CN101751656B - A watermark embedding and extraction method and device - Google Patents

Abstract

The invention discloses a watermark embedding and extraction method and device, which are applied in the copyright management process of digital content. The method includes: determining the available punctuation area in the text document to be embedded according to the set area determination rule; Determine the original position of the punctuation in each of the available punctuation areas respectively according to the position determination rule set; for each available punctuation area, according to the original position of the punctuation and the corresponding code to be embedded, the position of the punctuation in it is performed Adjust to realize embedding the code to be embedded into the corresponding available punctuation area. When extracting the watermark, the corresponding rules are used to extract the embedded codes in each available punctuation area. The above method is easy to operate, the embedded watermark information has good concealment and high stability, and can obtain good visual effects at the same time.

Description

A watermark embedding and extraction method and device

technical field

The invention relates to the field of digital copyright management, in particular to a text digital watermark embedding and extraction method and device based on punctuation for digital text.

Background technique

With the acceleration of global information digitization, a large number of text materials are emerging, such as personal files, medical records, academic certificates, patent certificates, handwritten signatures, library collections, confidential documents, etc. are all common forms of text. The importance is self-evident. In addition, with the increasing popularity of e-commerce and e-government, it is increasingly urgent to track the piracy of electronic publications issued on the Internet, and to judge the authenticity and integrity of electronic letters, official documents or faxes. Therefore, it is particularly important to embed watermarks (that is, add additional information) in text materials to achieve property rights protection and information security maintenance.

There are two common carrier forms for the above-mentioned text materials: one is paper, and the other is electronic documents. There are also two ways to embed watermarks: explicit and implicit. Explicitly embedded watermark means that the added information is clearly visible to human eyes, such as semi-transparent image watermark, unit name printed on the background, additional barcode, etc. Implicit embedded watermark means that the added information is difficult for human eyes to detect. For example, various implicit image watermarks need to be recognized by specific instruments combined with corresponding software. In the combination of the above situations, it is more difficult to implicitly embed watermarks in paper documents, among which embedding in black and white binary monochrome paper documents (hereinafter referred to as binary paper) is the most difficult. Embedding information in paper documents, the most common way is display, such as adding a specific background image on the paper background, writing specific text (for example, the sample documents that banks show customers will write "sample" in some areas typeface); or use special paper as a carrier, etc. But implicitly embedding information on binary paper, such as the common "black and white", is difficult.

In the existing technologies, there are also some technologies for embedding digital watermarks in binary paper.

For example: In the technical literature "Brassil J, Low S, Maxemchuk N F. Copyright Protection for the Electronic Distribution of Text Documents. Proceedings of the IEEE, I999, 87(7): 1181-1196", it is disclosed that by modifying line spacing, text Adjust the typesetting format of the document by adjusting the spacing, so as to realize the embedding of information. This method is simple in principle and easy to implement. The main problem is that the contradiction between the stability of the embedded information and the visual effect is difficult to solve: if the change of line spacing and word spacing is insufficient, it is difficult to achieve accurate information extraction; if the accuracy of extraction is to be ensured, more A large amount of change is easy to be noticed by readers, and cannot achieve the "implicit" effect.

The patent application "Method for Encrypting Electronic Official Documents or Documents and Identifying Authenticity" (publication number CN1588351) discloses that ordinary Chinese characters are slightly deformed, which is difficult for human eyes to detect, but can be identified by hand or OCR technology. Slight deformation, so as to achieve the purpose of implicitly embedding information. This method embeds a large amount of information and has high stability, but it involves technologies such as font library, OCR, printout, specific electronic text format (for printing), data training, etc. The process is cumbersome, the workload is heavy, and the production cost is very high. .

In the patent application "Text Digital Watermarking Technology Based on Character Topology" (publication number CN1684115), it is disclosed that the topological structure of characters can be changed by changing the continuity relationship between the strokes that make up characters (strings), so that the strokes of variant characters The method of transforming the number of connected domains to embed information. The main disadvantage of this method is that the production cost is high, and the robustness of information detection is not ideal: the slight pollution of the paper may cause the original unconnected strokes to be connected, and copying can easily cause the original connected strokes to be disconnected, making the Connected domains change, making detection by the number of connected domains unstable.

The patent application "A Digital Watermark Embedding and Extraction Method and Device" (publication number CNI945622A) discloses that each character in the text is divided into regions, and the points in each region are flipped, causing the black pixels in the region to change, thereby embedding information . The production cost of embedding watermark in this method is high, and due to the limitation of visual characteristics, the flipped point can only be at the edge of the character stroke, which actually changes the thickness of the stroke, and the visual effect is not ideal.

It can be seen that the method of encrypting (embedding watermark) on documents in the prior art has the disadvantages of large amount of hidden information, heavy workload, cumbersome production, and high production cost; its stability is also relatively poor, and it is difficult to maintain in the process of copying. ; And it is easy to be noticed by readers, the hiding effect is not good, and the visual effect is not good.

Contents of the invention

Embodiments of the present invention provide a method and device for embedding and extracting watermarks, which are used to solve the problems of poor stability and poor concealment existing in embedding watermark information in text documents in the prior art.

A watermark embedding method, comprising:

Determine the available punctuation area in the text document to embed information according to the set area determination rule;

According to the set position determination rules, respectively determine the original position of the punctuation in each of the available punctuation areas;

For each available punctuation area, the position of the punctuation point is adjusted according to the original position of the punctuation point and the corresponding code to be embedded, so as to embed the code to be embedded into the corresponding available punctuation area.

A watermark extraction method, comprising:

Determine the available punctuation area in the text document with embedded information according to the same area determination rules as when embedding watermark information;

According to the same position determination rule as when embedding watermark information, respectively determine the position of the punctuation in each of the available punctuation areas;

According to the position of the punctuation, the embedded codes in each available punctuation area are respectively determined.

A watermark embedding device, comprising:

an area determination module, configured to determine the available punctuation area in the text document to be embedded with information according to the set area determination rule;

A position determination module, configured to determine the original position of the punctuation points in each of the available punctuation areas according to the set position determination rules;

The information embedding module is used to adjust the position of the punctuation in each available punctuation area according to the original position of the punctuation and the corresponding code to be embedded, so as to embed the code to be embedded into the corresponding available punctuation area.

A watermark extraction device, comprising:

An area determination module, configured to determine the available punctuation area in the text document in which the information has been embedded according to the same area determination rule as when embedding the watermark information;

A position determination module, configured to determine the position of the punctuation points in each of the available punctuation areas according to the same position determination rules as when embedding watermark information;

The code extraction module is used to respectively determine the embedded codes in each available punctuation area according to the positions of the punctuation marks.

The watermark embedding and extraction method and device provided by the embodiments of the present invention realize embedding the code to be embedded into the corresponding available punctuation area by selecting the available punctuation area and adjusting the punctuation position in each available punctuation area. When extracting the watermark, the corresponding rules are used to extract the embedded codes in each available punctuation area according to the adjusted punctuation position. The above method is easy to operate, and since the human eye is far less sensitive to changes in punctuation positions than to changes in character positions, large changes can be made, so that the embedded watermark information has high stability and good concealment, and at the same time can ensure good visuals.

Description of drawings

Fig. 1 is the flowchart of watermark embedding method in the embodiment of the present invention;

Fig. 2 is an example diagram of determining the available punctuation area in the document fragment in the embodiment of the present invention;

FIG. 3 is a schematic diagram of frequency band division of determined available punctuation areas in an embodiment of the present invention;

FIG. 4 is an example of a text fragment after embedding information in an available punctuation area in an embodiment of the present invention;

FIG. 5 is a flowchart of a watermark extraction method in an embodiment of the present invention;

6 is a schematic structural diagram of a watermark embedding device in an embodiment of the present invention;

Fig. 7 is a schematic structural diagram of a watermark extraction device in an embodiment of the present invention.

Detailed ways

The watermark embedding and extraction method provided by the embodiment of the present invention selects the available punctuation area in the text document to be embedded with watermark information according to the set area selection rules, determines the position of the punctuation point in each available punctuation area, and then adjusts each The position of the punctuation in each available punctuation area to achieve the purpose of embedding watermark information; when extracting, the same rule is still used to determine the available punctuation area in the text document with embedded watermark information and the position of the punctuation in each available punctuation area. , get the embedded watermark information according to the position of the punctuation.

The watermark embedding method provided by the embodiment of the present invention achieves the purpose of embedding watermark information by adjusting the position of the punctuation in the text document. Its flow chart is shown in Figure 1, and the execution steps are as follows:

S101: Search and determine the available punctuation area in the text document to embed information according to the set area selection rule.

First, use OCR to perform layout recognition and analysis on the text document to embed information, remove the features of non-text areas such as borders, table lines, images, and lace in the text document, and obtain pure text areas.

Carry out text segmentation and punctuation analysis on the plain text area, find out all available punctuation points, and determine the available punctuation area.

Available punctuation means that there is at least one other character before and after the index point, that is, the position of the punctuation satisfies the positional relationship of "other characters, punctuation, and other characters". For punctuation that does not meet this condition, it can be discarded. For example, when the punctuation is at the end of the text line, it does not meet the conditions because there are no other characters behind it, so such punctuation is discarded. Wherein, other characters include all other symbols except punctuation, such as Chinese, numbers, and letters.

According to the available punctuation and two other adjacent characters before and after, define the start boundary and the end boundary, and obtain the available punctuation area. Wherein, the starting boundary of the available punctuation area may include: the left boundary, the right boundary, the center of gravity position or the center position of the preceding character, and the ending boundary of the available punctuation area may include: the left boundary, the right boundary, the center of gravity position or the center of the following character location etc.

For example: in the text document shown in Figure 2, it is determined that the available punctuation includes: "," between "show, yes", "." between "Fig. Shou", " ," and other 7 available punctuation points. What the starting boundary of available punctuation area got was the left boundary of preceding character, and what ending boundary got was the right boundary of back character, thereby determine " showing that is " shown in Fig. 2, " figure. First ", " first, "Text" and other 7 available punctuation areas.

S102: Determine the original position of the punctuation point in the available punctuation area according to the set position determination rule. Specifically include:

(1) Divide each available punctuation area into several frequency bands.

According to the set boundary rules, calculate the length of each available punctuation area, that is, the distance from the start boundary to the end boundary. According to the calculated distance, each available punctuation area is equally divided into several parts, and each part is a frequency band. For example, if it is divided into k parts, then from the first frequency band to the last frequency band, the corresponding frequency band indexes are 0, 1, 2, ...k-1 respectively.

Following the above example, the 7 available punctuation areas determined in Figure 2 are all divided into frequency bands, for example, each available punctuation area is divided into 16 frequency bands, as shown in Figure 3 after division.

(2) According to the coordinate positions of the punctuation points in each available punctuation area, respectively determine the frequency band where the punctuation points are in each available punctuation area and the corresponding frequency band index.

The position of the punctuation can be represented by position parameters such as the center of gravity, center, left boundary or right boundary of the punctuation, and the frequency band where the punctuation is located can be determined according to the position of the center of gravity, center, left boundary or right boundary of the punctuation, and the corresponding frequency band index can be determined .

When determining the frequency band where the punctuation is located according to the position of the center of gravity of the punctuation point, it is necessary to first calculate the coordinate position where the center of gravity of the punctuation point is located. The formula for calculating the barycentric coordinates of punctuation points is as follows:

${\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; C</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; S</span>}}$

Wherein, ΔS _i represents the number of black pixels whose horizontal coordinates are x _i included in the punctuation point;

x _i represents any horizontal coordinate value.

S is the sum of the black pixel points of punctuation;

x _C is the barycentric coordinates of the punctuation point.

Then, the frequency band where the punctuation point is located is obtained according to the frequency band where the calculated coordinate position is located. That is, according to the coordinate range of each frequency band, determine which frequency band the center of gravity coordinate x _C falls into, and the corresponding frequency band index.

For example: for the first area shown in FIG. 3 "show yes", after calculating the barycentric coordinates of the comma in the horizontal direction, the frequency band where the comma is located is determined as the frequency band whose frequency band index is 6 according to the barycentric coordinates.

In particular, for a horizontal text document, the horizontal center of gravity is calculated, while for a vertical text document, the vertical center of gravity needs to be calculated. The formula for calculating the vertical center of gravity can be obtained by analogy with the formula for calculating the horizontal center of gravity.

When it is necessary to calculate based on the center, there are also corresponding formulas, which will not be listed here.

It should be noted that many punctuation marks have a trailing end, for example commas have a sharper end. At this time, if the paper is scanned into an image by a device such as a scanner, the sharper tail is likely to disappear or break, that is, the tip of the tail is missing. At this time, if the center of the punctuation or the left and right boundaries are used to represent the position of the punctuation, errors will occur. If the center of gravity of the punctuation is used to represent the position of the punctuation point, the change of the center of gravity of the entire punctuation due to the absence or addition of individual points is very small. In the actual calculation process, after the value is rounded, the coordinate value of the center of gravity obtained will basically not occur. Change, so the position of the center of gravity of the punctuation is used to represent the position of the punctuation, which is the best choice for better stability.

S103: Determine the code to be embedded for each available punctuation area according to the set information embedding rules. Among them, the information embedding rules can be set and selected arbitrarily. Specifically include:

First, the binary number to be embedded is determined according to the information to be embedded corresponding to the text document; wherein the determined number of digits of the binary number to be embedded is less than or equal to the determined number of available punctuation regions.

If the information to be embedded is a binary number, and its number of digits is less than or equal to the determined number of available punctuation areas, then directly determine that the embedded information is a binary number to be embedded.

If the information to be embedded is a binary number, but its number of digits is greater than the determined number of available punctuation areas, then select a binary number whose number of digits is less than or equal to the determined number of available punctuation areas as the binary number to be embedded, and establish The corresponding relationship between the selected binary number and the information to be embedded is saved and recorded, for example, in a database.

If the information to be embedded is not a binary number itself, but can be converted into a binary number through base conversion, and the number of converted binary digits is less than or equal to the determined number of available punctuation areas, then it is determined that the converted binary number is the binary number to be embedded number.

If the information to be embedded is not a binary number itself, but can be converted into a binary number through base conversion, and the converted binary digits are greater than the determined number of available punctuation areas, then select a number that is less than or equal to the determined available punctuation areas Quantity of binary numbers, as the binary numbers to be embedded, and establish the corresponding relationship between the selected binary numbers and the information to be embedded, and record the information to be embedded and the corresponding relationship with the selected binary numbers, such as saving in the database.

If the information to be embedded is not a binary number and cannot be converted into a binary number; then select a binary number whose number of digits is less than or equal to the determined number of available punctuation areas as the binary number to be embedded, and establish the selected binary number and The correspondence between the information to be embedded and the correspondence between the information to be embedded and the selected binary number are recorded, for example, stored in a database.

Then, according to the binary number to be embedded and the set information embedding rules, the code to be embedded in each available punctuation area is determined. Specifically, according to the number of digits of the binary number to be embedded and the number of available punctuation areas, the code to be embedded in each available punctuation area is determined according to the set information embedding rules.

If the number of digits of the binary number to be embedded is equal to the number of available punctuation areas, the allocation is performed directly, that is, the binary codes contained in the binary number to be embedded are directly assigned to each available punctuation area as the code to be embedded.

If the number of binary digits to be embedded is less than the number of available punctuation areas, each available punctuation area is assigned a binary code contained in the binary number to be embedded as the code to be embedded for each available punctuation area through a redundancy algorithm. If the number of available punctuation areas is M, the number of digits of the binary number is N, and M<N; then calculate M/N to obtain the quotient and the remainder. The available punctuation area corresponding to the remainder is discarded, and then the code to be embedded is assigned to the remaining available punctuation area. For example, if the obtained quotient is 3, then the first binary code contained in the binary number is assigned to the 1st-3rd available punctuation area As the code to be embedded, assign the second binary code contained in the binary number to the 4th to 6th available punctuation area as the code to be embedded, ..., and so on.

Following the above example, if the number 123 is embedded in a text document as the information to be embedded, then convert 123 into a binary number: 1111011. Since the binary number has 7 digits, the determined available punctuation area is also 7 , so directly assign 1111011 to the 7 available punctuation areas shown in Figure 2, wherein the area "shown, yes" assigns the first binary code 1 of the binary number, and the area "Figure. The first" assigns the first bit of the binary number Two-bit binary code 1, ..., and so on.

For example: use the number 345 as the information to be embedded, then convert 345 into a binary number: 101011001, since the number of digits of the binary number is 9, and the determined available punctuation area is also 7, so a digit is randomly selected A binary number less than or equal to the number of available punctuation fields. Here, a 7-bit binary number can be selected or a binary number smaller than 7 bits can be selected. The corresponding relationship between the selected binary number and 345 the information to be embedded is established, and the information to be embedded 345 and the selected information are stored.

In particular, when the code to be embedded corresponding to each available punctuation area is known or has been randomly set, step S103 may not be omitted, and step S104 may be directly entered after step S102 is executed.

S104: Adjust the position of the punctuation in each available punctuation area according to the original position of the punctuation point in each available punctuation area and the corresponding code to be embedded. The code to be embedded corresponding to each available punctuation area can be embedded in each available punctuation area by adjusting the position of the punctuation mark.

Wherein, the code to be embedded corresponding to each available punctuation area includes 1 or 0, according to the parity of the code to be embedded and the parity of the frequency band index of the frequency band where the punctuation is located (referred to as index parity), the position of the punctuation is adjusted, so that Its position changes to achieve the purpose of embedding information. Specifically:

(i) If the frequency band index corresponding to the original position of the punctuation point in a certain available punctuation area is an odd number, and the code to be embedded corresponding to the available punctuation area is 0, then move the position of the punctuation point in the available punctuation area to the frequency band index that is an even number Frequency band, that is, move the position of the punctuation so that its frequency band index becomes an even number.

(ii) If the frequency band index corresponding to the original position of the punctuation point in an available punctuation area is an odd number, and the code to be embedded corresponding to the available punctuation area is 1, then the frequency band where the punctuation point is located in the available punctuation area is not changed, that is, the punctuation point is located The parity of the band index is unchanged. At this time, in order to improve the stability of the embedded information, the position of the punctuation should also be moved according to the situation, see the following description for details.

(iii) If the frequency band index corresponding to the original position of the punctuation point in a certain available punctuation area is an even number, and the code to be embedded corresponding to the available punctuation area is 0, then the frequency band where the punctuation point is located in the available punctuation area is not changed, that is, the punctuation point is located The parity of the band index is unchanged. At this time, in order to improve the stability of the embedded information, the position of the punctuation should also be moved according to the situation, see the following description for details.

(iv) If the frequency band index corresponding to the original position of the punctuation point in an available punctuation area is an even number, and the code to be embedded corresponding to the available punctuation area is 1, then move the position of the available punctuation area punctuation to the frequency band whose frequency band index is an odd number , that is, move the position of the punctuation so that its band index becomes even.

After all the available punctuation areas that need to be embedded with codes to be embedded are subjected to the above processing, the watermark information embedding of the text document is completed.

Using the above example, when the number of information to be embedded is 123, the code to be embedded corresponding to the first available punctuation area "show, yes" is 1, and the frequency band index 6 of the punctuation in this area is an even number. Article (iv) of the rule needs to be moved to the frequency band whose frequency band index is an odd number, then the punctuation in the available punctuation area is moved to the frequency band whose adjacent frequency band index is 5 or 7, in this embodiment is Move to the frequency band whose frequency band index is 5 as an example. Similarly, move the punctuation points in the other 6 areas according to the above rules, and the positions of the punctuation points in each available punctuation area after the movement are shown in Figure 4.

In the above two cases (ii) and (iii), it is not necessary to change the parity of the frequency band where the punctuation is located, but if the punctuation is just near the edge of the frequency band, it is very unstable, and it is likely to be caused by various interferences (such as printing, scanning, etc.) ) causes the parity flip of its band index. For example, the punctuation in an available punctuation area is in the third frequency band, and the specific value of its location is 3.01. Due to some kind of interference, it becomes 2.99, and the parity of its frequency band index changes from an odd number of 3 to 2. , an error will occur during subsequent checks. Therefore, adjusting the position of the punctuation point that does not need parity flipping of the frequency band index to the middle of the frequency band where it is located can greatly increase the stability. For example, the specific value of the above-mentioned punctuation in the third frequency band is 3.01, which can be adjusted to 3.5, which can effectively increase its stability.

Similarly, for the punctuation that needs to change the parity of the frequency band index of the frequency band, it should also be moved to the middle position of the frequency band to be adjusted, so that the stability of the position is better. For example, in the above available punctuation area "show, yes", when its index is adjusted from 6 to 5, the specific value of its position is 5.5.

The watermark extraction method provided by the embodiment of the present invention is used to extract the watermark information embedded by the above-mentioned watermark embedding method, the flow chart of which is shown in Figure 5, and the execution steps are as follows:

S201: Find and determine the available punctuation area in the text document in which the information has been embedded according to the same area selection rule as used for embedding the watermark information.

The process of determining the available punctuation area in this step is specifically the same as step S101, and will not be repeated here. The difference is that in step S101, the region selection rule can be selected arbitrarily, that is, the start boundary and the end boundary of the available punctuation area can be defined, and the boundary selection rule adopted when the processed text document must be embedded with watermark information must be adopted in this step, Use the same start and end boundaries for the usable punctuation area as when embedding watermark information.

Following the above example, seven available punctuation areas in the text segment shown in FIG. 4 can be determined.

S202: Determine the positions of the punctuation points in each available punctuation area according to the same position determination rules as when embedding the watermark information.

The process of determining the location of the punctuation in the available punctuation area is specifically the same as step S102.

The difference is: when determining the position of the punctuation, it is necessary to use the same position parameters (center of gravity, center, left or right boundary of the punctuation, etc.) position, the center of gravity must also be used to represent the position of the punctuation in this step. In order to ensure the accuracy of the frequency band where the determined punctuation points are located and the corresponding frequency band index.

Using the above example, for the first available punctuation area "show, yes" to determine the frequency band where the punctuation is located is 5 or 7.

S203: Determine the embedded codes in each available punctuation area according to the positions of the punctuation points in each available punctuation area. Specifically:

According to the parity of the frequency band index corresponding to the frequency band where the punctuation is located in the available punctuation area, the embedded code embedded in the available punctuation area is determined.

If the frequency band index corresponding to the frequency band where the punctuation is located in the available punctuation area is an even number, it is determined that the embedded code in the available punctuation area is 0.

If the frequency band index corresponding to the frequency band where the punctuation is located in the available punctuation area is an odd number, then it is determined that the embedded code in the available punctuation area is 1.

Following the above example, according to the frequency band where the punctuation in the determined first available punctuation area "show is" is 5 or 7, the embedded code in this area is obtained as 1.

The above-mentioned watermark extraction method may also include the following steps:

S204: According to the same information embedding rules when embedding the watermark information and the determined embedding codes in each available punctuation area, obtain the embedding information in the text document. Specifically:

According to the same information embedding rules when embedding watermark information and the determined embedding codes in each available punctuation area, the binary number embedded in the text document is obtained.

According to the obtained binary number, the embedded information corresponding to the text document is obtained. Specifically include: if the binary number is the embedded information, then the embedded information can be obtained directly; if the binary number is not the embedded information, search for the corresponding relationship between the stored embedded binary number and the embedded information, and obtain the embedded information.

Using the above example, according to the embedded codes in each available punctuation area and the allocation rules when embedding watermark information, the binary number embedded in the text segment is 1111011, and the information embedded in the text document can be recovered as 123.

In the above watermark embedding and extraction method, when embedding the watermark, the frequency band can be divided by analyzing the scanned image of the text through computer software, and can also be directly divided on paper with tools such as rulers; correspondingly, when extracting the watermark, it can also be divided by computer Extract by software or manually to obtain the embedded code corresponding to each available punctuation area; then the embedded binary number can be recovered.

According to the above watermark embedding method, a watermark embedding device can be constructed, as shown in FIG.

The area determination module 101 is configured to determine the available punctuation area in the text document to embed information according to the set area determination rule.

Preferably, the area determination module 101 may further include: an acquisition unit 1011 , a punctuation determination unit 1012 and an area determination unit 1013 .

The acquiring unit 1011 is configured to acquire the plain text area in the text document.

The punctuation determination unit 1012 is configured to perform text segmentation and punctuation analysis on the plain text area acquired by the acquisition unit 1011, and determine the included available punctuation; where there is at least one adjacent other character before and after the available punctuation.

The area determination unit is configured to define a start boundary and an end boundary according to the available punctuation determined by the punctuation determination unit 1012 and two other adjacent characters before and after, so as to obtain an available punctuation area.

The position determination module 102 is configured to determine the original position of the punctuation points in each available punctuation area according to the set position determination rules.

Preferably, the location determination module 102 may further include: a frequency band division unit 1021 and a location determination unit 1022 .

The frequency band division unit 1021 is configured to calculate the distance from the start boundary to the end boundary of each available punctuation area, and divide each available punctuation area into several frequency bands according to the distance.

The position determination unit 1022 is configured to respectively determine the frequency band where the punctuation is located in each available punctuation area and the corresponding frequency band index according to the coordinate position of the punctuation point in each available punctuation area.

Information embedding module 103, for each available punctuation area, according to the original position of wherein punctuation and corresponding code to be embedded, the position of wherein punctuation is adjusted, realizes that corresponding code to be embedded is embedded in the corresponding available punctuation area .

The above-mentioned watermark embedding device also includes: a code distribution module 104, which is used to determine the binary number to be embedded according to the information to be embedded corresponding to the text document; wherein, the number of digits of the binary number is less than or equal to the determined number of available punctuation areas; and According to the binary number to be embedded and the set information embedding rules, the code to be embedded in each available punctuation area is determined.

According to the above watermark extraction method, a watermark extraction device can be constructed, as shown in FIG. 7 , including: an area determination module 201 , a position determination module 202 and a code extraction module 203 .

The area determination module 201 is configured to determine the available punctuation area in the text document in which the information has been embedded according to the same area determination rule as when embedding the watermark information.

Preferably, the area determination module 201 may further include: an acquisition unit 2011 , a punctuation determination unit 2012 , and an area determination unit 2013 .

The acquiring unit 2011 is configured to acquire the plain text area in the text document.

The punctuation determination unit 2012 is configured to perform text segmentation and punctuation analysis on the plain text area acquired by the acquisition unit 2011, and determine the included available punctuation; wherein, there is at least one adjacent other character before and after the available punctuation.

The area determination unit 2013 is configured to determine the available punctuation area according to the available punctuation determined by the punctuation determination unit 2012, two other adjacent characters before and after it, and the start boundary and end boundary defined when embedding the watermark information.

The position determination module 202 is configured to determine the positions of the punctuation points in each available punctuation area according to the same position determination rules as when embedding the watermark information.

Preferably, the location determination module 202 may further include: a frequency band division unit 2021 and a location determination unit 2022 .

The frequency band division unit 2021 is configured to calculate the distance from the start boundary to the end boundary of each available punctuation area, and divide each available punctuation area into several frequency bands according to the distance.

The position determining unit 2022 is configured to determine the frequency band where the punctuation points are located in each available punctuation area and the corresponding frequency band index according to the coordinate positions of the punctuation points in each available punctuation area.

The code extraction module 203 is configured to respectively determine the embedded codes in each available punctuation area according to the position of the punctuation determined by the position determination module 202 .

The above-mentioned watermark extracting device also includes: an information recovery module 204, which is used to obtain the binary number embedded in the text document according to the same information embedding rules when embedding the watermark information and the determined embedding codes in each available punctuation area; and according to the above Binary number, get the embedded information corresponding to the text document.

The watermark embedding and extraction method and device provided by the embodiments of the present invention realize embedding the code to be embedded into the corresponding available punctuation area by selecting the available punctuation area and adjusting the punctuation position in each available punctuation area. When extracting the watermark, the corresponding rules are used to extract the embedded codes in each available punctuation area according to the adjusted punctuation position. The above method is easy to operate, and because the sensitivity of the human eye to changes in punctuation positions is much smaller than changes in character positions, large changes can be made, so that the embedded watermark information has high stability and good concealment, and at the same time can obtain good visuals.

In addition, the binary number to be embedded can also be determined according to the information to be embedded, and the code to be embedded can be assigned to each available punctuation area through the set information embedding rules. According to the value of the code to be embedded and the original position of the punctuation (where frequency band), adjust and move the position of the punctuation, and complete the information embedding; the safety and reliability of information embedding are high, and the accuracy of information extraction is relatively high. Especially when embedding watermark information by changing the center of gravity of punctuation, better, more stable and more accurate embedding and detection and extraction effects can be obtained.

The above-mentioned method and device provided by the embodiment of the present invention have good concealment of embedded watermark information, and have high robustness, and can resist multiple printing, copying, and scaling attacks. Watermark extraction can realize blind detection, and the operation is fast , suitable for occasions that require high vision and high robustness. It solves the problem that the existing technology is difficult to be used in practice due to high implementation cost or unresolved contradiction between stability and visual effect.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any skilled person in the technical field can easily think of changes, Replacement or application to other similar devices shall fall within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (24)

1. a watermark embedding method is characterized in that, comprising:

Confirm rule according to the zone of setting, determine the available punctuate zone in the text document of waiting the information of embedding;

Confirm rule according to the position of setting, determine the original position of punctuate in each said available punctuate zone respectively;

To each available punctuate zone, according to the wherein original position and the corresponding embedded coding of treating of punctuate, the position of punctuate is wherein adjusted, realize saidly treating that embedded coding is embedded in the available punctuate zone of correspondence.

2. the method for claim 1 is characterized in that, and is said according to the definite rule in the zone of setting, and determines the available punctuate zone in the text document of waiting the information of embedding, specifically comprises:

Obtain the plain text zone in the said text document;

Literal cutting and punctuate analysis are carried out in said plain text zone, determine the available punctuate that comprises; The front and back of said available punctuate all have other adjacent characters at least;

According to two adjacent other characters of said available punctuate and front and back thereof, define initial border and stop the border, obtain available punctuate zone.

3. method as claimed in claim 2 is characterized in that, the initial border in said available punctuate zone comprises: the left margin of front character, right margin, centre of gravity place or center;

The termination border in said available punctuate zone comprises: left margin, right margin, centre of gravity place or the center of back character.

4. method as claimed in claim 2 is characterized in that, and is said according to the definite rule in the position of setting, and determines the original position of punctuate in each said available punctuate zone respectively, specifically comprises:

The initial border of calculating each said available punctuate zone is several frequency bands according to said distance with each said available punctuate area dividing to the distance that stops the border;

And, determine the frequency band and the corresponding band index at punctuate place in each available punctuate zone respectively according to the coordinate position that the punctuate in each available punctuate zone belongs to.

5. method as claimed in claim 4 is characterized in that, the coordinate position at said punctuate place specifically comprises: the coordinate position at the center of gravity of said punctuate, center, left margin or right margin place.

6. method as claimed in claim 4 is characterized in that, said basis is the original position and the corresponding embedded coding of treating of punctuate wherein, and the position of punctuate is wherein adjusted, and specifically comprises:

If the corresponding band index in original position at said punctuate place is an odd number, and the corresponding embedded coding of treating is 0, and position to the band index that then moves said punctuate is the frequency band of even number;

If the corresponding band index in original position at said punctuate place is an odd number, and the corresponding embedded coding of treating is 1, then do not change the frequency band at said punctuate place;

If the corresponding band index in original position at said punctuate place is an even number, and the corresponding embedded coding of treating is 0, then do not change the frequency band at said punctuate place;

If the corresponding band index in original position at said punctuate place is an even number, and the corresponding embedded coding of treating is 1, and position to the band index that then moves said punctuate is the frequency band of odd number.

7. like the arbitrary described method of claim 1-6, it is characterized in that, also comprise:

The treat embedding information corresponding according to said text document is determined binary number to be embedded; The figure place of said binary number is smaller or equal to the said available punctuate region quantity of determining;

Information according to binary number said to be embedded and setting embeds rule, determines the embedded coding of treating in each said available punctuate zone.

8. method as claimed in claim 7 is characterized in that, said the treat embedding information corresponding according to said text document is determined binary number to be embedded, and specifically comprises:

If the said embedding information of treating itself is maybe to be converted into the binary number of figure place smaller or equal to said available punctuate region quantity, confirm that then the said binary number of treating that embedding information or conversion obtain is a binary number to be embedded;

If the said embedding information of treating itself is maybe to be converted into the binary number of figure place greater than said available punctuate region quantity, or the said embedding information of treating is not and can not be converted into binary number the time; Then select the binary number of a units,, and set up binary number and the said corresponding relation of waiting the information of embedding to be embedded as binary number to be embedded smaller or equal to said available punctuate region quantity.

9. method as claimed in claim 8 is characterized in that, said information according to binary number said to be embedded and setting embeds rule, determines the embedded coding of treating in each said available punctuate zone, specifically comprises:

If the figure place of binary number said to be embedded equals the quantity in available punctuate zone, the binary coding that binary number that then directly will be to be embedded comprises is distributed to each available punctuate zone respectively as treating embedded coding;

If binary number figure place to be embedded less than the quantity in available punctuate zone, then distributes the binary coding that comprises in the binary number to be embedded as treating embedded coding respectively through redundant arithmetic for each available punctuate zone.

10. a watermark extracting method is characterized in that, comprising:

Rule is confirmed in zone identical according to embed watermark information the time, determines the available punctuate zone in the text document of the information that embeds;

Rule is confirmed in position identical according to embed watermark information the time, determines the position at punctuate place in each said available punctuate zone respectively;

According to the position at said punctuate place, determine the embedded coding in each available punctuate zone respectively; The position at said punctuate place be when embed watermark according to the wherein original position and the corresponding embedded coding of treating of punctuate, obtain after the position of said punctuate adjusted.

11. method as claimed in claim 10 is characterized in that, said basis is confirmed rule in identical zone during with embed watermark information, determines the available punctuate zone in the text document of the information that embeds, specifically comprises:

Obtain the plain text zone in the said text document;

Literal cutting and punctuate analysis are carried out in said plain text zone, determine the available punctuate that comprises; The front and back of said available punctuate all have other adjacent characters at least;

According to said available punctuate, the initial border that defines when adjacent two other characters and embed watermark information before and after it with stop the border, determine available punctuate zone.

12. method as claimed in claim 11 is characterized in that, the initial border in said available punctuate zone comprises: the left margin of front character, right margin, centre of gravity place or center;

The termination border in said available punctuate zone comprises: left margin, right margin, centre of gravity place or the center of back character.

13. method as claimed in claim 11 is characterized in that, said basis is confirmed rule in identical position during with embed watermark information, determines the position at punctuate place in each said available punctuate zone respectively, specifically comprises:

The initial border of calculating each said available punctuate zone is several frequency bands according to said distance with each said available punctuate area dividing to the distance that stops the border;

And, determine the frequency band and the corresponding band index at punctuate place in each available punctuate zone respectively according to the coordinate position that the punctuate in each available punctuate zone belongs to.

14. method as claimed in claim 13 is characterized in that, the coordinate position at said punctuate place specifically comprises: the coordinate position at the center of gravity of said punctuate, center, left margin or right margin place.

15. method as claimed in claim 13 is characterized in that, said position according to said punctuate place is determined the embedded coding in each available punctuate zone respectively, specifically comprises:

If the corresponding band index of frequency band at the place of punctuate described in the available punctuate zone is an even number, confirm that then the embedded coding in this available punctuate zone is 0;

If the corresponding band index of frequency band at the place of punctuate described in the available punctuate zone is an odd number, confirm that then the embedded coding in this available punctuate zone is 1.

16. like the arbitrary described method of claim 10-15, it is characterized in that, also comprise:

Identical information embeds the embedded coding in rule and each available punctuate zone of determining during according to embed watermark information, obtains embedding the binary number in the said text document; And

According to said binary number, obtain the corresponding embedding information of said text document.

17. a watermark flush mounting is characterized in that, comprising:

The zone determination module is used for confirming rule according to the zone of setting, and determines the available punctuate zone in the text document of waiting the information of embedding;

Position determination module is used for confirming rule according to the position of setting, and determines the original position of punctuate in each said available punctuate zone respectively;

The information merge module is used for to each available punctuate zone, according to the wherein original position and the corresponding embedded coding of treating of punctuate, the position of punctuate is wherein adjusted, and realizes saidly treating that embedded coding is embedded in the available punctuate zone of correspondence.

18. device as claimed in claim 17 is characterized in that, said regional determination module comprises:

Acquiring unit, the plain text zone that is used for obtaining said text document;

Punctuate is confirmed the unit, is used for literal cutting and punctuate analysis are carried out in said plain text zone, determines the available punctuate that comprises; The front and back of said available punctuate all have other adjacent characters at least;

The unit is confirmed in the zone, is used for according to two adjacent other characters of said available punctuate and front and back thereof, defines initial border and stops the border, obtains available punctuate zone.

19. device as claimed in claim 18 is characterized in that, said position determination module comprises:

The frequency band division unit, the initial border that is used to calculate each said available punctuate zone is several frequency bands according to said distance with each said available punctuate area dividing to the distance that stops the border;

Position determination unit is used for the coordinate position according to the punctuate place in each available punctuate zone, determines the frequency band and the corresponding band index at punctuate place in each available punctuate zone respectively.

20. like the arbitrary described device of claim 17-19, it is characterized in that, also comprise:

The coding assignment module is used for the treat embedding information corresponding according to said text document, determines binary number to be embedded; The figure place of said binary number is smaller or equal to the said available punctuate region quantity of determining; And embed rule according to the information of binary number said to be embedded and setting, determine the embedded coding of treating in each said available punctuate zone.

21. a watermark extraction apparatus is characterized in that, comprising:

The zone determination module, rule is confirmed in identical zone when being used for basis with embed watermark information, determines the available punctuate zone in the text document of the information that embeds;

Position determination module, rule is confirmed in position identical when being used for basis with embed watermark information, determines the position at punctuate place in each said available punctuate zone respectively;

The coding extraction module is used for the position according to said punctuate place, determines the embedded coding in each available punctuate zone respectively; The position at said punctuate place be when embed watermark according to the wherein original position and the corresponding embedded coding of treating of punctuate, obtain after the position of said punctuate adjusted.

22. device as claimed in claim 21 is characterized in that, said regional determination module comprises:

Acquiring unit, the plain text zone that is used for obtaining said text document;

Punctuate is confirmed the unit, is used for literal cutting and punctuate analysis are carried out in said plain text zone, determines the available punctuate that comprises; The front and back of said available punctuate all have other adjacent characters at least;

The unit is confirmed in the zone, is used for according to said available punctuate, the initial border that defines when adjacent two other characters and embed watermark information before and after it and stops the border, determines available punctuate zone.

23. device as claimed in claim 22 is characterized in that, said position determination module comprises:

The frequency band division unit, the initial border that is used to calculate each said available punctuate zone is several frequency bands according to said distance with each said available punctuate area dividing to the distance that stops the border;

Position determination unit is used for and according to the coordinate position at the punctuate in each available punctuate zone place, determines the frequency band and the corresponding band index at punctuate place in each available punctuate zone respectively.

24. like the arbitrary described device of claim 21-23, it is characterized in that, also comprise:

Information is recovered module, and identical information embeds the embedded coding in rule and each available punctuate zone of determining when being used for according to embed watermark information, obtains embedding the binary number in the said text document; And, obtain the corresponding embedding information of said text document according to said binary number.

Images