CN111242825A - A zero-watermark method for ENC electronic nautical charts based on bathymetric features - Google Patents

Abstract

The zero-watermark method of ENC electronic navigation chart based on water depth feature, including zero-watermark construction and zero-watermark extraction test method, not only overcomes the certain data (spatial) disturbance caused by the reversible watermark algorithm of spatial characteristics to the electronic chart data, but also improves the attribute characteristics The resistance to precision reduction of lossless watermarking algorithms.

Description

A zero-watermark method for ENC electronic nautical charts based on water depth features

technical field

The invention relates to the technical field of marine surveying and mapping.

Background technique

The ENC electronic nautical chart zero watermark technology belongs to a kind of digital watermarking technology within a special application range, which has a wide range of applications in the fields of ENC electronic nautical chart copyright protection or content authentication.

At present, the research on digital watermarking algorithm for two-dimensional electronic chart mainly includes reversible watermarking algorithm based on spatial feature, lossless watermarking algorithm based on attribute feature and digital watermarking algorithm based on semantics. In the reversible watermarking algorithm based on spatial features, the electronic chart is essentially evolved from discrete vector nodes through continuous changes. By analyzing the evolution process of the chart, the embedding nodes of the watermark can be obtained, and then the embedding of the reversible watermarking algorithm can be realized. , in order to avoid excessive data disturbance to the electronic chart, the embedding process of the watermark needs to be kept within a certain disturbance range, and after the watermark information is successfully extracted, the similarity calculation of the digital watermark information is performed. The attribute feature-based lossless watermarking algorithm adopts the zero-bit dynamic lossless watermarking implementation method based on the gradient algorithm, uses the zero-bit extension technology for watermarking processing, and combines with the embedding point selection technology of the digital watermarking gradient algorithm, which has better watermarking performance. Invisibility and robustness. Semantic-based digital watermarking algorithm, the use of semantic watermarking is limited to legitimate users, and the propagation environment is controlled. On this basis, each user's authority is limited. Before opening the map data, users cannot initiate direct attacks on semantic watermarking. At the same time, the semantic watermark adopts the method of redundant embedding. Before the map is distributed, some redundant vertices are embedded into the map. Before the user uses the map, the permission watermark is extracted, and the redundant vertices are synchronously cleared, which does not interfere with the normal use of the map.

The above three algorithms still have many shortcomings: the reversible watermarking algorithm based on spatial features causes a certain range of data (spatial) disturbance to the electronic chart data; the lossless watermarking algorithm based on attribute features is difficult to add zero-bit watermarks in ENC electronic navigation charts, because The unit used for the position in the S57 specification is the latitude and longitude coordinates and is converted into an integer value by the multiplier factor COMF, and the precision reduction resistance, format conversion resistance and interpretation attack resistance are poor. The semantic-based digital watermarking algorithm focuses on the use of image matching technology. The digital watermarking of wider semantic information needs to be studied. At the same time, the chart update ER is the most common operation, and the correctability of ENC chart data is insufficient. In terms of experimental environment, the three algorithms all use MapInfo software, and the main form of electronic chart is SHP file.

SUMMARY OF THE INVENTION

In order to overcome the shortcomings of the above three methods, the present invention provides a zero-watermark technology for ENC electronic nautical charts based on water depth characteristics.

The technical scheme adopted by the present invention to achieve the above-mentioned purpose is: a zero-watermark method for ENC electronic nautical charts based on water depth characteristics, including a zero-watermark construction method, and the steps are as follows:

Step 1: Preprocess ENC electronic nautical chart to obtain sounding point data;

Step 2: Construct a Delaunay triangulation for the water depth data with zero watermark, first extract the aspect feature line by using the sound depth interpolation, and then use the extracted feature line as the skeleton line to judge the feature point of the aspect between the point pairs on the line and then extract the feature points;

Step 3: Select an appropriate threshold for the extracted feature points to perform point constraint segmentation, and then sort the sounding points in each block through ZigZag, and then obtain a binary sequence of latitude and longitude coordinates of the sounding point to ensure the length of the binary sequence of coordinates in each block. greater than the length of the watermark sequence;

Step 4: Arnold scramble the original watermark image to obtain the encrypted watermark image, and then obtain the watermark sequence through group transformation;

Step 5: Complement the binary sequence of latitude and longitude coordinates and the watermark sequence to obtain a zero-watermark image, that is, a zero-watermark; Step 6: Save the zero-watermark image and the original watermark image generated by each block;

Also included is a zero-watermark extraction inspection method, comprising the following steps:

Step 1: Preprocess the ENC electronic nautical chart to be detected to obtain sounding point data;

Step 2: Construct a Delaunay triangulation network for the sounding point data, and use the sounding interpolation to extract the slope aspect feature line to further judge and extract the sounding feature points;

Step 3: perform point constraint segmentation on the extracted feature points, and then sort the sounding points in each block by ZigZag, and then obtain a binary sequence of latitude and longitude coordinates of the sounding points;

Step 4: Grouping and transforming the watermark images to obtain a watermark sequence, performing a complementary XOR operation on the binary sequence of latitude and longitude coordinates and the watermark sequence, extracting several encrypted watermark images according to the size of the original watermark, and then performing Arnold inverse scrambling to obtain the data to be detected. watermark image;

Step 5: Compare and verify the plurality of extracted watermark images with the original watermark images, use the normalized correlation coefficient NC value to measure the similarity of the two images, and then determine the copyright ownership of the data to be detected.

In the third step of constructing the zero watermark, the unit used for the position in the ENC electronic nautical chart is a decimal degree converted into an integer type, the converted coordinate integer value is encoded in binary, and the integer value before the significant digits of precision is converted into binary.

In the step 4 and step 5 of the zero-watermark construction, the watermark is encrypted by the Arnold transformation method, and Arnold scrambling is used for image data with equal length and width, and the Arnold transformation formula is:

Among them, N is the length and width of the picture, X and Y are the horizontal and vertical coordinates of a point on the original image. After transformation, another point coordinate X', Y' is formed, and the transformed image is binarized, with The zero-watermark image is obtained by performing the complement XOR operation on the coordinate binary sequence after the block.

In the step 4 of the zero-watermark extraction and inspection, when the extracted scrambled image is restored, an inverse matrix is obtained from the forward permutation formula to obtain the inverse transformation formula of Arnold permutation, and the original image is directly formed by the inverse transformation formula, The inverse transform formula of Arnold transform is as follows:

In the described zero watermark extraction test step 5, whether the watermark data has copyright authentication is judged by comparing the similarity value with the set threshold, and the calculation formula of similarity (NC) is as follows:

where W(i,j) represents the pixel value at (i,j) in the original watermark data, W′(i,j) represents the pixel value at (i,j) of the watermark data extracted from the vector chart, m, n represents the row and column values of the watermark image data,

When NC is 1, it means that the extracted watermark image data is exactly the same as the original watermark image data, and the image has not changed after being attacked; when the NC value is between 0 and 1, the closer it is to 1, the higher the data similarity is. , the robustness of the watermark is strong; otherwise, the lower the similarity, the poorer the robustness of the watermark, where the NC value takes 0.7 as the threshold for identifying the watermark.

In the fifth step of the zero-watermark extraction test, the comparative verification is measured by the structural similarity SSIM, and the SSIM is:

SSIM(W,W')=l(W,W')c(W,W')s(W,W')#(4)

where W is the watermark image, W′ is the extracted watermark image,

The closer the value of SSIM is to 1, it can be said that the two images are more similar and the quality of the image is better.

The zero watermarking method for ENC electronic nautical charts based on water depth features of the present invention not only overcomes certain data (space) disturbances to electronic chart data caused by the reversible watermarking algorithm of spatial features, but also improves the accuracy reduction of the lossless watermarking algorithm of attribute features. resistance.

Description of drawings

FIG. 1 is a flow chart of the zero-watermark construction of the ENC electronic nautical chart of the present invention.

Fig. 2 is the flow chart of zero watermark extraction and inspection of ENC electronic nautical chart.

Figure 3 is a schematic diagram of the SuperMap software operating platform.

Figure 4(a) is a schematic diagram of the original watermark.

Figure 4(b) is a schematic diagram of the watermark after Arnold scrambles the watermark image once.

Figure 4(c) is a schematic diagram of the watermark after Arnold scrambles the watermark image twice.

Figure 4(d) is a schematic diagram of the watermark after Arnold scrambles the watermark image 10 times.

Figure 4(e) is a schematic diagram of the watermark after Arnold scrambles the watermark image 20 times.

Figure 4(f) is a schematic diagram of the watermark after Arnold scrambles the watermark image 30 times.

Detailed ways

The zero-watermark method for ENC electronic nautical charts based on the water depth feature of the present invention includes a zero-watermark structure and a zero-watermark extraction check, wherein the zero-watermark structure is shown in Figure 1 and includes the following steps:

Step 1: Preprocess ENC electronic nautical chart through SuperMap software to obtain sounding point data;

Step 2: Construct a Delaunay triangulation for the water depth data with zero watermark, first extract the aspect feature line by using the water depth interpolation, and then use the extracted feature line as the skeleton line to judge the feature points of the aspect between the points on the line and then extract the feature points. Aspect characteristic line is a conceptual line used to describe the relationship tree of water depth points, and it is a transitional virtual connection line that provides a basis for judgment of feature shallow point extraction. The water depth data in the upper closed isobaric area is classified as one type, so as to avoid the influence of the data outside the interval on the data in the interval. After constructing the triangulation network and extracting the characteristic line, the characteristic line is used as the skeleton line of the seabed topography to judge the characteristic of the slope aspect between the point pairs on the line. " to extract;

Step 3: Select an appropriate threshold for the extracted feature points to perform point constraint segmentation, and then sort the sounding points in each block through ZigZag, and then obtain a binary sequence of latitude and longitude coordinates of the sounding point to ensure the length of the binary sequence of coordinates in each block. Greater than the length of the watermark sequence, the unit used for the position in the ENC electronic nautical chart is the decimal degree converted into an integer, and the converted coordinate integer value is coded in binary. The integer value of the bit) is converted into binary, and the latitude and longitude binary sequence in each block is sorted and recorded by ZigZag in turn, to ensure that the length of the coordinate sequence in each block is greater than the length of the watermark sequence, and the latitude and longitude of a single sounding point is passed through the multiplier factor COMF(10 ⁷ ) Convert to an integer, within the allowable range of precision, the last one (two) bits can be discarded and then encoded in binary;

Step 4: Arnold scramble the original watermark image to obtain an encrypted watermark image, and then obtain a watermark sequence through grouping transformation, and use the Arnold transformation method to encrypt the watermark, as shown in Figure 4;

Step 5: Complement the binary sequence of latitude and longitude coordinates and the watermark sequence to obtain a zero-watermark image, that is, a zero-watermark. In order to simplify the calculation, Arnold scrambling is used for image data with equal length and width. Because the Arnold scrambling method is periodic, According to the different users of ENC electronic charts, the number of scrambling is mapped to different users and stored as a key. The Arnold transformation formula is:

Among them, N is the length and width of the picture, X and Y are the horizontal and vertical coordinates of a point on the original image, after transformation, another point coordinate X', Y' is formed, and the transformed image is binarized, and The zero-watermark image is obtained by complementing the binary sequence of coordinates after the block by the XOR operation. The larger N is, the larger the amount of watermark information is, and the longer the cycle period T is. many;

Step 6: Save the zero-watermark image generated by each block and the original watermark image to the Copyright Protection Center (IPR), and add a legally effective time stamp to resist interpretation attacks.

The zero-watermark extraction test is shown in Figure 2 and includes the following steps:

Step 1: Preprocess the ENC electronic nautical chart to be detected to obtain sounding point data;

Step 2: Construct a Delaunay triangulation network for the sounding point data, and use the sounding interpolation to extract the slope aspect feature line to further judge and extract the sounding feature points;

Step 3: perform point constraint segmentation on the extracted feature points, and then sort the sounding points in each block by ZigZag, and then obtain a binary sequence of latitude and longitude coordinates of the sounding points;

Step 4: Grouping and transforming the watermark images to obtain a watermark sequence, performing a complementary XOR operation on the binary sequence of latitude and longitude coordinates and the watermark sequence, extracting several encrypted watermark images according to the size of the original watermark, and then performing Arnold inverse scrambling to obtain the data to be detected. For the watermark image, when restoring the extracted scrambled image, if n times of replacement have been performed before, T-n iterations are required to restore the original image. Obviously, only one forward formula requires too much computation. The inverse matrix is obtained from the permutation formula, and the inverse transformation formula of Arnold permutation is obtained. Then the original image can be formed by directly performing the same number of transformations n through the inverse transformation formula. The inverse transformation formula of Arnold transformation is as follows:

Step 5: Compare and verify the multiple extracted watermark images with the original watermark images, use the normalized correlation coefficient NC (Normalized Correlation) value to measure the similarity of the two images, and then determine the copyright ownership of the data to be detected. The purpose of the detection is to test Robustness of the embedded digital watermark, after geometric transformation (translation, rotation, zoom), map cropping, data deletion and addition of vector charts, the normalized correlation coefficient between the extracted watermark data and the original watermark data is calculated (Normalized Correction, NC), by comparing the similarity value with the set threshold value to determine whether the watermark data has copyright authentication. The calculation formula of similarity (NC) is as follows:

where W(i,j) represents the pixel value at (i,j) in the original watermark data, W′(i,j) represents the pixel value at (i,j) of the watermark data extracted from the vector chart, m, n represents the row and column values of the watermark image data,

When NC is 1, it means that the extracted watermark image data is exactly the same as the original watermark image data, and the image has not changed after being attacked; when the NC value is between 0 and 1, the closer it is to 1, the higher the data similarity is. , the watermark has strong robustness; anyway, the lower the similarity, the poorer the watermark robustness. The NC value generally takes 0.7 as the threshold for identifying watermarks.

Comparative verification can also be measured by structural similarity (SSIM), which is defined as:

SSIM(W,W')=l(W,W')c(W,W')s(W,W')#(4)

where W is the watermark image, W′ is the extracted watermark image,

The closer the value of SSIM is to 1, it can be said that the two images are more similar and the quality of the image is better, in general, the value of SSIM is greater than 0.5, people can accept the quality of the image.

The present invention is described through the embodiments, and those skilled in the art will appreciate that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, in the teachings of this invention, these features and embodiments may be modified to adapt a particular situation and material without departing from the spirit and scope of the invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of the present application belong to the protection scope of the present invention.

Claims (7)

1. An ENC electronic nautical chart zero-watermarking method based on water depth features is characterized by comprising the following steps: the method comprises a zero watermark construction method, and comprises the following steps:

the method comprises the following steps: preprocessing the ENC electronic chart to obtain water depth data;

step two: constructing a Delaunay triangulation network for the water depth data constructing the zero watermark, extracting a slope characteristic line by utilizing water depth interpolation, and judging the slope characteristic between the point pairs on the line according to the extracted characteristic line as a skeleton line so as to extract characteristic points;

step three: selecting a proper threshold value for the extracted feature points to carry out point constraint blocking, and sequencing the water depth points in each block through ZigZag to further obtain a longitude and latitude coordinate binary sequence of the water depth points, so as to ensure that the length of the coordinate binary sequence in each block is greater than that of the watermark sequence;

step four: performing Arnold scrambling on an original watermark image to obtain an encrypted watermark image, and performing grouping conversion to obtain a watermark sequence;

step five: carrying out bit-filling XOR operation on the longitude and latitude coordinate binary sequence and the watermark sequence to obtain a zero watermark image, namely a zero watermark;

step six: and storing the zero watermark image generated by each block and the original watermark image.

2. The ENC electronic nautical chart zero-watermarking method based on the water depth feature of claim 1, wherein: the zero watermark extraction and inspection method comprises the following steps:

the method comprises the following steps: preprocessing an ENC electronic chart to be detected to obtain water depth data;

step two: constructing a Delaunay triangulation network for the water depth data, and further judging and extracting water depth feature points by extracting a slope characteristic line through water depth interpolation;

step three: performing point constraint blocking on the extracted feature points, and sequencing the water depth points in each block through ZigZag to obtain a binary sequence of longitude and latitude coordinates of the water depth points;

step four: grouping and converting the watermark images to obtain a watermark sequence, performing complementary XOR operation on the longitude and latitude coordinate binary sequence and the watermark sequence, extracting a plurality of encrypted watermark images according to the size of the original watermark, and performing Arnold reverse scrambling to obtain a watermark image of data to be detected;

step five: comparing and verifying the extracted watermark images with the original watermark image, and measuring the similarity of the two images by using the normalized correlation coefficient NC value so as to judge the copyright ownership of the data to be detected.

3. The ENC electronic nautical chart zero-watermarking method based on the water depth feature of claim 1, wherein: in the third step of zero watermark construction, the unit used in the position in the ENC electronic nautical chart is decimal degree converted into integer, the integer value of the converted coordinate is encoded by binary system, and the integer value before the precision effective bit is converted into binary system.

4. The ENC electronic nautical chart zero-watermarking method based on the water depth feature of claim 1, wherein: in the fourth and fifth zero watermark construction steps, the watermark is encrypted by adopting an Arnold conversion mode, picture data Arnold with equal length and width are scrambled, and the Arnold conversion formula is as follows:

and N is the length and width of the picture, X and Y are the horizontal and vertical coordinates of one point on the original image, another point coordinate X 'and Y' is formed through transformation, the transformed image is subjected to binarization processing, and the binary sequence of the coordinates after the transformation and the binary sequence of the coordinates after the block are subjected to complementary exclusive OR operation to obtain a zero watermark image.

5. The ENC electronic nautical chart zero-watermarking method based on the water depth feature of claim 2, wherein: in the fourth step of zero watermark extraction and inspection, when the extracted scrambled image is subjected to reduction processing, an inverse matrix is obtained for the forward replacement formula to obtain an inverse transformation formula of Arnold replacement, and the original image is directly formed through the inverse transformation formula, wherein the inverse transformation formula of Arnold transformation is as follows:

6. the ENC electronic nautical chart zero-watermarking method based on the water depth feature of claim 2, wherein: in the step five of zero watermark extraction and inspection, whether the watermark data has copyright authentication is judged by comparing the similarity value with a set threshold value, and the calculation formula of the similarity (NC) is as follows:

wherein W (i, j) represents the pixel value at (i, j) in the original watermark data, W' (i, j) represents the pixel value at (i, j) in the watermark data extracted from the vector chart, m, n represents the values of the row and column of the watermark picture data,

when NC is 1, the extracted watermark picture data is completely the same as the original watermark picture data, and the picture is not changed after being attacked; when the NC value is between 0 and 1, the closer to 1, the higher the data similarity is, and the watermark robustness is strong; and conversely, the lower the similarity is, the lower the watermark robustness is, wherein the NC value is 0.7 as the threshold value for identifying the watermark.

7. The ENC electronic nautical chart zero-watermarking method based on the water depth feature of claim 2, wherein: in the step five of zero watermark extraction and inspection, the comparison and verification are measured by the structural similarity SSIM, and the SSIM is as follows:

SSIM(W,W')＝l(W,W')c(W,W')s(W,W')#(4)

where W is the watermark image, W' is the extracted watermark image,

the closer the value of SSIM is to 1, the more similar the two images are and the better the quality of the images is.

Images