KR101925011B1 - Method and apparatus for inserting and detecting wartermark - Google Patents

Abstract

A watermark detection method according to an embodiment of the present invention includes dividing a panorama image into a plurality of partial images that are not overlapped with each other, calculating a first partial image having the largest number of minutiae corresponding to the target image among the plurality of partial images Generating a target image by inversely rendering a target image on the basis of the viewpoint information, and generating a panorama image based on the target image, And checking whether the target image corresponds to a part of the panoramic image by comparing the watermark pattern included in the image.

Description

[0001] METHOD AND APPARATUS FOR INSERTING AND DETECTING WARTERMARK [

More particularly, the present invention relates to a method and an apparatus for inserting / detecting a watermark, and more particularly, a watermark is inserted into a panoramic image, a watermark is detected from a target image in order to check whether a target image originates from a panoramic image, And a watermark embedding / detection method and apparatus.

The panoramic image is an image capable of securing a 360 ° foreground view, allowing the user to freely select a viewpoint within the image, thereby providing a feeling that the user is located in a virtual space and is looking directly.

Recently, interest in panoramic images has increased, and copyright protection for panoramic images has become important. However, compared to the rapidly developing panoramic imaging industry, the technology to protect the copyright has not been established yet.

Digital watermarking is a technique to protect copyrights from illegal copying and illegal distribution of existing media contents. Currently, digital watermarking technology for content such as 2D, Streoscopic 3D, 3D mesh, However, watermarking technology for protecting panoramic images has not existed at present.

Particularly, due to the nature of the panorama image, the entire panorama image is not leaked, but the partial image of the panorama image is leaked, resulting in copyright infringement. For example, when a concert hall is made into a panorama image, the partial image in the stage direction has content value even if it is not a whole panorama image, and partial images of various angles can be collected and reproduced as panorama contents. There is a need for a technique capable of detecting a watermark in an image.

A problem to be solved by an embodiment of the present invention is to provide a technique capable of detecting a watermark from a partial image of a panoramic image.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are intended to provide further explanation of the invention as claimed.

The watermark embedding method according to an embodiment of the present invention includes dividing an original image to be rendered into a 360-degree panoramic image into a plurality of blocks, and for each of the plurality of blocks, Converting a spatial domain into a frequency coefficient, inserting a watermark pattern into the transformed frequency coefficient in each block, restoring the frequency coefficient into which the watermark pattern is inserted into the spatial domain And generating the 360 ° panoramic image by rendering the watermark pattern embedded image in each block after the reconstructing step.

In this case, the original image includes an isosamental image, a conical image, a cylindrical image, a quasi-cylindrical image, a cylindrical image, a fisheye image, a Mercator image, and a sinusoidal image.

Also, the plurality of blocks may be in a lattice form that do not overlap with each other.

In addition, the converting into the frequency coefficient may include converting the spatial domain into the frequency coefficient through a discrete cosine transform.

A watermark detection method according to an embodiment of the present invention includes dividing a 360 ° panoramic image into a plurality of partial images that are not overlapped with each other, Determining a partial image based on the first partial image, deriving the viewpoint information at which the target image is located in the 360 ° panoramic image based on the first partial image, rendering the target image inversely based on the viewpoint information, And comparing the watermark pattern included in the target original image with the watermark pattern included in the 360 ° panoramic image to determine whether the target image is included in a portion of the 360 ° panoramic image And determining if it is applicable.

At this time, the feature points may be extracted through a scale-invariant feature transform (SIFT).

The step of deriving the viewpoint information may include deriving the viewpoint information by calculating conversion information between the target image and the first partial image.

In addition, the step of deriving the viewpoint information comprises the steps of: (a) calculating transformation information between the target image and the ith partial image (i is incremented from 1 in ascending order) through an euclidean transformation matrix; ) Generating an (i + 1) -th partial image obtained by transforming the i-th partial image based on the transform information, and (c) generating a peak signal-to-noise ratio (PSNR) between the target image and the 1) th partial image and the horizontal angle of the (i + 1) -th partial image as the viewpoint information of the target image when the PSNR does not exceed the predetermined threshold, And the step (c) may be performed again.

In addition, the step of generating the target original image may include a step of reconstructing a pixel of a hole through interpolation.

The step of determining whether the target image corresponds to a portion of the 360 ° panoramic image may include dividing the target original image into a plurality of blocks so as to correspond to positions of the plurality of blocks, Selecting a block in which pixels included in the plurality of blocks are restored by a first threshold value or more, converting a spatial region of an image in each selected block into a frequency coefficient for each of the selected blocks, Calculating a correlation between the frequency coefficients in each of the selected blocks and the watermark pattern included in the 360 ° panoramic image, and averaging the correlation coefficients calculated from each selected block to a second Determining that the target image is a portion of the 360 ° panorama image if the threshold value is greater than a threshold value It can be included.

Meanwhile, the step of converting into the frequency coefficient may include a step of converting the spatial domain into the frequency coefficient through a discrete cosine transform.

At this time, the target original image includes an isotropic image, a conical image, a cylindrical image, a quasi-cylindrical image, a cylindrical image, a fisheye image, a Mercator image, and a sinusoidal image.

A watermark embedding apparatus according to an embodiment of the present invention includes a block assigning unit for dividing a plurality of blocks into an original image in a format capable of generating a 360 panoramic image, A transform unit for transforming a spatial domain of an included image into a frequency coefficient, an inserting unit for inserting a watermark pattern into the transformed frequency coefficient in each block, And a panoramic image generation unit for rendering a 360 ° panoramic image by rendering an original image in which a watermark pattern is embedded in each of the blocks.

A watermark detection apparatus according to an embodiment of the present invention includes a divider for dividing a 360 ° panoramic image into a plurality of divided partial images so as not to overlap with each other, A feature point comparison unit for calculating a first partial image, a time point information derivation unit for deriving time point information on which the target image is located in the 360 ° panorama image based on the first partial image, And a controller for comparing the watermark pattern included in the 360 ° panorama image with the watermark pattern included in the 360 ° panorama image to determine whether the target image corresponds to a portion of the 360 ° panorama image, And a mark detection unit.

According to the embodiment of the present invention, not only the entire panorama image but also partial images corresponding to a specific time point of the panorama image can be watermarked to prevent the copyright infringement problem on the panorama image.

1 is a functional block diagram of a watermark embedding apparatus according to an embodiment of the present invention.
2 is a flow chart showing the process of a watermark embedding method according to an embodiment of the present invention.
3 is a functional block diagram of a watermark detection apparatus according to an embodiment of the present invention.
FIG. 5 is a diagram illustrating an example of time information of a target image in a panoramic image.
FIG. 6 is an exemplary view for explaining a partial image in which panorama images are divided so as not to overlap with each other according to an embodiment of the present invention.
FIG. 7 is a diagram illustrating an example of comparing feature points between a partial image and a target image according to an exemplary embodiment of the present invention. Referring to FIG.
FIG. 8 is a diagram illustrating an example of restoring a pixel of a hole through an image interpolation method according to an embodiment of the present invention. Referring to FIG.

DETAILED DESCRIPTION OF THE EMBODIMENTS Reference will now be made in detail to the preferred embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment according to the present invention will be described in detail with reference to the accompanying drawings.

The embodiments disclosed herein should not be construed or interpreted as limiting the scope of the present invention. It will be apparent to those of ordinary skill in the art that the description including the embodiments of the present specification has various applications. Accordingly, any embodiment described in the Detailed Description of the Invention is illustrative for a better understanding of the invention and is not intended to limit the scope of the invention to embodiments.

The functional blocks shown in the drawings and described below are merely examples of possible implementations. In other implementations, other functional blocks may be used without departing from the spirit and scope of the following detailed description. Also, although one or more functional blocks of the present invention are represented as discrete blocks, one or more of the functional blocks of the present invention may be a combination of various hardware and software configurations that perform the same function.

In addition, the expression "including any element" is merely an expression of an open-ended expression, and is not to be construed as excluding the additional elements.

Further, when a component is referred to as being connected or connected to another component, it may be directly connected or connected to the other component, but it should be understood that there may be other components in between.

Also, the expressions such as 'first, second', etc. are used only to distinguish a plurality of configurations, and do not limit the order or other features between configurations.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

1 is a functional block diagram of a watermark embedding apparatus 100 according to an embodiment of the present invention.

1, a watermark embedding apparatus 100 according to an embodiment of the present invention includes a block allocating unit 110, a transforming unit 120, an inserting unit 130, a restoring unit 140, And an image generating unit 150.

The block allocation unit 110 divides the original image to be rendered into a panorama image into a plurality of blocks. At this time, 'block' is used as the meaning of 'patch' in the term 'patch image' which means a predetermined part of a specific image. On the other hand, the original image in a format that can be rendered as a panoramic image is composed of an isotropic image, a conical image, a cylindrical image, a similar cylindrical image, a cylindrical image, a fisheye image, a Mercator image, sinusoidal) images.

The conversion unit 120 converts the spatial domain of the image included in each block into a frequency coefficient. For example, the spatial domain is transformed into a frequency coefficient through a discrete cosine transform.

 The insertion unit 130 inserts the watermark pattern into the transformed frequency coefficients in each block, and the reconstruction unit 140 restores the frequency coefficient into which the watermark pattern is inserted in each block into the spatial domain, The image generation unit 150 generates a panorama image by rendering an original image in which a watermark pattern is embedded in frequency coefficients within each block.

Hereinafter, a concrete procedure for operating each configuration of the watermark embedding apparatus 100 will be described with reference to FIG.

The block allocation unit 110, the conversion unit 120, the insertion unit 130, the decompression unit 140, and the panorama image generation unit 150, which are included in the above-described embodiments, And a microprocessor for carrying out these instructions.

2 is a flow chart showing the process of a watermark embedding method according to an embodiment of the present invention. Each step of the watermark embedding method according to Fig. 2 can be performed by the watermark embedding apparatus 100 described with reference to Fig.

Before describing each step, the panoramic image is produced in such a form that it can be viewed at 360 ° x 180 ° (horizontal angle x vertical angle) by photographing a real space using a camera and rendering it. At this time, the original image in the format capable of generating the panorama image is composed of an isotropic image, a conical image, a cylindrical image, a pseudo cylindrical image, a cylindrical image, a fisheye image, a meridional image obtained by projecting a sphere on a two- Mercator images, sinusoidal images, and the like.

Particularly, the most widely used format when dealing with panoramic images is an isometric image. The width of the image is 360 °, which is the width of the image in the direction of the longitude, and the height of the image is 180 °, which is the angle between the ceiling and floor of the sphere. . In the drawings of the present specification, an original image is shown as an original image. However, the present invention can be applied to an original image in a format other than the isometric image. Each step will be described below.

First, the block allocation unit 110 divides the original image to be rendered into a panorama image into a plurality of blocks (S210). The plurality of blocks are positions where watermark is inserted. When a watermark is inserted in a plurality of blocks, a watermark can be detected in an arbitrary area of the panorama image, and watermark can be efficiently protected from a cropping attack. can do.

In this case, the block allocation unit 110 may allocate a plurality of blocks in a non-overlapping grid format. By splitting the multiple blocks into a grid format, the watermark can be extracted from the partial images at all angles, since the watermark can be inserted into all regions without missing portions in the original image.

Also, when the partial image of the panoramic image is inversely rendered, the area and quality of the reconstructed image vary depending on the vertical angle of the partial image. When the range of the vertical angle of the panoramic image is from -90 degrees to +90 degrees, the partial image whose vertical angle is close to -90 degrees and +90 degrees is restored to a large area in the inverse rendering, but the quality is low, In the near partial image, the area to be reconstructed in the reverse rendering is small, but the quality is high. That is, the quality and area of the reconstructed image are inversely proportional to the inverse rendering. In this case, when a plurality of blocks are allocated in a lattice format and a watermark is inserted into each block, a large area is restored even if the reconstructed quality is low, so that the number of watermarks that can be extracted increases. Therefore, when a watermark is inserted into each block divided into a grid, it is possible to detect a watermark from a larger number of blocks at a vertical angle restored to a low quality, thereby enhancing the robustness of watermark detection.

Next, the transforming unit 120 transforms the spatial domain of the image included in each block into a frequency coefficient (S220). For example, the spatial domain can be transformed into a frequency coefficient through a discrete cosine transform. (Cox, Ingemar J., et al., "Secure spread spectrum watermarking for multimedia." IEEE transactions on image processing 6.12 (1997): 1673-1687.)

The inserter 130 generates a watermark pattern from the secret key through the reference pattern generator (S225), and inserts the watermark pattern into the transformed frequency coefficients in each block (S230). At this time, when the watermark is inserted into the low frequency band, the invisibility of the watermark can be lowered. When the watermark is inserted into the high frequency band, the watermark is susceptible to attacks such as compression, A watermark pattern can be inserted into a frequency coefficient of an intermediate frequency band through a zigzag scanning method. (Barni, Mauro, et al., "A DCT-domain system for robust image watermarking." Signal processing 66.3 (1998): 357-372.)

The restoration unit 140 restores the frequency coefficient into which the watermark pattern is inserted in each block into the spatial domain (S240). For example, a frequency coefficient in which a watermark pattern is inserted through an inverse discrete cosine transform can be restored into a spatial domain.

Thereafter, the panoramic image generation unit 150 combines the images of the restored blocks through the restoration unit 140 (S245), renders the original image having the watermark pattern embedded in the frequency coefficients within each block, An image can be generated (S250).

The watermark detection apparatus 300 according to an exemplary embodiment of the present invention confirms whether a target image corresponds to a portion of a panoramic image generated according to the embodiment of FIGS. 1 and 2 through watermark detection. 3 is a functional block diagram of a watermark detection apparatus 300 according to an embodiment of the present invention.

3, the watermark detection apparatus 300 according to an exemplary embodiment of the present invention includes a segmentation unit 310, a feature point comparison unit 320, a viewpoint information derivation unit 330, an inverse rendering unit 340 And a watermark detection unit 350.

The division unit 310 divides the panoramic image into a plurality of partial images that are not overlapped with each other. The minutiae point comparator 320 compares the minutiae corresponding to the target image among the plurality of partial images divided by the division unit 310 A large number of first partial images are determined. At this time, feature points can be extracted through SIFT (scale-invariant feature transform).

The viewpoint information deriving unit 330 derives the viewpoint information on which the target image is located in the panorama image based on the first partial image and the inverse rendering unit 340 derives viewpoint information derived by the viewpoint information deriving unit 330 The target image is inversely rendered to generate a target original image corresponding to the target image.

The watermark detection unit 350 compares the watermark pattern included in the target original image generated by the inverse rendering unit 340 with the watermark pattern included in the panoramic image to determine whether the target image corresponds to a part of the panoramic image.

A concrete procedure of the operation of each configuration of the watermark detection apparatus 300 will be described with reference to FIG.

Meanwhile, the division unit 310, the feature point comparison unit 320, the viewpoint information derivation unit 330, the inverse rendering unit 340, and the watermark detection unit 350 included in the above-described embodiment are programmed to perform these functions A memory including instructions, and a microprocessor for carrying out these instructions.

4 is a flowchart illustrating a process of a watermark detection method according to an embodiment of the present invention. Each step of the watermark detection method according to Fig. 4 can be performed by the watermark detection apparatus 300 described with reference to Fig.

1 and 2, a watermark embedded in a panoramic image can be detected from an original image in which a panoramic image is inversely rendered. Therefore, in order to detect a watermark from a target image assumed to have flowed out, To detect the watermark from the target original image.

At this time, in order to restore the target original image, for example, the appearing rectangular image, it is necessary to set a certain value and inverse rendering setting value, so that the point of view indicating the target portion of the panorama image is necessarily required Do. FIG. 5 is a diagram illustrating an example of time information of a target image in a panoramic image. As shown in FIG. 5, when the panoramic image is a spherical shape, the viewpoint information includes a vertical angle v and a horizontal angle of the line segment connecting the center of the sphere, which is determined to be the target image, h. < / RTI >

Then, based on the view information of the target image, a watermark may be detected from the target original image inversely rendering the target image to check whether the target image corresponds to a part of the panorama image. Each step will be described below.

In order to know which portion of the panoramic image the target image corresponds to, the division unit 310 divides the panoramic image into a plurality of partial images that are not overlapped with each other (S410). In order to obtain the viewpoint information of the target image, it is very inefficient to find the viewpoint information of the target image by comparing with all the parts of the 360 ° panorama image. Accordingly, in the embodiment of the present invention, as shown in FIG. 6, the view information of the target image is closely located through the partial image.

6 is a diagram illustrating an example of a partial image obtained by dividing the panorama images so that the divided images do not overlap with each other according to an embodiment of the present invention. Referring to FIG. 6, the segmentation unit 310 divides the panoramic image into partial images such that each partial image includes the entire panorama image, but no distortion exists in the region where the adjacent partial images overlap each other. The values below the partial image in FIG. 6 indicate the vertical angle v and the horizontal angle h of each partial image in the panoramic image.

In step S420, the feature point comparison unit 320 determines a first partial image that is most similar to the target image among the partial images divided by the divided unit 310. For this, the feature point comparator 320 may extract the feature point from the partial image and the target image, and determine the partial image having the largest feature point coinciding with the target image as the first partial image.

That is, since the partial image is divided in advance before comparing the target image and the minutiae point, the viewpoint information of the target image and the first partial image may not perfectly match. However, if the two images have close viewpoint information, . Therefore, it is possible to find a part including the same object through a scale-invariant feature transform (SIFT) feature point matching with RST (rotation, scale, translation). Using this characteristic, the partial image having the largest number of the feature points coinciding with the target image can be determined as the first partial image having the viewpoint information close to the target image.

Next, the viewpoint information derivation unit 330 derives the viewpoint information of the target image on the panorama image based on the first partial image (S430). For this, the viewpoint information derivation unit 330 calculates transformation information indicating how far the first partial image is vertically transformed from the target image, how it is horizontally transformed, and how much it is rotated from the target image, It is possible to derive accurate viewpoint information of the target image based on the basis.

), 수직 변환 정보(

) 및 회전 변환 정보(

)를 포함한다. To do this, in step S430, transformation information between the target image and the i-th partial image (i is incremented by 1 every time the step S430 is restarted, starting from 1) can be calculated through an euclidean transformation matrix (S431). At this time, the conversion information includes horizontal conversion information

), Vertical conversion information (

) And rotation conversion information (

).

,

를 부분 영상의 특징점이라 하고,

,

를 타겟 영상에서 대응되는 특징점이라 할 때, 아래 수학식 1의 유클리디안 변환 행렬을 통해 전체 특징점의 매칭쌍을 가장 잘 근사하는 a, b, c, d를 구할 수 있다. E.g,

,

Is a feature point of a partial image,

,

A, b, c, and d that best approximate a matching pair of all the minutiae points can be obtained through the Euclidean transformation matrix of Equation (1) below.

) 및 수직 변환 정보(

)를 의미하며,

방정식의

는 회전 변환 정보(

)를 의미한다. Here, c and d are horizontal conversion information (

) And vertical conversion information (

),

Of the equation

Rotation conversion information (

).

Next, an i + 1 partial image obtained by converting the i-th partial image is generated based on the conversion information obtained in step S431 (S432). FIG. 7 is a diagram illustrating an example of comparing feature points between a partial image and a target image according to an exemplary embodiment of the present invention. Referring to FIG. Referring to FIG. 7, the target image and the partial image are in a state in which a plurality of feature points are matched with each other, but the horizontal angle is remarkably different. This is because the target image is rotated relative to the origin relative to the partial image. Therefore, the viewpoint of the target image can be accurately grasped by converting the viewpoint of the partial image through the transformation information between the target image and the partial image.

)가 충분히 작다면 수평 변환 정보(

) 및 수직 변환 정보(

)만을 사용하여 제i 부분 영상을 변환할 수 있다. 하나의 축을 기준으로 부분 영상을 변경하면 보다 정확하게 타겟 영상과 부분 영상이 일치하는 시점 정보를 찾아낼 수 있기 때문이다.At this time, in one embodiment of the present invention,

) Is sufficiently small, horizontal conversion information (

) And vertical conversion information (

) Can be used to transform the i < th > partial image. If the partial image is changed on the basis of one axis, the time information in which the target image and the partial image coincide more accurately can be found.

)가 수평 변환 정보(

)의 2배보다 크다면 아래 수학식 2와 같이 제i 부분 영상의 수직 각도 v만 변환하여 제i+1 부분 영상을 생성하고, 수직 변환 정보(

)가 수평 변환 정보(

)의 2배보다 작다면 아래 수학식 3과 같이 제i 부분 영상의 수평 각도 h만 변환하여 제i+1 부분 영상을 생성할 수 있다. At this time,

) ≪ / RTI >

(I + 1) th partial image by transforming only the vertical angle v of the i th partial image as shown in Equation (2) below,

) ≪ / RTI >

), It is possible to generate the (i + 1) th partial image by transforming only the horizontal angle h of the ith partial image as shown in Equation (3) below.

)가 수평 변환 정보(

)에 2배를 한 값을 비교하여 영향이 더 큰 축에 해당하는 각도를 변경할 수 있다. The degree of freedom (360 °) of the horizontal angle is two times larger than the degree of freedom (180 °) of the vertical angle on the characteristic of the 360 ° panoramic image. Therefore,

) ≪ / RTI >

), And then change the angle corresponding to the larger axis.

)를 곱하여 부분 영상의 수직 각도 또는 수평 각도를 변환할 수 있는데, 변환 가중치(

)가 작으면 각도 수정을 조금씩 하게 되어 방향적 정확도를 올릴 수 있고, 변환 가중치(

)가 크면 각도 수정을 상대적으로 크게 하여 연산을 줄일 수 있다. On the other hand, as shown in Equations (2) and (3), conversion information

) To convert the vertical angle or the horizontal angle of the partial image. The conversion weight (

) Is small, it is possible to increase the directional accuracy by slightly modifying the angle, and the conversion weight

) Is large, it is possible to reduce the calculation by relatively increasing the angle correction.

If the peak signal-to-noise ratio (PSNR) between the target image and the (i + 1) th partial image exceeds a predetermined threshold after generating the (i + 1) th partial image in step S432, The viewpoint information of the target image in the panoramic image may be derived based on the vertical angle and the horizontal angle converted into one partial image (S433). In this case, if the PSNR does not exceed the predetermined threshold value, steps S431 to S433 may be performed again to more accurately derive the viewpoint information when the target image and the viewpoint information of the partial image coincide with each other.

The inverse rendering unit 340 inversely renders the target image based on the view information derived by the view information deriving unit 330 to generate a target original image corresponding to the target image in operation S440. At this time, the target original image includes an isometric image, a conical image, a cylindrical image, a similar cylindrical image, a cylindrical image, a fisheye image, a Mercator image, and a sinusoidal image. Since the inverse rendering unit 340 knows exactly where the target image is located in the panorama image through the viewpoint information, the target image can be generated by inverse rendering by reflecting the viewpoint information.

Since only the pixel information included in the target image can be inversely rendered, only the partial region of the original image, on which the panorama image is based, is restored as shown in FIG. In addition, since distortions occur during inverse rendering, holes are generated in the region corresponding to the restored portion of the target original image. Since the hole has no pixel information at the time of watermark detection, the robustness of watermark detection can be lowered.

In order to prevent this, the inverse rendering unit 340 may restore the pixel of the hole through interpolation (S445). FIG. 8 is a diagram illustrating an example of restoring a pixel of a hole through an image interpolation method according to an embodiment of the present invention. Referring to FIG. Referring to FIG. 8, the inverse rendering unit 340 may interpolate the hall pixel with the average value of the pixel information, not the hole, among the pixels around the pixel to be interpolated through the image interpolation method.

Then, the watermark detection unit 350 compares the watermark pattern included in the target original image generated by the inverse rendering unit 340 with the watermark pattern included in the panoramic image to determine whether the target image is a part of the panoramic image (S450).

In more detail, in step S450, the target original image is divided into a plurality of blocks so as to correspond to the positions of the plurality of blocks allocated to the original image, and the pixels included in the block among the plurality of blocks of the target original image are restored to a predetermined threshold value or more The block can be selected (S451). Thereafter, the spatial domain of the image within each selected block can be transformed into a frequency coefficient (S452).

In order to compare the transformed frequency coefficient with the watermark pattern included in the panoramic image, a watermark pattern is generated from the secret key used for the panoramic image through the reference pattern generator (S453). In step S453, The correlation coefficient may be calculated by comparing the frequency coefficient with the watermark pattern generated in step S453 (S454). In order to remove the outlier, a part of the block with the highest correlation coefficient and a part of the lowest block are excluded, and the average of the remaining correlation coefficients is averaged. If the average of the correlation coefficients is equal to or greater than the predetermined threshold value, .

According to the embodiment described above, not only the entire panorama image but also the partial image corresponding to the specific time point of the panorama image can be watermarked to prevent the copyright infringement problem on the panorama image.

The above-described embodiments of the present invention can be implemented by various means. For example, embodiments of the present invention may be implemented by hardware, firmware, software, or a combination thereof.

In the case of hardware implementation, the method according to embodiments of the present invention may be implemented in one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs) , FPGAs (Field Programmable Gate Arrays), processors, controllers, microcontrollers, microprocessors, and the like.

In the case of an implementation by firmware or software, the method according to embodiments of the present invention may be implemented in the form of a module, a procedure or a function for performing the functions or operations described above. The software code can be stored in a memory unit and driven by the processor. The memory unit may be located inside or outside the processor, and may exchange data with the processor by various well-known means.

Thus, those skilled in the art will appreciate that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is therefore to be understood that the embodiments described above are to be considered in all respects only as illustrative and not restrictive. The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

100: watermark embedding device
110: block allocation unit
120:
130:
140:
150: panoramic image generation unit
300: a watermark detection device
310: Split installment
320: Feature point comparator
330: viewpoint information derivation unit
340:
350: watermark detection unit

Claims (16)

A watermark embedding method performed by one or more processors,
By dividing the original image to be rendered with the 360-degree panoramic image into a plurality of blocks in a lattice form that do not overlap with each other, an image corresponding to each block when rendered with the 360-degree panoramic image is divided into a vertical angle and a horizontal angle To have different values;
Converting a spatial domain of an image included in each of the plurality of blocks into a frequency coefficient for each of the plurality of blocks;
Inserting a watermark pattern in each of the transformed frequency coefficients in the plurality of blocks;
Restoring each of the frequency coefficients into which the watermark pattern is inserted in the plurality of blocks into a spatial domain; And
And rendering the 360-degree panoramic image by rendering the watermark pattern embedded in each of the plurality of blocks after the reconstructing step
The method includes the steps of: when the 360 ° panorama image is divided into a plurality of partial images to determine whether the target image corresponds to a portion of the 360 ° panorama image, To be watermarked.
The method according to claim 1,
The original image,
One of the following types of images: an isometric image, a conical image, a cylindrical image, a pseudocylinder image, a cylindrical image, a fisheye image, a mercator image, and a sinusoidal image
How to insert a watermark.
delete The method according to claim 1,
The step of converting into the frequency coefficient may comprise:
And transforming the spatial domain into the frequency coefficients via a discrete cosine transform
How to insert a watermark.
A watermark detection method performed by one or more processors to determine if a target image corresponds to a portion of a 360 ° panoramic image generated according to the method of claim 1,
Dividing the 360 ° panoramic image into a plurality of partial images that are not overlapped with each other;
Determining a first partial image having the largest number of feature points coinciding with the target image among the plurality of partial images;
Deriving the viewpoint information at which the target image is located in the 360 ° panoramic image based on the first partial image;
Generating a target original image that is an image corresponding to the target image by inversely rendering the target image based on the viewpoint information; And
And comparing the watermark pattern included in the target original image with a watermark pattern included in the 360 ° panoramic image to determine whether the target image corresponds to a portion of the 360 ° panoramic image
A watermark detection method.
6. The method of claim 5,
The feature point may be,
Extracted through a scale-invariant feature transform (SIFT)
A watermark detection method.
6. The method of claim 5,
Wherein the step of deriving the viewpoint information comprises:
And deriving the viewpoint information by calculating transformation information between the target image and the first partial image
A watermark detection method.
8. The method of claim 7,
Wherein the step of deriving the viewpoint information comprises:
(a) calculating transformation information between the target image and an ith partial image (i is incremented from 1 in ascending order) through an euclidean transformation matrix;
(b) generating an (i + 1) -th partial image obtained by transforming the i-th partial image based on the transformation information; And
(c) when a peak signal-to-noise ratio (PSNR) between the target image and the (i + 1) th partial image exceeds a predetermined threshold value, a vertical angle and a horizontal angle of the (i + And if the PSNR does not exceed the predetermined threshold, performing the steps (a) to (c) again
A watermark detection method.
6. The method of claim 5,
Wherein the generating of the target original image comprises:
And restoring the pixels of the hole through interpolation.
A watermark detection method.
6. The method of claim 5,
Wherein the step of determining whether the target image corresponds to a portion of the 360 ° panoramic image comprises:
Dividing the target original image into a plurality of blocks so as to correspond to positions of the plurality of blocks;
Selecting a block among the plurality of blocks of the target original image in which pixels included in the plurality of blocks are restored by a first threshold value or more;
Converting, for each of the selected blocks, a spatial domain of an image in each of the selected blocks into a frequency coefficient;
Calculating a correlation between the frequency coefficients within each of the selected blocks and a watermark pattern included in the 360 ° panoramic image; And
Determining that the target image is a portion of the 360 ° panoramic image if the average of the correlation coefficients computed from each of the selected blocks is greater than or equal to a second threshold value
A watermark detection method.
11. The method of claim 10,
The step of converting into the frequency coefficient may comprise:
And transforming the spatial domain into the frequency coefficients via a discrete cosine transform
A watermark detection method.
6. The method of claim 5,
The target image may include:
One of the following types of images: an isometric image, a conical image, a cylindrical image, a pseudocylinder image, a cylindrical image, a fisheye image, a mercator image, and a sinusoidal image
A watermark detection method.
An original image in a format capable of generating a 360 ° panoramic image is divided into a plurality of blocks in a lattice form that do not overlap with each other so that an image corresponding to each block when rendered with the 360 ° panoramic image is divided into vertical A block assigning unit for setting the angle and the horizontal angle to have different values;
A transform unit for transforming a spatial region of an image included in each of the plurality of blocks into frequency coefficients for each of the plurality of blocks;
An inserting unit for inserting a watermark pattern into each of the transformed frequency coefficients in the plurality of blocks;
A restoring unit for restoring each of the frequency coefficients into which the watermark pattern is inserted in the plurality of blocks into a spatial domain; And
And a panorama image generation unit for rendering the 360 ° panorama image by rendering an image in which the watermark pattern is embedded in each of the plurality of blocks,
When the 360 ° panoramic image is divided into a plurality of partial images so as to determine whether the target image corresponds to a part of the 360 ° panoramic image, each partial image has time information based on the vertical angle and the horizontal angle A watermark embedding device.
13. A watermark detection apparatus for verifying whether a target image corresponds to a part of a 360 占 panoramic image generated by the apparatus of claim 13,
Dividing the 360 占 panoramic image into a plurality of partial images which are not overlapped with each other;
A feature point comparing unit for determining a first partial image having the largest number of feature points coinciding with the target image among the plurality of partial images;
A time information deriving unit for deriving time point information of the target image in the 360 ° panorama image based on the first partial image;
An inverse rendering unit for inversely rendering the target image based on the view information and generating a target original image corresponding to the target image; And
And a watermark detection unit for comparing the watermark pattern included in the target original image with a watermark pattern included in the 360 ° panoramic image to determine whether the target image corresponds to a portion of the 360 ° panoramic image
A watermark detection apparatus.
A program stored in a computer-readable medium for causing a processor to perform the method of any one of claims 1 to 12 and 12 to 12.
A computer-readable recording medium having recorded thereon a program for causing a processor to perform the method of any one of claims 1 to 12,

Images