A method for selecting a printing image anti-counterfeiting watermark embedding and extraction channel

Abstract

This invention discloses a method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images, belonging to the field of anti-counterfeiting printing technology. The method of this invention breaks with the default approach of extracting watermarks from the same channel in which they are embedded. The watermark extraction position and embedding position proposed in this invention can be the same or different. Instead, regardless of whether the watermark is embedded in RGB or CMYK color mode, it is extracted in RGB color mode, resulting in better extraction performance.

Description

Technical Field

[0001] This invention relates to the field of anti-counterfeiting printing technology, and in particular to a method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images. Background Technology

[0002] An invisible watermark is embedded in the printed image. After printing and scanning, the watermark is extracted to verify the authenticity of the printed product, thus achieving the purpose of anti-counterfeiting printing. This method does not require the addition of new materials or processes during printing. This anti-counterfeiting method does not increase printing costs or the difficulty of printing processes, therefore, this technology has received much attention in recent years.

[0003] In recent years, scholars have realized that printed images are in CMYK color mode, while traditional digital watermarks are in RGB color mode. Furthermore, scanners that scan printed images are also driven by RGB color mode. In other words, regardless of whether the watermark is embedded in RGB or CMYK color mode, it must undergo a color mode conversion after printing and scanning.

[0004] This invention takes into account the influence of different color modes and channels during the scanning process, and proposes a method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images, aiming to reduce the impact of color mode conversion on the robustness of watermarks. Summary of the Invention

[0005] The purpose of this invention is to address the technical deficiencies in the existing technology by providing a method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images.

[0006] The technical solution adopted to achieve the purpose of this invention is: A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images includes the following steps: Step 1: Separate the host image into channels, select one channel as the embedding channel, and extract the diagonal high-frequency coefficients of the host image in the embedding channel. If the host image is in RGB color mode, it is separated into three channels: R, G, and B, and the B channel is selected as the embedding channel; if the host image is in CMYK color mode, it is separated into four channels: C, M, Y, and K, and the C channel is selected as the embedding channel. Step 2: Encrypt the watermark to be embedded to obtain a holographic watermark image, and extract the diagonal high-frequency coefficients of the holographic watermark image. The high frequency coefficients of the diagonal of the holographic watermark image Embedding the diagonal high-frequency coefficients of the host image obtained in step 1 In the process, the separated channels are merged to obtain a watermarked image; Step 3: Print the watermarked image obtained in Step 2, and then scan it in RGB color mode to obtain the scanned copy; Step 4: When the embedded channel is the B channel, extract the watermark from the B channel of the RGB color mode of the scanned document obtained in Step 3; when the embedded channel is the C channel, extract the watermark from the R channel of the RGB color mode of the scanned document obtained in Step 3.

[0007] In the above technical solution, in step 1, the embedded channel is subjected to a second-order Discrete Wavelet Transformation (DWT) to extract the diagonal high-frequency coefficients of the host image. ; In the above technical solution, step 1, specifically the process of performing a two-level discrete wavelet transform on the embedded channel, involves first performing a single-level transform to obtain the low-frequency subband, and then performing a second-level transform to extract the diagonal high-frequency coefficients of the host image. .

[0008] In the above technical solution, in step 2, the watermark to be embedded is encrypted using Borg coding, and a holographic watermark image is obtained through a holographic watermarking algorithm. The holographic watermark image is then subjected to a discrete wavelet transform to extract the diagonal high-frequency coefficients of the holographic watermark image. .

[0009] In the above technical solution, in step 2, the separated channels are merged after discrete wavelet inverse transform.

[0010] In the above technical solution, in step 2, the diagonal high-frequency coefficients of the holographic watermark image are transformed by substitution embedding. Embedding the diagonal high-frequency coefficients of the host image obtained in step 1 .

[0011] In the above technical solution, the equation for the substitution embedding is:

[0012] in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. This indicates the strength of the watermark embedding.

[0013] In the above technical solution, in step 3, the scanning resolution is above 300ppi and not greater than the optical resolution of the scanner.

[0014] In the above technical solution, in step 2, the scanned document needs to be cropped to obtain a borderless scanned image with watermark, and then the watermark needs to be extracted.

[0015] In the above technical solution, step 4, extracting the watermark from the scanned document requires a discrete wavelet two-level transform. The first-level transform yields the low-frequency sub-band, and then the second-level transform extracts the diagonal high-frequency coefficients of the scanned document.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention proposes a method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images, breaking away from the default method of extracting watermarks from the same channel as the embedding channel. The proposed method allows the watermark extraction position to be the same as the embedding position, or they may be different. Instead, regardless of whether the watermark is embedded in RGB or CMYK color mode, it is always extracted in RGB color mode, resulting in better extraction performance. Attached Figure Description

[0017] Figure 1 This is a flowchart of the selection method of the present invention.

[0018] Figure 2 The image is an embodiment of the present invention, where a is the host image and b is the watermark to be embedded.

[0019] Figure 3 The images are watermarked images of the present invention, wherein a is the watermarked image of Example 3 and b is the watermarked image of Example 1.

[0020] Figure 4 This is a comparative flowchart of the present invention.

[0021] Figure 5 The watermarks and Nc values ​​extracted from each channel of Cases 1-7 and Reference Example 1 of the present invention are shown.

[0022] Figure 6 The watermark images and Nc values ​​extracted from each channel of Cases 8-14 and Reference Example 2 of the present invention are shown. Detailed Implementation

[0023] The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0024] Reference example 1 The watermark to be embedded is inserted into the C channel of the host image (CMYK color mode) to obtain a watermarked image. The watermark is extracted directly from the C, M, Y, and K channels without printing or scanning. (Refer to Group 1) Example 1 like Figure 1 As shown, a method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images includes the following steps: Step 1, select a host image in CMYK color mode (e.g., Figure 2 As shown in 'a' (the standard test image), the ICC profile file is set to "Japan Color 2001 Coated". The CMYK color mode is then separated into four channels: C, M, Y, and K. Channel C is selected as the embedding channel. A single-level discrete wavelet transform is applied to the embedding channel to obtain the low-frequency subband. A second-level transform is then used to extract the diagonal high-frequency coefficients of the host image. ; Step 2, use Borg coding to embed the watermark (binary image, such as...) Figure 2 (as shown in b) is encrypted, and a holographic watermark image is obtained through a holographic watermarking algorithm. The holographic watermark image is then subjected to a discrete wavelet transform to extract the diagonal high-frequency coefficients of the holographic watermark image. Substitution embedding step 1 obtained diagonal high frequency coefficients of the host image In the process, after discrete wavelet inverse transform, the separated channels are merged to obtain a watermarked image (such as...). Figure 3 (as shown in b); where the equation for the substitution embedding is:

[0025] in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. In this embodiment, the strength of the watermark embedding is indicated. =1.

[0026] Step 3: Print the watermarked image obtained in Step 2 using a KONICA MINOLTA C4080Series printer. The substrate is 157g coated paper. Scan the printed image in RGB color space using a Microtek Phantom 9900XL scanner. The scanning parameters are as follows: RGB color space, scanning resolution of 1200ppi, scaling ratio of 100%. The scanned document is then cropped to obtain a borderless watermarked scanned image. Step 4: Extract the watermark from the watermarked scan image obtained in Step 3 in the R channel of the RGB color space. For comparison, the watermark is also extracted from the G and B channels. (Case 1) Example 2 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images includes the following steps: Step 1: Select the ICC profile file of the host image in CMYK color mode as "JapanColor 2001 Coated". Then, separate the CMYK color mode into four channels: C, M, Y, and K. Select the C channel as the embedding channel. Perform a discrete wavelet transform on the embedding channel to obtain the low-frequency subband, and then perform a second-level transform to extract the diagonal high-frequency coefficients of the host image. , wherein the host image; Step 2: The watermark to be embedded (binary image) is encrypted using Bocchi coding, and a holographic watermark image is obtained using a holographic watermarking algorithm. The holographic watermark image undergoes a discrete wavelet transform to extract the diagonal high-frequency coefficients. Substitution embedding step 1 obtained diagonal high frequency coefficients of the host image In the process, after discrete wavelet inverse transform, the separated channels are merged to obtain a watermarked image; wherein, the equation for the substitution embedding is:

[0027] in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. In this embodiment, the strength of the watermark embedding is indicated. =1.

[0028] Step 3: Print the watermarked image obtained in Step 2 using a KONICA MINOLTA C4080Series printer on 157g coated paper. Create an ICC profile file for the Phantom 9900XL scanner, named "Phantom 9900XL_2412". Then, use the scanner to scan the printed image in RGB color space. The scanning parameters are as follows: RGB color space, scanning resolution of 1200ppi, scaling ratio of 100%. Obtain the scanned copy, crop the scanned copy, and obtain a borderless watermarked scanned image. Step 4: Extract the watermark from the watermarked scan image obtained in Step 3 in the R channel of the RGB color space. For comparison, the watermark is also extracted from the G and B channels. (Case 4) Comparative Example 1 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images. This method is the same as steps 1-2 in Embodiment 1, but steps 3 and 4 have the following differences: Case 2: Step 3 involves scanning in the CMYK color space; Step 4 involves extracting the watermark from the scanned image with the watermark obtained in Step 3 in the four channels C, M, Y, and K of the CMYK color space.

[0029] Case 3: Step 3 involves scanning in the LAB color space; Step 4: Extract the watermark from the watermarked scan image obtained in Step 3 in the L, A, and B channels of the LAB color space.

[0030] Case 5: Step 3 involves scanning in the CMYK color space; Step 4 involves taking the watermarked scanned image obtained in Step 3 and, in Adobe Photoshop CC, clicking "Edit" and then "Specify Profile" in the menu bar, specifying "Japan Color 2001Coated" as the profile file, and then extracting the watermark from the C, M, Y, and K channels respectively. Case 6: Step 3 involves scanning in the LAB color space; Step 4 involves clicking "Edit" and then "Specify Profile" in the Adobe Photoshop CC software to convert the watermarked scanned image obtained in Step 3 into a profile file named "Japan Color 2001 Coated" for the host image, and extracting the watermark from the C, M, Y, and K channels respectively. Comparative Example 2 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images, which is the same as steps 1-3 in Embodiment 2, except for the following difference in step 4: Case 7: Step 4, in Adobe Photoshop CC, click "Edit" and "Specify Profile" in the menu bar to convert the watermarked scanned image obtained in Step 3 into a profile file of the host image, "Japan Color 2001Coated", and extract the watermark from the C, M, Y, and K channels respectively.

[0031] The normalized correlation coefficient (Nc) was used to evaluate the similarity between the watermarks extracted in Example 1, Example 2, Comparative Example 1, and Comparative Example 2 and the original watermark of the reference group. The specific formula is as follows, and the calculation results are as follows. Figure 5 .

[0032] in, For embedding watermarks (original watermarks). The watermarks extracted for Example 1, Example 2, Comparative Example 1 and Comparative Example 2 have Nc values ​​between 0 and 1. Theoretically, the closer the Nc value is to 1, the higher the similarity and the better the robustness of the watermark.

[0033] Depend on Figures 4-5 It is known that embedding a watermark in the C channel of the CMYK color space results in optimal watermark extraction in the R channel of the RGB color space, which is non-original channel extraction. Furthermore, configuring the profile file directly after scanning does not affect the watermark quality; however, if color space conversion is involved, creating a profile file for the scanner is necessary (case 7 is better than case 6).

[0034] Reference example 2 The watermark to be embedded is inserted into the B channel of the host image (RGB color mode) to obtain a watermarked image. The watermark is extracted directly from the R, G, and B channels without printing or scanning. (Ref. Group 2) Example 3 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images includes the following steps: Step 1: Select the ICC profile file of the host image in RGB color mode as "sRGBICE61966-2.1". Then, separate the RGB color space into three channels: R, G, and B. Select the B channel as the embedding channel. Perform a discrete wavelet transform on the embedding channel to obtain the low-frequency subband, and then perform a second-level transform to extract the diagonal high-frequency coefficients of the host image. ; Step 2: The watermark to be embedded (binary image) is encrypted using Bocchi coding, and a holographic watermark image is obtained using a holographic watermarking algorithm. The holographic watermark image undergoes a discrete wavelet transform to extract the diagonal high-frequency coefficients. Substitution embedding step 1 obtained diagonal high frequency coefficients of the host image In the process, after discrete wavelet inverse transform, the separated channels are merged to obtain a watermarked image (such as...). Figure 3 (as shown in a); where the equation for the substitution embedding is:

[0035] in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. In this embodiment, the strength of the watermark embedding is indicated. =1.

[0036] Step 3: Print the watermarked image obtained in Step 2 using a KONICA MINOLTA C4080Series printer. The substrate is 157g coated paper. Scan the printed image in RGB color space using a Microtek Phantom 9900XL scanner. The scanning parameters are as follows: RGB color space, scanning resolution of 1200ppi, scaling ratio of 100%. The scanned document is then cropped to obtain a borderless watermarked scanned image. Step 4: Extract the watermark from the watermarked scan image obtained in Step 3 in the B channel of the RGB color space. For comparison, the watermark is also extracted from the G and R channels. (Case 8) Example 4 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images includes the following steps: Step 1: Select the ICC profile file of the host image in RGB color mode as "sRGBICE61966-2.1". Then, separate the RGB color space into three channels: R, G, and B. Select the B channel as the embedding channel. Perform a discrete wavelet transform on the embedding channel to obtain the low-frequency subband, and then perform a second-level transform to extract the diagonal high-frequency coefficients of the host image. ; Step 2: The watermark to be embedded (binary image) is encrypted using Bocchi coding, and a holographic watermark image is obtained using a holographic watermarking algorithm. The holographic watermark image undergoes a discrete wavelet transform to extract the diagonal high-frequency coefficients. Substitution embedding step 1 obtained diagonal high frequency coefficients of the host image In the process, after discrete wavelet inverse transform, the separated channels are merged to obtain a watermarked image; wherein, the equation for the substitution embedding is:

[0037] in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. In this embodiment, the strength of the watermark embedding is indicated. =1.

[0038] Step 3: Print the watermarked image obtained in Step 2 using a KONICA MINOLTA C4080Series printer on 157g coated paper. Create an ICC profile file for the Phantom 9900XL scanner, named "Phantom 9900XL_2412". Then, use the scanner to scan the printed image in RGB color space. The scanning parameters are as follows: RGB color space, scanning resolution of 1200ppi, scaling ratio of 100%. Obtain the scanned copy, crop the scanned copy, and obtain a borderless watermarked scanned image. Step 4: Extract the watermark from the watermarked scan image obtained in Step 3 in the B channel of the RGB color space. For comparison, the watermark is also extracted from the G and R channels. (Case 11) Example 5 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images includes the following steps: Step 1: Select the ICC profile (profile file is "sRGBICE61966-2.1") of the host image in RGB color mode. Then, separate the RGB color space into three channels: R, G, and B. Select the B channel as the embedding channel. Perform a discrete wavelet transform on the embedding channel to obtain the low-frequency subband. Then, perform a second-level transform to extract the diagonal high-frequency coefficients of the host image. ; Step 2: The watermark to be embedded (binary image) is encrypted using Bocchi coding, and a holographic watermark image is obtained using a holographic watermarking algorithm. The holographic watermark image undergoes a discrete wavelet transform to extract the diagonal high-frequency coefficients. Substitution embedding step 1 obtained diagonal high frequency coefficients of the host image In the process, after discrete wavelet inverse transform, the separated channels are merged to obtain a watermarked image; wherein, the equation for the substitution embedding is:

[0039] in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. In this embodiment, the strength of the watermark embedding is indicated. =1.

[0040] Step 3: Print the watermarked image obtained in Step 2 using a KONICA MINOLTA C4080Series printer on 157g coated paper. Create an ICC profile file for the Phantom 9900XL scanner, named "Phantom 9900XL_2412". Then, use the scanner to scan the printed image in RGB color space. The scanning parameters are as follows: RGB color space, scanning resolution of 1200ppi, scaling ratio of 100%. Obtain the scanned copy, crop the scanned copy, and obtain a borderless watermarked scanned image. Step 4: In Adobe Photoshop CC, click "Edit" and then "Assign Profile" in the menu bar to convert the watermarked scanned image obtained in Step 3 into a profile file named "sRGB ICE61966-2.1" for the host image. Extract the watermark from the B channel. For comparison, the watermark from the G and R channels is also extracted. (Case 14) Comparative Example 3 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images. This method is the same as steps 1-2 in Example 3, but steps 3 and 4 have the following differences: Case 9: Step 3 involves scanning in the CMYK color space; Step 4 involves extracting the watermark from the watermarked scan image obtained in Step 3 in the four channels C, M, Y, and K of the CMYK color space.

[0041] Case 10: Step 3 involves scanning in the LAB color space; Step 4 involves extracting the watermark from the watermarked scan image obtained in Step 3 in the L, A, and B channels of the LAB color space.

[0042] Case 12: Step 3 involves scanning in the CMYK color space; Step 4: In Adobe Photoshop CC, take the watermarked scanned image obtained in Step 3, click "Edit" in the menu bar, then "Specify Profile," specify "Japan Color 2001Coated" as the profile file, and extract the watermark from the C, M, Y, and K channels respectively. Comparative Example 4 A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images, which is the same as steps 1-3 in Embodiment 2, except for the following difference in step 4: Case 7: Step 4, in Adobe Photoshop CC, click "Edit" and "Specify Profile" in the menu bar to convert the watermarked scanned image obtained in Step 3 into a profile file of the host image, specify "Japan Color2001 Coated", and extract the watermark from the C, M, Y, and K channels respectively.

[0043] The similarity between the watermarks extracted in Examples 1-5, Comparative Examples 1-4, and Reference Examples 1-2 and the watermark to be embedded (the original watermark) was evaluated using the normalized correlation coefficient (Nc).

[0044] in, For the watermark to be embedded, For the extracted watermark, the Nc value is between 0 and 1. Theoretically, the closer the Nc value is to 1, the higher the similarity and the better the robustness of the watermark.

[0045] If the calculated Nc value is 5 and Figure 6 As shown, by Figure 5 and Figure 6 It can be seen that the watermarks extracted in Examples 3-5, Comparative Examples 3 and 4 are embedded in the B channel of the RGB color space. Therefore, extracting the watermark in the B channel of the RGB color space is optimal, i.e., the Nc value is the highest. Furthermore, configuring the profile file directly after scanning does not affect the watermark quality; if color space conversion is involved, creating a profile file for the scanner is necessary (case 14 is better than case 13).

[0046] The above description is only a preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for selecting channels for embedding and extracting anti-counterfeiting watermarks in printed images, characterized in that, Includes the following steps: Step 1: Separate the host image into channels, select one channel as the embedding channel, and extract the diagonal high-frequency coefficients of the host image in the embedding channel. If the host image is in RGB color mode, it is separated into three channels: R, G, and B, and the B channel is selected as the embedding channel; if the host image is in CMYK color mode, it is separated into four channels: C, M, Y, and K, and the C channel is selected as the embedding channel. Step 2: Encrypt the watermark to be embedded to obtain a holographic watermark image, and extract the diagonal high-frequency coefficients of the holographic watermark image. The high frequency coefficients of the diagonal of the holographic watermark image Embedding the diagonal high-frequency coefficients of the host image obtained in step 1 In the process, the separated channels are merged to obtain a watermarked image; Step 3: Print the watermarked image obtained in Step 2, and then scan it in RGB color mode to obtain the scanned copy; Step 4: When the embedded channel is the B channel, extract the watermark from the B channel of the RGB color mode of the scanned document obtained in Step 3; when the embedded channel is the C channel, extract the watermark from the R channel of the RGB color mode of the scanned document obtained in Step 3.

2. The selection method according to claim 1, characterized in that, In step 1, the embedded channel is subjected to a discrete wavelet second-order transform to extract the diagonal high-frequency coefficients of the host image. .

3. The selection method according to claim 1, characterized in that, In step 1, the specific process of performing a discrete wavelet two-level transform on the embedded channel is as follows: first, a low-frequency sub-band is obtained through a single-level transform, and then the high-frequency diagonal coefficients of the host image are extracted through a second-level transform. .

4. The selection method according to claim 1, characterized in that, In step 2, the watermark to be embedded is encrypted using Borge coding, and a holographic watermark image is obtained using a holographic watermarking algorithm. The holographic watermark image is then subjected to a discrete wavelet transform to extract the diagonal high-frequency coefficients of the holographic watermark image. .

5. The selection method according to claim 1, characterized in that, In step 2, the separated channels are merged after discrete wavelet inverse transform.

6. The selection method according to claim 1, characterized in that, In step 2, the high-frequency coefficients of the holographic watermark image are transformed by substitution embedding. Embedding the diagonal high-frequency coefficients of the host image obtained in step 1 .

7. The selection method according to claim 6, characterized in that, The equation for the substitution embedding is: in, These are the diagonal high-frequency coefficients of the watermarked channel after a two-layer wavelet transform. The diagonal high-frequency coefficients of the wavelet transform layer of the holographic watermark image are given. This indicates the strength of the watermark embedding.

8. The selection method according to claim 1, characterized in that, In step 3, the scanning resolution is 300ppi or higher and not greater than the optical resolution of the scanner.

9. The selection method according to claim 1, characterized in that, In step 4, watermark extraction of the scanned document requires a discrete wavelet two-level transform. The first-level transform yields the low-frequency subband, and the second-level transform extracts the diagonal high-frequency coefficients of the scanned document.

Images