone converted to the cylindrical coordinate system.

The processor may resample the vertex coordinates of the 3D model at regular intervals.

In addition, when the spread spectrum watermark is inserted into the 3D model, the processor can detect the embedded watermark using a frequency analysis algorithm and a correlation technique for the aligned digital pattern.

As described above, according to various embodiments of the present disclosure, a watermark embedding apparatus and method can insert a watermark which is robust and is not visually recognized, regardless of the reproduction of the 3D model.

And, the watermark detection apparatus and method can easily estimate the printing direction of the 3D model based on the stack manufacturing method.

Also, the watermark detection apparatus and method can detect a watermark without reference to an original model by using a print axis estimation technique, and do not require any separate hardware or special material, and can be applied to all of a printer based on a lamination manufacturing method .

In addition, the watermark inserting apparatus, the watermark detecting apparatus and the method can strengthen the copyright protection of the 3D model contents by providing a robust 3D mesh watermarking technique both in the 3D printing environment and the online environment.

Further, the watermark inserting apparatus, the watermark detecting apparatus, and the method can protect the rights of the copyright holder by enhancing the copyright protection of the 3D model contents, thereby making it possible to produce higher quality contents.

FIG. 1 is an exemplary view showing an original mesh model, a digital slicing process for a manufacturing process, and a stacking noise generated in a stacking process.
2 is a diagram illustrating a watermarking technique in accordance with one embodiment of the present disclosure;
3 is a block diagram of a watermark embedding device according to an embodiment of the present disclosure;
4 is a block diagram of a watermark embedding device processor in accordance with one embodiment of the present disclosure.
5 is a block diagram of a watermark detection apparatus according to an embodiment of the present disclosure;
6 is a block diagram of a watermark detection device processor in accordance with an embodiment of the present disclosure;
FIGS. 7A and 11B are diagrams for explaining a process of detecting a printing direction according to an embodiment of the present disclosure.
12 and 13 are views for explaining a watermark embedding process according to an embodiment of the present disclosure.
14 is a flowchart of a watermark embedding method according to an embodiment of the present disclosure;
15 is a flowchart of a watermark detection method according to an embodiment of the present disclosure.
16 is a flowchart of a watermark detection method according to another embodiment of the present disclosure.

Various embodiments will now be described in detail with reference to the accompanying drawings. The embodiments described herein can be variously modified. Specific embodiments are described in the drawings and may be described in detail in the detailed description. It should be understood, however, that the specific embodiments disclosed in the accompanying drawings are intended only to facilitate understanding of various embodiments. Accordingly, it is to be understood that the technical idea is not limited by the specific embodiments disclosed in the accompanying drawings, but includes all equivalents or alternatives falling within the spirit and scope of the invention.

Terms including ordinals, such as first, second, etc., may be used to describe various elements, but such elements are not limited to the above terms. The above terms are used only for the purpose of distinguishing one component from another.

In this specification, the terms " comprises " or " having ", and the like, are intended to specify the presence of stated features, integers, steps, operations, elements, parts, or combinations thereof, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof. It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

In the meantime, " module " or " part " for components used in the present specification performs at least one function or operation. Also, " module " or " part " may perform functions or operations by hardware, software, or a combination of hardware and software. Also, a plurality of " modules " or a plurality of " parts ", other than a " module " or " part ", to be performed in a specific hardware or performed in at least one processor may be integrated into at least one module. The singular expressions include plural expressions unless the context clearly dictates otherwise.

In addition, in the description of the present invention, when it is judged that the detailed description of known functions or constructions related thereto may unnecessarily obscure the gist of the present invention, the detailed description thereof will be abbreviated or omitted.

2 is a diagram illustrating a watermarking technique in accordance with one embodiment of the present disclosure;

Referring to FIG. 2, the watermarking technique includes the steps of inserting a watermark into a 3D model, estimating a printing direction or printing reference axis for a digital version of the 3D model scanned by the 3D scanner, And detecting a watermark.

The insertion of the watermark for the 3D model can be performed by the following process. In one embodiment, the watermark embedding device aligns the 3D model in the printing direction based on the stacking direction of the stacked 3D models. The watermark embedding device inserts watermarks of a predetermined pattern in a direction orthogonal to the printing direction in the 3D model aligned so that the inserted watermark is not related to the printing direction. In some cases, the watermark embedding device can simultaneously perform 3D printing and watermark embedding. The watermark embedding device determines the axial direction for printing the 3D model (S10), and prints the 3D model while inserting watermarks of a predetermined pattern in a direction orthogonal to the printing direction (S12).

A watermarked model in which a watermark is inserted is generated by inserting a watermark into the 3D model, and a watermark can be inserted based on the printing direction of the 3D printer or the printing reference axis. For example, A spread spectrum watermark may be inserted in a direction orthogonal to the direction of the spread spectrum. The watermark can be generated by a separate watermark key. And, the watermark may include different patterns that can be distinguished by individual or group.

Then, the watermark detection apparatus can detect the watermark embedded in the 3D model. The watermark detection apparatus performs a process of estimating a print axis and a process of detecting a watermark embedded in a 3D printed model.

The process of estimating the printing direction can be performed by the following process. In one embodiment, the watermark detection apparatus scans a 3D model printed by a 3D printer that wants to detect a watermark using a 3D scanner to obtain a digital version (S14). Then, the watermark detection apparatus can estimate the printing direction or printing reference axis of the 3D model using the obtained digital pattern (S16).

The process of estimating the printing direction can estimate the printing direction of the digital pattern obtained by the 3D scanning by analyzing the lamination noise appearing on the surface of the digital pattern.

The watermark detection apparatus can perform the watermark detection process after the estimation process of the printing direction. The watermark detection apparatus may perform blind watermark detection on the 3D model based on the aligned digital scan data after aligning the digital pattern in the estimated print direction (S18) (S20).

In order to detect a watermark pattern embedded in a model acquired in a three-dimensional printing environment, a watermark synchronization (watermark synchronization) system should be restored first. This disclosure makes use of laminated noise patterns appearing on the model surface (or object surface, digital bone surface) without additional information insertion or model changes for watermark sync (sync) recovery. Since the laminated noise pattern appears in a direction orthogonal to the printing direction, if the watermark pattern is inserted in a direction orthogonal to the printing direction, the coordinate noise of the inserted watermark pattern can be restored by analyzing the laminated noise pattern. Of course, the watermark in the present disclosure is not limited to being inserted only in the direction perpendicular to the printing direction, and any watermark embedding method and detection method that can be inserted based on the printing direction can be used.

The configuration of the watermark inserting apparatus and the watermark detecting apparatus will be described below.

3 is a block diagram of a watermark embedding device according to an embodiment of the present disclosure;

Referring to FIG. 3, the watermark embedding apparatus 100 includes an alignment unit 110 and a processor 120.

The alignment unit 110 aligns the 3D model in the printing direction based on the stacking direction of the stacked 3D models. The watermark described in this disclosure can be inserted in a direction perpendicular to the printing direction so as to have toughness to the lamination noise. Therefore, the watermark embedding apparatus 100 can identify the printing direction of the 3D model and align the 3D model so that the printing direction is perpendicular to the bottom surface.

The processor 120 inserts a watermark of a predetermined pattern in a direction orthogonal to the printing direction in the 3D model aligned so that the watermark to be inserted is not related to the printing direction. Generally, a 3D printer creates a 3D model in a stacking manner. Therefore, in the printed 3D model, noise is generated for each layer in a direction parallel to the printing direction, so that the inserted data may be lost. Therefore, the data inserted in the direction parallel to the printing direction may not be lost or discerned by copying of the 3D model or the like. On the other hand, the watermark pattern set in the direction orthogonal to the printing direction is not affected by the lamination noise generated in the horizontal direction, and can be identified without being lost in the copying of the 3D model or the like.

The processor 120 converts the coordinate system of the aligned 3D model into a cylindrical coordinate system, and inserts a watermark of a predetermined pattern into a radial component of the 3D model converted into the cylindrical coordinate system. If a spread spectrum watermark is inserted into the 3D model, the processor 120 converts the watermark of a predetermined pattern into a spread spectrum signal, synthesizes the watermark into a sinusoidal signal formed of a radial component, Can be inserted.

In addition, the processor 120 may perform uniform remeshing on the 3D model to uniformize the vertex distribution of the 3D model. The watermark detection method of the present disclosure is a technique related to blind watermarking without a reference 3D model. Generally, if the vertex information of the 3D model generated is not known because the vertex for 3D printing is not uniform, the watermark detection apparatus may be difficult to detect the inserted watermark. Therefore, the processor 120 of the watermark embedding apparatus can perform uniform reshaping to uniformize the vertex distribution of the 3D model so that the watermark detecting apparatus can detect the watermark even in the absence of information.

Up to now, a technique of inserting a watermark into a pre-created 3D model has been described. However, the watermark embedding apparatus 100 can simultaneously perform 3D printing and watermark embedding. For example, the watermark embedding apparatus 100 may be a 3D printer. The alignment unit 110 determines the printing direction, and the processor 120 can perform the printing process of the 3D model based on the determined printing direction and insert a predetermined watermark.

4 is a block diagram of a watermark embedding device processor in accordance with one embodiment of the present disclosure.

4, the processor 120 of the watermark embedding apparatus according to the present disclosure may include an execution unit 121, a dividing unit 122, a conversion unit 123, and an insertion unit 124. FIG.

The execution unit 121 may perform uniform remeshing on the 3D model in which the watermark is to be inserted. The dividing unit 122 may divide the 3D model aligned in the printing direction into a plurality of predetermined slices. The watermark embedding apparatus 100 may divide the 3D model into a plurality of slices and insert predetermined watermarks into each of the sliced slices. Therefore, when the watermark embedding apparatus 100 divides the 3D model into a plurality of slices to insert a watermark, the dividing section 122 can divide the 3D model into a plurality of slices. The dividing unit 122 may divide the 3D model aligned in the printing direction into slices based on the printing direction of the 3D printer.

The converting unit 123 may convert the coordinate system of the 3D model aligned in the printing direction into the cylindrical coordinate system. The inserting unit 124 may insert a predetermined watermark into the 3D model aligned in the printing direction. The inserting unit 124 may insert a watermark into the divided slices, and may specifically insert a watermark into a radial component of slices converted into a cylindrical coordinate system. For example, the watermark may be information capable of identifying the 3D model, and the identification information may include a type of printing apparatus, user information, an IP address, a MAC address, a usage history, specific information of the model, and the like.

In one embodiment, the execution unit 121, the partition unit 122, the conversion unit 123, and the insertion unit 124 may be implemented as a hardware module or a software module.

On the other hand, the watermark embedding apparatus 100 of the present disclosure may include all the operations and functions for the watermark detection process even if the description thereof is not described in FIG. 3 or FIG. That is, the watermark inserting apparatus and the watermark detecting apparatus may be physically implemented as a single apparatus. The specific procedure for inserting a watermark is described in detail below.

5 is a block diagram of a watermark detection apparatus according to an embodiment of the present disclosure;

Referring to FIG. 5, the watermark detection apparatus 200 includes a 3D scanner 210 and a processor 220.

The 3D scanner 210 scans the stacked 3D model to obtain a digital pattern. In some cases, the watermark detection apparatus 200 may detect an embedded watermark using a 3D digital pattern scanned in a separate 3D scanner 210. [

The processor 220 rotates the obtained digital pattern by a predetermined angle unit to generate a plurality of projection images. Then, the processor 220 detects a projection image having a periodicity of the laminated noise pattern among the plurality of generated projection images. Then, the processor 220 identifies the printing direction based on the laminated noise pattern included in the projected image having the periodicity. The processor 220 can detect that the peak value has a periodicity when a peak value larger than a preset threshold value is detected at a constant cycle through frequency analysis. The processor 220 can identify the printing direction (or reference axis direction) of the 3D model through the above-described process. Alternatively, the generating unit 221 may generate a plurality of surface images instead of a plurality of projection images to estimate the printing direction. The process of estimating the printing direction using a plurality of surface images is the same as the process of estimating the printing direction using a plurality of projection images, and thus a detailed description thereof will be omitted.

On the other hand, the printing direction may be estimated in a different manner. In one embodiment, the watermark detection apparatus 200 rotates the 3D model. Then, the watermark detection apparatus 200 performs a filtering process for detecting a specific signal (e.g., printing trace) on the surface of the rotating 3D model. The watermark detection apparatus 200 analyzes output values of the filtering process performed. The watermark detection apparatus 200 can identify the alignment direction of the model having the maximum value among the analyzed output values in the print direction.

Alternatively, the watermark detection apparatus 200 performs a digital slicing process on the 3D model to generate a digital slicing model. Then, the watermark detection apparatus 200 scans the 3D model. The watermark detection apparatus 200 analyzes the correlation between the generated digital slicing model and the scan model to identify the printing direction. That is, the watermark detection apparatus 200 can identify the alignment direction of the model having the maximum correlation in the print direction.

The processor 220 may align the digital bones in the identified printing direction. The processor 220 may then detect the watermark embedded in the 3D model based on the aligned digital pattern.

On the other hand, the inserted watermark can be inserted after being converted into a cylindrical coordinate system in the watermark embedding apparatus. The processor 220 may convert the coordinate system for the aligned digital bone into the cylindrical coordinate system and detect the inserted watermark from the radial component of the digital bone converted to the cylindrical coordinate system.

The watermark detection apparatus 200 of the present disclosure identifies a watermark embedded in a 3D model without a reference 3D model. Accordingly, the watermark detection apparatus 200 can resample the vertex coordinates of the 3D model at regular intervals, and detect the inserted watermark.

On the other hand, if a spread spectrum watermark is inserted into the 3D model, the processor 220 may detect the embedded watermark using a frequency analysis algorithm and correlation technique for the aligned digital pattern.

6 is a block diagram of a watermark detection device processor in accordance with an embodiment of the present disclosure;

Referring to FIG. 6, the processor 220 of the watermark detection apparatus may include a generating unit 221, an estimating unit 222, a dividing unit 223, a converting unit 224, and a detecting unit 225.

The generation unit 221 may generate a plurality of projection images by rotating the obtained digital version by a predetermined rotation angle unit.

The estimating unit 222 can estimate the printing direction of the digital copy based on the laminated noise pattern for the digital copy. The estimating unit 222 can estimate the printing direction of the digital pattern based on the laminated noise pattern for a plurality of projected images. For example, the estimation unit 222 may detect a projection image having a periodicity of the laminated noise pattern among the plurality of generated projection images. In one embodiment, the estimator 222 may determine that the peak value is greater than a predetermined threshold value at a predetermined period of time through frequency analysis. The estimating unit 222 can estimate the printing direction based on the laminated noise pattern included in the projected image having periodicity. Then, the estimating unit 222 can align the digital version to the printing direction based on the estimated printing direction. On the other hand, as described above, the processor 220 of the watermark detection apparatus may estimate the print direction and align the digital version using the surface image instead of the projected image.

In one embodiment, the watermark embedding device may divide the 3D model into a plurality of slices and insert predetermined watermarks into each of the sliced slices. When the watermark embedding device inserts watermarks into the divided slices, the dividing section 223 can divide the digital document aligned in the estimated printing direction into a plurality of predetermined slices.

The converting unit 224 may convert the coordinate system for the digital pattern aligned in the printing direction into a cylindrical coordinate system. The detection unit 225 can detect the watermark embedded in the 3D model based on the digital pattern aligned in the printing direction. The detecting unit 225 can detect the watermark embedded in the 3D model from the divided slices. The detection unit 225 may detect the watermark embedded in the 3D model from the radius component of the slices converted into the cylindrical coordinate system. In addition, when the spread spectrum watermark is inserted into the 3D model, the detection unit 225 can detect the watermark embedded in the 3D model using a frequency analysis algorithm and a correlation technique for the digital pattern aligned in the printing direction.

The generating unit 221, the estimating unit 222, the dividing unit 223, the converting unit 224, and the detecting unit 225 may be implemented as a hardware module or a software module.

As described above, the watermark detection apparatus 100 of the present disclosure may include all the operations and functions for the watermark embedding process. That is, the watermark inserting apparatus and the watermark detecting apparatus may be physically implemented as a single apparatus. The specific procedure for detecting a watermark is described in detail below.

First, a process of detecting the printing direction will be described.

FIGS. 7A and 11B are diagrams for explaining a process of detecting a printing direction according to an embodiment of the present disclosure. The printing direction of the 3D model means the axial direction. Therefore, detection of the printing direction means detection in the axial direction.

The watermarking technique according to the present disclosure can start from an intuitive and simple theorem that " 3D printing distortion in a rotating projection system can be interpreted as lines having periodicity ", and various signal processing techniques To provide a completely new method of input variable estimation.

The 3D printer sets input parameters such as a layer thickness, a fabrication direction, and a print size and a position for laminate manufacturing. Based on the input information, the digital 3D model can be divided into a thin layer shape in the horizontal direction as shown in FIG. 7A. The stair-step effect created in this process is a distortion that necessarily occurs in the lamination process.

The present disclosure analyzes a step-like distortion using a projection and rotation system (or method, apparatus), as shown in Figure 7B. That is, a projection system having a rotation space as shown in FIG. 7B can be defined to analyze the 3D model. The projection direction is the y-axis, and the space can rotate about the z-axis. An arbitrary plane (P 占 R3) in the system can be interpreted as a line at a particular angle of rotation. The above description can be proved through the characteristic of a plane having a normal vector after applying a rotation transformation to remove the θ value for an arbitrary vector v _s = (r, θ, φ) as shown in FIG. 7B .

That is, as shown in FIG. 8, a plurality of laminated layers may appear as a line at a specific rotation angle through rotation. In the present disclosure, a 3D model for detecting a watermark through such a laminated noise pattern The printing direction or printing reference axis of the printed 3D printer can be estimated.

An algorithm for estimating the printing direction of the layered 3D model (layering artifact) using the projection system can be designed. First, the digitized 3D surface data can be obtained from the 3D scanner to analyze the surface of the 3D model. As a pre-processing step, the center of the scanned data M may be located at the center of the mass. The scanned data M may be a digital version. The reference axis can be estimated through the process of rotation and projection, amplification, frequency analysis and axial estimation.

First, a rotation and projection process can be performed. The scanned data M of the 3D model can be rotated around a z axis within a predetermined angle unit within a range of 0 to 180 degrees. Then, the rotated model M &_amp;_thetas; can also be generated as shown in Fig. 10 (b). M _? Is projected in the xz plane of FIG. 7 (b), and the thickness map P _? Is obtained.

Fig. 9 is a view for explaining the process of calculating the thickness. The thickness of M _? Can be calculated by subtracting the depth map of the back-faces from the depth map of the front-faces. The front depth map P _f can be computed by the z-buffer algorithm. Invalidate the z-axis of _θ M, the back may be rendered in order to obtain a depth map, P _b. Then, the thickness map P _? Can be obtained by subtracting Pb from Pf.

Next, an amplification process can be performed. Various filtering techniques can be applied to emphasize periodic characteristics in the layered 3D model. First, a high pass filter may be applied to the thickness map P _&thetas; to reveal the traces of the stacked 3D model. Alternatively, a similar result may be obtained by a second derivative or a Laplace operator. Histogram thresholding has been found to have the effect of removing some meaningless components at the filtered P _[theta] . Two opposite motions of histogram stretching and histogram thresholding have been used in this disclosure. The histogram thresholding can be defined by the following equation.

(1)

(One)

Here, τ denotes a threshold, and (x, y) denotes an x-th row and a y-th column. The histogram stretching and histogram threshold painter is sequentially applied to the P _θ, it is amplified P _'θ can be obtained.

Next, a frequency analysis process can be performed. If the input scan data is a stacked fabricated model, the 3D model includes a special strong periodicity. The stacked 3D model can be expressed in a periodic line shape through analysis using 2D DFT (Discrete Fourier Transform).

First, a Hanning window can be applied to P '_&thetas; to prevent spectral leakage. Then, a 2D FFT (Fast Fourier Transform) is performed, and the magnitude can be calculated as follows.

(2)

(2)

Here, u and v are frequency variables for x and y, and F ( . ) Is a 2D FFT. H () is a Butterworth high-pass filter that removes important low-frequency components. F _θ (u, v) is the polar coordinate system F _θ (r, φ) satisfying

(3)

(3)

Where r and φ are the radius and angular coordinates.

A threshold-based peak detector is applied to search for a local maximum value (n peaks greater than the local average intensity). Then, a set of locations P ^[theta] ₀ for the candidate peaks is obtained. Peak ρ _θ can be chosen from the biggest strength that satisfies the following:

(4)

(4)

θ is between 0 and 180 degrees, and P ^θ ₀ is a set of positions for the candidate peak. pos represents the position value of the peak ρ _θ , and mag represents the magnitude value. And pos (ρ _θ ) represents the maximum peak position from the rotated model M _θ , and mag (ρ _θ ) represents the intensity of the laminated 3D model.

FIG. 10 shows a diagram for each process of input variable estimation. Figure 10 (a) has been shown that the surface information data of the input model, Figure 10 (b), the projection was a thickness map P _θ of the image shown, Fig. 10 (c) there was a high-pass filtered P _θ shown, Fig. 10 (d) shows the final result after filtering.

를 위한 F_θ(r, φ)의 샘플 플롯이 도시되어 있다. 주기적 에너지들은 국부 지점으로 수렴하고, F_θ(r, φ)의 전역 최고 지점(global maximum point)은 추정될 수 있다. 즉, 축 추정 방법은 주파수 분석을 통해 일정한 주기로 기 설정된 임계값보다 큰 피크값이 검출되는 경우 주기성을 가지는 것으로 검출할 수 있다. Next, the orientation of the stacked 3D model can be estimated. 11A and 11B show estimated parameters

A sample plot of F _[theta] (r, [phi]) is shown. Periodic energies converge to local points, and the global maximum point of F _θ (r, φ) can be estimated. That is, the axis estimation method can detect that a periodicity is detected when a peak value larger than a preset threshold value is detected at a constant cycle through frequency analysis.

을 추정한다. 피크가 탐지되지 않는 경우, 워터마크 검출 장치는 3D 모델이 적층 방식으로 제조되지 않은 것으로 판단할 수 있다.Specifically, the process of estimating the printing direction or the printing reference axis of the 3D printer that has printed the 3D model through the periodicity analysis in the projection data will be described. First, the watermark detection apparatus calculates projection data M _? (M, n ) on for converting the obtained 2D Fourier (Fourier) map intensity (magnitude) to perform the transformation F _θ (u, v), and the intensity maps F _θ (u, v) in the polar coordinate system F _θ (r, φ) The local peaks (or candidate peaks) are selected through a peak-point-based peak detector. Then, the watermark detection apparatus selects the peak ρ _θ having the largest value among the regional peaks, and then, through the assignment, the 3D model is printed in the stacking direction (or printing direction)

. If no peak is detected, the watermark detection apparatus can determine that the 3D model is not manufactured in a laminated manner.

In order to reduce the axis estimation time, the axis estimation method described above may be applied by sampling only a part of the scanned data.

Next, the watermark embedding process will be described.

12 and 13 are views for explaining a watermark embedding process according to an embodiment of the present disclosure.

The watermark embedding method according to an embodiment of the present disclosure can perform uniform remeshing of an input 3D model when a 3D model to insert a watermark is input. Then, the 3D model that has been reformed can be aligned and rotated in the printing direction of the 3D printer. As shown in FIG. 12 (a), uniform re-meshing means a process of resampling the input 3D model surface, that is, the mesh model. Uniform re-meshing can evenly change the vertex distribution of the 3D model surface.

The right half shown in FIG. 12 (a) is a resampled mesh model for convenience of explanation. The mesh model actually input may be represented by a dot and a line as shown in the left half of FIG. 12 (a) The remeshed 3D model can resample the mesh model by uniformly changing the vertex distribution of the mesh model surface as shown in FIG. 12 (b). By resampling the mesh model and inserting the watermark, the watermark detection apparatus can detect the watermark without information on the reference 3D model.

The watermark embedding device can arrange the uniformly resized 3D model by rotating it in the printed direction. Alternatively, the watermark embedding device may include a 3D printer. Thus, the watermark embedding device can align the 3D model in the printing direction and print the 3D model in the aligned direction. The 3D model shown in Fig. 12 (b) can be rotated and aligned in the printing direction as shown in Fig. 12 (c).

When the 3D model is aligned with the printing direction of the 3D printer, the 3D model aligned in the printing direction can be divided into a plurality of predetermined slices. The watermark embedding device can divide the 3D model into slices based on the printing direction of the 3D printer or aligned with respect to the printing reference axis. For example, the 3D model shown in FIG. 12C may be divided into a predetermined number of slices in a stacked state in the vertical direction. The center of each of the slices thus divided can be shifted about the printing reference axis.

When the 3D model is divided into slices, the coordinate system information for the 3D model aligned in the printing direction is converted into a cylindrical coordinate system, and a watermark is inserted into the radius component (rho) of the slices converted into the cylindrical coordinate system . The watermark pattern can be generated using a watermark key. On the other hand, the watermark embedding apparatus can insert a watermark using x, y coordinates in a Cartesian coordinate system (orthogonal coordinate system) without converting it into a cylindrical coordinate system. In addition, the watermark embedding device may embed a watermark in all slices or may insert a watermark in at least one slice in a certain area.

The watermark embedding device can convert a 3D model into a cylindrical coordinate system to insert a watermark into a radius component of slices, and insert a watermark into a radius component of a 3D model vertex of a cylindrical coordinate system. In addition, the watermark embedding device can insert the spread spectrum watermark into the radial component of the slices.

The watermark insertion can be described by the following equation.

(5)

(5)

Here, i is an index of a vertex and i is a function for visual masking. Then, Ψ can be synthesized into the form of a sine wave signal in the frequency band [f _s +1, f _s + l _w ] as the spread spectrum signal in the following equation.

(6)

(6)

Here, wm is the watermark, fs is the minimum frequency band, Φ _{i, l} is the fine phase parameters, (m + f _s) is a frequency.

Conceptually, FIG. 13 (a) shows one slice (or one layer) of a 3D model printed by a 3D printer. The watermark embedding device can insert a watermark of a predetermined pattern generated by the watermark key into one slice (or one layer). 13 (b) shows one slice (or one layer) into which a watermark is inserted. As shown in FIG. 13 (b), the watermark embedded in the radial component of the 3D model is not visible to the naked eye and can have robustness to laminated noise. Therefore, even if the 3D model is copied, the watermark embedded in the radial component can be maintained.

The watermark embedding method according to the present disclosure is not limited to embedding a watermark in a radial component of a cylindrical coordinate system, and various methods of embedding a watermark can be applied. The watermark embedding method according to the present disclosure can arrange the 3D model in the printing direction of the 3D printer and insert a watermark to be inserted into the 3D model aligned in the printing direction.

As described above, the watermark embedding method according to the present disclosure can apply various watermark embedding schemes based on the printing direction or the printing reference axis with respect to the 3D model aligned in the printing direction of the 3D printer.

The watermark embedding method has been described so far. Hereinafter, a watermark detection method will be described.

The watermark detection method according to the present disclosure may require a printed version of the 3D mesh model protected by the watermark embedding method in order to detect the watermark. Basically, the watermark detection method can be performed in the reverse order of the watermark embedding method.

2, the watermark detection method according to an embodiment of the present disclosure can be performed in the order of 3D model scanning, printing direction (reference axis) identification, coordinate system conversion, watermark detection, and detected response signal analysis have.

The 3D scanner scans the surface information of the 3D printed by the 3D printer to obtain a digital version of the 3D model. That is, the 3D scanner digitizes the surface information of the 3D model printed by the 3D printer. For example, the 3D scanner acquires a digital version of the 3D model as shown in FIG. 10 (a) by scanning the surface information of the 3D model by 3D scanning. As described above, the 3D scanner may be a part of the configuration included in the watermark detection apparatus. Alternatively, the 3D scanner may be a separate device separate from the watermark detection apparatus, and the watermark detection apparatus may transmit or receive the scanned data from the 3D scanner to perform the detection process.

When the digital version of the 3D model is acquired, the watermark detection apparatus rotates the obtained digital version by a predetermined rotation angle unit to generate a plurality of projection images for the digital version and calculates a thickness map of the projection image. For example, the watermark detection apparatus generates a projection image for every rotation [theta] along any particular axis of the digital bone shown in Fig. 10 (a) and then generates a thickness map P _[theta ] of the projection image. Here, the thickness map of the projected image can be generated or obtained by projecting the thickness information of the 3D model (or digital pattern) on the two-dimensional plane.

The watermark detection apparatus estimates the print direction of the digital version based on a laminated noise pattern for the projected images when a thickness map of the projected images is generated. For example, as shown in Figs. 10 (c) and 10 (d), the watermark detection apparatus obtains P _? Emphasizing a stepped effect through a filtering process on the thickness map of all projection images. Then, the watermark detection apparatus obtains the position of the peak on the analysis data through the periodicity analysis of the thickness map P _? Of the projection data, and estimates the printing direction or printing reference axis of the 3D model as the input variable through the position of the obtained peak. Here, the peak may correspond to? And? In the spherical coordinate system for P _? Emphasizing the step effect.

을 추정한다. 피크가 탐지되지 않는 경우, 워터마크 검출 장치는 3D 모델이 적층 방식으로 제조되지 않은 것으로 판단할 수 있다.Specifically, the process of estimating the printing direction or the printing reference axis of the 3D printer that has printed the 3D model through the periodicity analysis in the projection data will be described. First, the watermark detection apparatus calculates projection data M _? (M, n ) on for converting the obtained 2D Fourier (Fourier) map intensity (magnitude) to perform the transformation F _θ (u, v), and the intensity maps F _θ (u, v) in the polar coordinate system F _θ (r, φ) The local peaks (or candidate peaks) are selected through a peak-point-based peak detector. Then, the watermark detection apparatus selects the peak ρ _θ having the largest value among the regional peaks, and then, through the assignment, the 3D model is printed in the stacking direction (or printing direction)

. If no peak is detected, the watermark detection apparatus can determine that the 3D model is not manufactured in a laminated manner.

If the printing direction of the digital version is estimated, the watermark detection device can align the digital version in the estimated printing direction. The watermark detection apparatus can align the digital version in the direction in which the watermark is inserted by aligning the axis of the digital version with the printing reference axis or the printing direction of the 3D printer.

The watermark detection apparatus can divide a digital pattern aligned in the printing direction into a plurality of preset slices. The watermark detection apparatus may skip the process of dividing the watermark into a plurality of slices if it knows information that the watermark is not sliced and inserted. The watermark detection apparatus can divide the 3D model into slices based on the printing direction of the 3D printer or aligned with respect to the printing reference axis.

When the digital pattern is divided into slices, the watermark detection apparatus converts the coordinate system for the digital pattern aligned in the printing direction into a cylindrical coordinate system, and extracts the 3D model from the radial component (rho) of the slices converted into the cylindrical coordinate system Can be detected.

The watermark detection apparatus may re-sample the coordinate information of the vertex for the digital pattern into a lattice pattern at regular intervals. At this time, the coordinate information used may be the z-axis and the phi component of the cylindrical coordinate system.

The watermark detection apparatus can detect a watermark embedded in a 3D model using a frequency analysis algorithm and a correlation technique for a digital pattern aligned in the printing direction when a spread spectrum watermark is inserted into the 3D model.

The detection of the embedded watermark using the correlation technique can be performed by the following process.

First, a 1D FFT is applied to the phi axis and the intensity is calculated.

(7)

(7)

Where F is a 1D FFT and H is a high pass filter.

Next, the intensity of the z-axis is averaged as follows.

(8)

(8)

Next, the subvector M * is selected at M.

(9)

(9)

Here, the 1≤m≤l _w, l _w is the length of the watermark. And fs is a start frequency.

Next, the detection response r is calculated.

(10)

(10)

Here, w is a watermark pattern.

As described above, the watermarking method according to the embodiment of the present disclosure can have the following advantages.

First, it is possible to restore the sync information of the watermark using only patterns generated in the 3D printing process without adding additional reference axis restoration information. Since all of the current industrial / popular laminate manufacturing methods have laminated noise, this disclosure can be applied to most existing models and additional patterns required for restoring watermark reference axes are not needed, Can have a great advantage.

Second, since the watermark is inserted in a direction orthogonal to the printing axis, the digital slicing process occurring before the stacking process can be robust. In addition, since the watermark pattern is not separated even in the physical lamination process, it can be robust against off-line printing noise.

Heretofore, various embodiments for inserting a watermark and detecting an inserted watermark have been described. A flowchart of the watermark embedding method and the watermark embedding method will be described below.

14 is a flowchart of a watermark embedding method according to an embodiment of the present disclosure;

Referring to Fig. 14, the watermark embedding apparatus aligns the 3D model in the printing direction based on the stacking direction of the stacked 3D models (S1410). The printing direction of the 3D model means the direction in which the layers are stacked, which means the reference axis direction.

The watermark embedding device inserts watermarks of a predetermined pattern in a direction orthogonal to the printing direction in the 3D model aligned so that the watermark to be inserted is not related to the printing direction (S1420). The watermark embedding apparatus can convert a coordinate system for the aligned 3D model into a cylindrical coordinate system, and insert a watermark of a predetermined pattern into a radial component of the 3D model converted into the cylindrical coordinate system. On the other hand, the watermark embedding apparatus may insert a watermark into the x and y components on a Cartesian coordinate system (orthogonal coordinate system) without transforming the cylindrical coordinate system.

In addition, the watermark embedding apparatus can perform uniform resizing processing for the 3D model to uniformize the vertex distribution of the 3D model. In one embodiment, the watermark embedding device may convert a watermark of a predetermined pattern into a spread spectrum signal, and insert the watermark into a sinusoidal signal formed of a radial component. On the other hand, the watermark embedding apparatus may include a 3D printer, and may simultaneously perform watermark embedding along with printing of the 3D model.

15 is a flowchart of a watermark detection method according to an embodiment of the present disclosure.

The watermark detection apparatus can estimate the printing direction (or reference axis direction) and detect the inserted watermark.

Referring to Fig. 15, the watermark detection apparatus acquires a digital pattern by scanning the 3D model of the laminated 3D model (S1510). The watermark detection apparatus can scan a 3D model including a 3D scanner. In some cases, the 3D scanner may be an independent device, and the watermark detection apparatus may receive or transmit digital watermark data from the 3D scanner.

The watermark detection apparatus detects the periodicity of the laminated noise pattern based on the predetermined scheme by rotating the obtained digital pattern by a predetermined angle unit (S1520). Generally, since the 3D model is manufactured in a laminated manner, the obtained digital pattern includes a laminated noise pattern. Thus, if the watermark detection apparatus rotates the digital pattern by a certain angle unit to generate a plurality of projection images, the watermark detection apparatus can identify the laminated noise pattern from the projected image. For example, the watermark detection apparatus can determine that the peak value is greater than a preset threshold value at regular intervals through frequency analysis, and that the watermark detection apparatus has a periodicity.

On the other hand, the watermark detection apparatus may sample and analyze the projection image data to reduce the time of the axis estimation process, and may determine the periodicity as a projection image having a relatively largest periodicity.

The predetermined method may include various methods. For example, it is possible to generate a plurality of projection images rotated in predetermined angular units, and to detect the periodicity of the laminated noise pattern by detecting a projection image having a periodicity of the laminated noise pattern among the plurality of generated projection images. Alternatively, the predetermined method may include generating a plurality of surface images rotated in a predetermined angle unit, detecting a periodic feature of the laminated noise pattern by detecting a surface image having a periodicity of the laminated noise pattern among the plurality of generated surface images have.

The periodicity of the stacked noise pattern may be estimated in other manners. In one embodiment, the watermark detection apparatus 200 rotates the 3D model. Then, the watermark detection apparatus 200 performs a filtering process for detecting a specific signal (e.g., printing trace) on the surface of the rotating 3D model. The watermark detection apparatus 200 analyzes output values of the filtering process performed. When the watermark detection apparatus 200 has the maximum value among the analyzed output values, it can be assumed that the laminated noise pattern has periodicity. Alternatively, the watermark detection apparatus 200 performs a digital slicing process on the 3D model to generate a digital slicing model. Then, the watermark detection apparatus 200 scans the 3D model. The watermark detection apparatus 200 may estimate the periodicity of the laminated noise pattern by analyzing the correlation between the generated digital slicing model and the scan model.

The watermark detection apparatus identifies the printing direction based on the periodicity of the detected lamination noise pattern (S1530).

16 is a flowchart of a watermark detection method according to another embodiment of the present disclosure.

Referring to Fig. 16, the watermark detection apparatus identifies the printing direction and then aligns the digital copy in the identified printing direction (S1610). The watermark detection apparatus detects a watermark embedded in the 3D model based on the aligned digital pattern (S1620). The watermark detection apparatus can convert the coordinate system for the aligned digital pattern into the cylindrical coordinate system and detect the inserted watermark from the radial component of the digital pattern converted into the cylindrical coordinate system. Further, the watermark detection apparatus may resample the vertex coordinates of the 3D model at regular intervals.

On the other hand, when a spread spectrum watermark is inserted in the 3D model, the watermark detection apparatus may detect the embedded watermark using a frequency analysis algorithm and a correlation technique for the digital pattern. Since the correlation technique has been described above, detailed description is omitted.

The watermark embedding method or the watermark embedding method according to the above-described various embodiments may be implemented as a program, and a non-transitory computer readable medium in which the program is stored may be provided.

A non-transitory readable medium is a medium that stores data for a short period of time, such as a register, cache, memory, etc., but semi-permanently stores data and is readable by the apparatus. In particular, the various applications or programs described above may be stored on non-volatile readable media such as CD, DVD, hard disk, Blu-ray disk, USB, memory card, ROM,

Various watermark embedding methods and watermark detection methods described above can be used variously. For example, the watermark insertion method and the watermark detection method may be used individually or together as independent software. Alternatively, the watermark embedding method and watermark embedding method can be used as an accessory program of 3D model design or editing program, or the watermark embedding method and watermark embedding method can be used as an accessory program of 3D scanner control software. When the insertion method and the detection method are used as an accessory program, the insertion method and the detection method can be implemented in the form of firmware in a desktop computer or a scanner. And, when the print model is scanned, a watermark can be detected. Alternatively, the insertion method and the detection method can be used as an accessory program of the 3D model sharing portal. When the insertion method and the detection method are used as an accessory program of the 3D shared portal, the detection method can confirm the identification number inserted when the model is uploaded.

On the other hand, a watermark can be inserted or detected in a smartphone using an IoT scanner dedicated to a smart phone or augmented reality technology. 3D models can be captured using smartphone apps and phone cameras. Alternatively, a 3D model may be captured using a 2D or dual camera. A watermark can be detected using the captured 3D model.

The watermark embedding method and watermark embedding method according to the present disclosure can be used in various embodiments. As a specific example, Alice may be a 3D model maker and copyright holder, and Bob may be a 3D model receiver. The 3D model that Bob receives can be leaked by a third party. In the case of the above example, existing digital security technology can be used. However, in order to detect the watermark in the existing non-blind method, it is necessary to construct a 3D model database. However, the blind watermark detection method according to the present disclosure does not require a 3D model database construction. Also, in the case of the above-described example, when the 3D printing process is included in the distribution process, watermark detection is impossible in the conventional method. That is, when a 3D model is copied, distributed, shared, or sold using an online sharing platform, protection is lost when a watermark is inserted in a conventional manner. However, if a watermark (or identification information) is inserted by the watermark embedding method according to the present disclosure, the identification information remains even if the third party copies it by 3D printing. Thus, in the case of the above example, if the watermark embedding method and the detection method of the present disclosure are used, the model maker can resolve the conflict by checking the mark displayed on the surface of the 3D model of the digital copying in the case of a copyright dispute of the 3D model.

As another example, the 3D model produced by Alice may be digitally copied and received by Bob, and the copy model received by Bob may be printed through at least one 3D printing or 3D scanning process, or the scanned model may be leaked . In general, 3D models can be distributed, shared or sold using an online sharing platform. If a protective device is inserted with existing technology, the blanket can disappear when Bob prints the 3D model. Alternatively, when the copying process of the 3D model includes a 3D scanning process, Bob may maliciously perform 3D printing to remove digital identification information.

Thus, in the case of the above example, Alice may insert copyright holder identification information into the model prior to delivery to the online sharing platform or Bob. The copyright holder identification information may be inserted using Alice's personal program, an editing program, using the watermark embedding method of the present disclosure. Alternatively, Alice can insert identification information when uploading to an online sharing platform. Alternatively, immediately before Bob can download the 3D model from the shared platform, identification information may be inserted. If the identification information is inserted using the watermark embedding method of the present disclosure, even if Bob uses 3D printing, the identification information remains. Accordingly, when a dispute occurs, it is possible to resolve the dispute by detecting the identification information inserted in the 3D model.

In another embodiment, a 3D model printed by Alice may be delivered to Bob through an off-line distribution process. Bob can spill out the printed 3D model, and can leak the scanned digital model after 3D scanning. Alternatively, Bob may leak the printing model after 3D scanning and 3D printing. Typically, after Alice prints a 3D model, she distributes, shares, or sells print models. However, the identification information inserted in the conventional manner disappears when the printing model is 3D-printed.

Therefore, in the case of the above example, Alice can print and distribute the digital model after inserting the copyright holder identification information into the digital model before printing. Even after the 3D model is printed and distributed via offline and copied to 3D printing, watermark insertion (or identification information) inserted in accordance with the present disclosure remains, so that the watermark embedding method and detection method of the present disclosure can be used for copyright dispute .

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It will be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the present invention.

100: watermark embedding device 200: watermark detection device
110: aligner 120: processor
210: 3D scanner 220: processor

Claims (22)

Aligning the 3D model in the printing direction based on the stacking direction of the stacked 3D model; And
Inserting a watermark of a predetermined pattern in a direction orthogonal to the printing direction in the aligned 3D model so that an inserted watermark is not related to the printing direction,
Wherein the inserting step comprises:
Transforming the coordinate system for the aligned 3D model into a cylindrical coordinate system and inserting the watermark of the predetermined pattern into a radial component of the 3D model with respect to an axis orthogonal to the printing direction of the cylindrical coordinate system. delete The method according to claim 1,
Wherein the inserting step comprises:
Wherein the predetermined pattern watermark is converted into a spectrum spread spectrum signal, and the watermark is synthesized with a sinusoidal signal formed by the radial component, and the watermark of the predetermined pattern is inserted. The method according to claim 1,
And performing uniform remeshing on the 3D model to uniformize the vertex distribution of the 3D model. 3D scanning a stacked 3D model to obtain a digital pattern;
Detecting the periodicity of the laminated noise pattern based on a predetermined method by rotating the obtained digital pattern by a predetermined angle unit; And
And identifying a printing direction based on the detected periodicity of the laminated noise pattern. 6. The method of claim 5,
In the predetermined method,
A method of generating a plurality of projection images rotated in units of the predetermined angle and detecting a periodic feature of a laminated noise pattern by detecting a projection image having a periodicity of a laminated noise pattern among the plurality of generated projection images, Wherein a plurality of surface images rotated in units of angles are generated and a periodic feature of a laminated noise pattern is detected by detecting a surface image having a periodicity of a laminated noise pattern among the plurality of generated surface images. 6. The method of claim 5,
Wherein the detecting the periodicity comprises:
Detecting a peak value larger than a preset threshold value at a predetermined cycle through frequency analysis as having periodicity. 6. The method of claim 5,
Aligning the digital version in the identified printing direction; And
And detecting a watermark embedded in the 3D model based on the aligned digital pattern. 9. The method of claim 8,
Wherein the step of detecting the watermark comprises:
Converting the coordinate system of the aligned digital bone into a cylindrical coordinate system and detecting the inserted watermark from a radius component of the digital bone converted into the cylindrical coordinate system. 9. The method of claim 8,
And resampling the vertex coordinates of the 3D model at regular intervals. 9. The method of claim 8,
Wherein the step of detecting the watermark comprises:
Wherein the inserted watermark is detected using a frequency analysis algorithm and a correlation technique for the aligned digital version when a spread spectrum watermark is inserted in the 3D model. An aligning unit for aligning the 3D model in a printing direction based on a stacking direction of the stacked 3D models; And
And a processor for inserting a predetermined pattern of watermarks into the aligned 3D model in a direction orthogonal to the printing direction so that an inserted watermark is not associated with the printing direction,
The processor comprising:
Wherein the coordinate system for the aligned 3D model is converted into a cylindrical coordinate system and the watermark of the predetermined pattern is inserted into a radius component of the 3D model with respect to an axis orthogonal to the printing direction of the cylindrical coordinate system. delete 13. The method of claim 12,
The processor comprising:
Wherein the predetermined watermark pattern is converted into a spectrum spread spectrum signal, and the watermark is combined with a sinusoidal wave signal formed by the radial component to insert the predetermined pattern watermark. 13. The method of claim 12,
The processor comprising:
And performs uniform remeshing on the 3D model to uniformize the vertex distribution of the 3D model. A 3D scanner for acquiring a digital pattern by scanning the 3D model of the laminated 3D model; And
And a processor for detecting the periodicity of the laminated noise pattern based on a predetermined scheme by rotating the obtained digital pattern by a predetermined angle unit,
The processor comprising:
And identifies the printing direction based on the periodicity of the detected laminated noise pattern. 17. The method of claim 16,
In the predetermined method,
A method of generating a plurality of projection images rotated in units of the predetermined angle and detecting a periodic feature of a laminated noise pattern by detecting a projection image having a periodicity of a laminated noise pattern among the plurality of generated projection images, Wherein a plurality of surface images rotated in units of angles are generated and a periodic feature of a laminated noise pattern is detected by detecting a surface image having a periodicity of a laminated noise pattern among the plurality of generated surface images. 17. The method of claim 16,
The processor comprising:
And detects that a peak value larger than a predetermined threshold value is detected at regular intervals through frequency analysis as having periodicity. 17. The method of claim 16,
The processor comprising:
Aligning the digital version in the identified printing direction and detecting a watermark embedded in the 3D model based on the aligned digital version. 20. The method of claim 19,
The processor comprising:
Converts the coordinate system for the aligned digital bone into a cylindrical coordinate system and detects the inserted watermark from the radial component of the digital bone converted into the cylindrical coordinate system. 20. The method of claim 19,
The processor comprising:
And resample the vertex coordinates of the 3D model at regular intervals. 20. The method of claim 19,
The processor comprising:
Wherein the inserted watermark is detected using a frequency analysis algorithm and a correlation technique for the aligned digital pattern when a spread spectrum watermark is inserted in the 3D model.

Images