Method for embedding invisible fingerprints by using character typesetting

Abstract

The invention relates to a method for embedding invisible fingerprints by using character typesetting, which comprises the following steps of: acquiring fingerprint information and characters to be typeset, generating binary messages of 0 and 1 from the fingerprint information, and generating a positive group of fingerprints and a negative group of fingerprints from the same group of fingerprint information; performing typesetting output, calculating the display blank interval width of 0 and 1 during output, and performing reverse coding and the same operation; generating an image from the visual characters containing the invisible fingerprints, carrying out OCR (Optical Character Recognition) analysis, and then processing through an image binary method to generate a black-and-white image only containing the characters; the characters are changed into a pure black square frame; extracting the blank interval width between the black square frames and the square frames, corresponding to a numerical value of 0 or 1, and extracting a binary message; and inputting a fingerprint system for translation to obtain real information of hidden fingerprints. The problems that a current copyright and authorization information adding method affects impression, is easy to tamper and is not strong in protection are solved, reading and viewing are not affected, and a pirate cannot find and perceive the copyright and authorization information easily.

Description

Technical Field

[0001] The present invention relates to the technical fields of text typesetting and copyright, and particularly to a method for embedding invisible fingerprints by using text typesetting. Background Art

[0002] With the development of the information field, current various cultural dissemination technologies have developed rapidly, and the information dissemination speed is extremely fast. Information can be rapidly disseminated in the Internet field in the forms of documents, pictures, videos, and book publications. The progress of technology makes the pirated dissemination of copyrighted information very simple. Just with one copy, others' hard work can be sold. When rights protection is carried out, it is very difficult to collect evidence. The current technology is to embed invisible fingerprint information in documents, pictures, and videos by using the text typesetting technology inside. When the information is pirated, the hidden fingerprint can be used to indicate the copyright ownership, and in addition, it can also be traced back to find the information leakage source.

[0003] Currently, the main methods for adding copyright and authorization information are as follows:

[0004] 1) Indicate the copyright owner and the authorized user in the title headers of documents, pictures, and videos, indicating that the information has copyright information and cannot be copied and disseminated at will. However, this method is easily tampered with by pirates, and the modified result is disseminated, circumventing the copyright information.

[0005] 2) Use the watermark technology to add the watermarks of the copyright owner and the authorized user to documents, pictures, and videos to identify the copyright information. Because the watermarks are repeated all the time, the quantity and area are large, the tampering difficulty is increased, and the piracy cost consumption will increase. However, because the watermarks are directly displayed, they will interfere with the observers, which is not conducive to information viewing. In addition, with the development of technology, the cost of removing the watermark technology is also decreasing, and the copyright protection power is also decreasing.

[0006] 3) Use the two-dimensional code technology to add two-dimensional codes to documents, pictures, and videos. The two-dimensional codes contain copyright information. The two-dimensional codes can be in multiple positions to enhance the copyright awareness, but they are also easily recognized and covered and removed by technical means, and the copyright protection strength is not great. Summary of the Invention

[0007] Aiming at the problems that the current methods for adding copyright and authorization information affect the visual experience, are easily tampered with, and have little protection strength, a method for embedding invisible fingerprints by using text typesetting is proposed. The text typesetting technology can be used to embed invisible fingerprint information in the text typesetting, which is difficult for pirates to discover and perceive. Because the fingerprints are invisible, they have no impact on the observers, and the information can be extracted and processed comfortably. In addition, the erasure technology will damage the whole information, and there is no simple method to remove it currently.

[0008] By using the typesetting technology of text, the invisible information is embedded by taking advantage of the inconsistent width of the blank space between the texts. Because the distance can be controlled very accurately when embedding, the difference between the two sets of values ​​0 and 1 can be controlled at the pixel level. The human eye cannot distinguish such a small difference, so it will not affect reading and viewing. The invisible information can be embedded between one or more lines of text, but if the encoding is a simple loop repetition, it will cause multiple lines of typesetting to be repeated, and it is easy to find patterns between the lines of text. At this time, only the invisible information needs to be retrieved. The two sets of fingerprints correspond to the same information. After processing and recycling, the errors can be disordered, the lengths are different, and the sorting is more natural and coordinated. In addition, the fingerprint encoding should not be continuous 0 or continuous 1, but the signal 0 and signal 1 should be evenly distributed. In this way, the overall text layout will become very natural and coordinated, and the invisible fingerprint information cannot be seen even if the font is enlarged.

[0009] The technical solution of the present invention is:

[0010] A method for embedding invisible fingerprints using text layout, comprising the following steps:

[0011] Invisible fingerprint embedding stage:

[0012] Step 1.1: Obtain fingerprint information and text to be typeset, generate binary messages of 0 and 1 from the fingerprint information, set the encoding rule to evenly distribute the 0 and 1 values ​​in the binary in the message, and generate two sets of positive and negative fingerprints from the same set of fingerprint information;

[0013] Step 1.2: Layout and output the first set of fingerprint information and the actual displayed text. When outputting, first calculate the display blank interval width of 0 and 1. Use the first three characters of the actual displayed text to fix the width of 0 and 1 of the first set of fingerprint information. The two sets of values ​​are different. Calculate the corresponding blank width of the corresponding position of the fingerprint code to layout the text.

[0014] Step 1.3: Embed the reverse code of the invisible fingerprint into the text layout according to the same rules as step 2, so that the invisible fingerprint is embedded in the carrier containing the text layout;

[0015] Invisible fingerprint extraction stage:

[0016] Step 2.1: Generate an image from the visualized text containing the invisible fingerprint to obtain an image containing the text information of the invisible fingerprint;

[0017] Step 2.2: Perform OCR analysis on the image to find the image area containing text, and then use the image binary method to process the entire image except the text into white and the text into black, generating a black and white image containing only the text;

[0018] Step 2.3: Further process the image. Scan it in both the horizontal and vertical directions, and process the start and end coordinates of the positions occupied by the text into black, so that the text becomes a pure black rectangle.

[0019] Step 2.4: Take out the first three black rectangles in a row. The blank interval width between two rectangles corresponds to the value 0 or 1. Use the blank interval widths of 0 and 1 values as the reference values, and compare the subsequent blank interval widths one by one. If it is less than or equal to the 0-value reference width, it is 0 or 1; if it is greater than or equal to the 1-value reference width, it is the other value, and extract the binary message.

[0020] Step 2.5: Input the binary message extracted in Step 4 into the fingerprint system for translation to obtain the real information of the hidden fingerprint.

[0021] Furthermore, in Step 1.1, the coding information can be customized. The pre-written data can be first encoded into digital information and then encoded into binary again.

[0022] Furthermore, in Step 1.2, generally, the blank interval width of the 0 value is short, and the blank interval width of the 1 value is long. The difference between the two groups of blank widths can be defined by oneself. In addition, the width of the basic blank can be defined according to the typesetting requirements, and only the relative value is taken, not the absolute value.

[0023] Furthermore, in Step 1.3, if a line of text is too short and there is not enough available blank space, the invisible fingerprint can be considered abandoned and equal-length interval typesetting can be used. When a line of text is too long and there are extra blank positions after embedding, all 1 values or 0 values can be embedded. According to actual needs, cyclic embedding can also not be performed, and invisible fingerprints can be embedded only in a certain sentence on each page to speed up the typesetting speed.

[0024] Furthermore, in Step 1.3, the carriers containing text typesetting include ordinary documents and printed publications, and also include pictures and videos. Pictures and videos are embedded according to the method in Step 2. Pictures can be embedded in the explanatory text. For horizontal typesetting of the explanatory text, normal horizontal blank intervals are used, and for vertical typesetting, vertical blank intervals are used. Videos can be embedded in the subtitles below.

[0025] Furthermore, generate images of the visual text containing invisible fingerprints. For documents, the screenshot method can be used; for publications, taking pictures can be done; for video streams, the screenshot method can be used.

[0026] Further, the method of embedding hidden fingerprint information includes: only embedding a set of hidden fingerprints, and one can choose a relatively complete line of explanatory text, and use the blank spaces between the texts to complete forward coding or reverse coding embedding; selectively cyclic embedding, select the paragraphs that can embed fingerprint information, and find that the blank spaces in a whole line can accommodate the hidden fingerprint message for embedding, and the typesetting of the extra-long part can choose equal-length blank spaces; embed hidden fingerprints throughout the article, and the embedding method is carried out alternately in positive and negative groups for adjacent paragraph lines.

[0027] The beneficial effects of the present invention are as follows:

[0028] For traditional methods of prompting copyright information such as text reminders, watermarks, and two-dimensional codes, they are easy to be discovered when stolen and are easy to be erased by technical means. In addition, these information will also cause certain interference to the observers, with a poor experience. Ordinary people will have a sense of disgust when seeing watermarks all over the text, pictures, and videos. While invisible fingerprints can also embed copyright and other information, but do not affect the viewing and reading of the information, just as if it does not exist, and can only be extracted through technical means. Because it is not easy to detect, the pirated party will inadvertently bring the invisible watermark into their own publications, so as to safeguard rights through invisible fingerprints and trace the leakage method. Brief Description of the Drawings

[0029] Figure 1 is the flowchart of invisible fingerprint embedding of the present invention;

[0030] Figure 2 is the flowchart of invisible fingerprint extraction of the present invention;

[0031] Figure 3 is the schematic diagram of invisible fingerprint embedding into text typesetting of the present invention;

[0032] Figure 4 is the overall typesetting effect diagram after invisible fingerprint embedding of the present invention;

[0033] Figure 5 is the overall vertical typesetting effect diagram of text after invisible fingerprint embedding of the present invention;

[0034] Figure 6 is the overall typesetting effect diagram after invisible fingerprint embedding of the present invention. Detailed Embodiment

[0035] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and gives detailed implementation manners and specific operation processes, but the protection scope of the present invention is not limited to the following embodiments.

[0036] A method of embedding invisible fingerprints using text typesetting includes the following steps:

[0037] I. Invisible fingerprint embedding, as shown below: Figure 1 shown as follows:

[0038] Step 1: Obtain fingerprint information and text to be typeset. Generate a binary message of 0s and 1s from the fingerprint information. The encoding information can be customized. The pre-written data can be first encoded into digital information and then encoded into binary again, so as to ensure that the fingerprint encoding is as short as possible. The encoding rule needs to be simply designed to ensure that the 0 values and 1 values in the binary are evenly distributed in the message. The same set of fingerprint information generates two sets of fingerprints, positive and negative. For example, if one set is 011010, the other set is 100101.

[0039] 1) Fingerprint information: The hidden information that the author wants to embed in documents, pictures, and videos, generally including company and author information, as well as the internal journal number of the published content, etc. The Chinese ASCII encoding (GBK encoding) of such text data can be regarded as fingerprint information, which has an identity identifier and can correctly identify the original source and copyright information of the published content. For example, when Wind Information Co., Ltd. publishes an article, the binary corresponding to the GBK Chinese character encoding of Wind, "0XCD 0XF2 0XB5 0XC3", i.e., "1100 1101 11110010 1011 0101 1100 0011", can be regarded as fingerprint information.

[0040] 2) Text to be typeset: The display text information in documents, pictures, and videos (captions), which is published to the public in an open form.

[0041] 3) Two sets of fingerprints (encodings): For example, the positive encoding "011010" and the negative encoding "100101" (both have the same meaning and represent fingerprint information). In order to ensure more natural and harmonious typesetting in documents and pictures, the positive encoding is used for typesetting in the first line of typeset text, and the negative encoding is used in the second line. Such an effect will make the entire paragraph look natural and beautiful, without obvious vertical alignment of characters in a unified pattern.

[0042] Step 2: Typeset and output the first set of fingerprint information and the actual display text. When outputting, first calculate the display blank interval widths for 0 values and 1 values. The first 3 characters of the actual display text are needed to fix the widths of 0 values and 1 values of the first set of fingerprint information. Generally, the blank interval width for 0 values is short, and the blank interval width for 1 values is long. Of course, it can also be the other way around, as long as the two values are different. The difference in the two blank widths can be defined by oneself, and the basic blank width can be defined according to the typesetting requirements, only taking relative values, not absolute values. Perform typesetting on the text by finding the corresponding blank widths for the corresponding bits of the fingerprint encoding. The effect is as shown in Figure 3 .

[0043] 1) Actual displayed text: It is the text information to be published. Fingerprint information is generally short information and can be embedded within one or two lines by utilizing the text layout gaps.

[0044] 2) Fix the interval width between the first 3 characters "0" and "1": In typesetting, a real blank gap must be specified (in the monitor, it can be defined in pixels; in printed publications, it is defined in millimeters as the length unit) to identify the "0" value and "1" value. It can be defined by the two blank intervals in the first 3 characters of a line. After defining the actual occupied width, the subsequent text is typeset based on this blank interval distance, and different blank widths are taken according to the fingerprint numbers at different position information.

[0045] 3) Figure 3 Explanation: In the figure, it shows the typesetting of two lines of text "I love Wind economic data", successfully embedding the fingerprints "00110" and "11001", and the representative meaning is the number "3". Because the picture is made manually and is time-consuming, the complete fingerprint is not embedded. It only explains how to embed the binary code and the text typesetting effect after embedding the fingerprint.

[0046] Step 3: Embed the reverse code of the invisible fingerprint into the text layout according to the same rules as in Step 2, thus completing the embedding of the invisible fingerprint into ordinary documents and printed publications. Additionally, it should be noted that if a line of text is too short and there is not enough available blank space, the invisible fingerprint can be considered abandoned and equal-length interval typesetting can be used. When a line of text is too long and there are extra blank spaces after completion of the embedding, all 1 values or 0 values can be embedded. According to actual requirements, non-cyclic embedding can also be not carried out, and the invisible fingerprint can be embedded only in a certain sentence on each page to speed up the typesetting. The overall typesetting effect is as Figure 4 。

[0047] Step 4: Pictures and videos can also be embedded according to the method in Step 2. Pictures can be embedded in the explanatory text. For horizontal typesetting of the explanatory text, normal horizontal blank intervals are used, and for vertical typesetting, vertical blank intervals are used. Videos can be embedded in the subtitles below. The effect is as Figure 5 。

[0048] The carriers containing text layout include ordinary documents and printed publications, and also include pictures and videos:

[0049] A. Ordinary electronic documents (including word, excel, pdf, html, txt, etc.) or corresponding publications: In the information disseminated by such text-based media, whether it is horizontal typesetting or vertical typesetting, fingerprint information can be well embedded.

[0050] B. Pictures or corresponding publications: It is impossible to embed fingerprint information into a picture itself, but fingerprint information can be embedded using the additional explanatory text in the picture. If it is a pure picture without text information, the above method cannot be used for fingerprint embedding.

[0051] C. Videos: Similar to pictures, use the subtitle information in the video for fingerprint information embedding.

[0052] II. Invisible fingerprint extraction, as Figure 2 shown:

[0053] Step 1: Generate an image from the visual text containing the invisible fingerprint. For documents, the screenshot method can be used; for publications, taking pictures; for video streams, the screenshot method can be used to obtain an image containing the text information of the invisible fingerprint.

[0054] 1) Image: It can be understood as a picture. Mainly considering the maturity of image processing technology, it is convenient for "invisible fingerprint" extraction. Of course, extraction can also be carried out without relying on pictures. For example, through manual comparison, the blank intervals between "0" and "1" can be enlarged, and the human eye can distinguish them. Essentially, it is still to distinguish the blank intervals between characters. Image processing can enable the machine to distinguish very subtle information processing and achieve automated extraction, similar to the barcode scanning effect in supermarkets.

[0055] Step 2: Conduct OCR analysis on the image to find the image area containing text, and then process it through image binarization. All parts of the image except the text are processed into white, and the text is processed into black to generate a black-and-white image containing only text.

[0056] Step 3: Further process the image. Scan it in both the horizontal and vertical directions, and process the start and end coordinates of the positions occupied by the text into black, turning the text into a pure black square.

[0057] Step 4: Take out the first three black squares in a row. The width of the blank interval between two squares corresponds to the value 0 or 1. Use the width of the blank intervals of 0 and 1 values as the reference values, and compare the width of the subsequent blank intervals one by one. If it is less than or equal to the reference width of 0 value, it is 0; if it is greater than or equal to the reference width of 1 value, it is 1, and extract the binary message. The effect is as Figure 6 .

[0058] 1) Extract the binary message: The two blank intervals (one long and one short, representing "1" and "0" respectively. If the short value is "0", the long value is "1"; conversely, if the short value is "1", the long value is "0". The two resulting messages are complementary binary with the same meaning. In the legend, the short value is taken as "0" and the long value as "1") in the first three black boxes (from "I Love Wind Economic Database") are used as the reference calibration lengths. For the entire line of text, if the blank interval is less than or equal to the short value length, it is "0", and if it is greater than or equal to the long value length, it is "1". The embedded data fingerprint in the first line of the final parsing result is "00110", and the embedded invisible fingerprint in the second line is "11001".

[0059] Step 5: Input the binary message extracted in Step 4 into the fingerprint system for translation to obtain the true information of the hidden fingerprint.

[0060] Several methods and means for embedding hidden fingerprint information:

[0061] Scenario 1: When only one set of hidden fingerprints is embedded in an article, you can choose a relatively complete explanatory text at the beginning, middle, or end of the article and use the blank intervals between the words to complete forward coding or reverse coding embedding. The fingerprint information can be embedded in a paragraph of text, and it is not necessary to complete the embedding of all fingerprint message in a single line. It can be completed in a paragraph. For the typesetting of the extra-long part, equal-length blank intervals (all "0" values or all "1" values) can be selected. The blank intervals between the words in other text paragraphs can be processed with equal lengths without embedding any information, as shown in Examples 1, 2, and 3.

[0062] Example 1:

[0063] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0064] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0065] 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 -> Contains fingerprint information (0011), extra-long part set to 0

[0066] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0067] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0068] Example 2:

[0069] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -> The paragraph line does not contain fingerprint information

[0070] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -> The paragraph line does not contain fingerprint information

[0071] 1 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 -> Contains fingerprint information (0011), set to 1 when exceeding the length

[0072] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -> The paragraph line does not contain fingerprint information

[0073] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -> The paragraph line does not contain fingerprint information

[0074] Example 3:

[0075] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0076] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information 0 1 0 0 1 1 0 1 1 0 1 0 1 0 0 1 1 1 0 1 0 1 1 1 0 1 1 0 1 0 1 0 1 0

[0079] 0 1 0 1 1 1 0 0 1 0 1 0 0 0 0 0 0 -> Contains fingerprint information (0011 01101010 00111011 1011 0101 1010 0111 0010 1000), set to 0 when exceeding the length

[0080] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0081] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0082] Scenario 2: Selective loop embedding. Select paragraphs that can embed fingerprint information. It is found that the blank spaces in a whole line of text can accommodate the embedding of hidden fingerprint messages. For the typesetting of the extra-long part, equal-length blank spaces (all "0" values or all "1" values) can be selected. The embedding of hidden fingerprint messages is carried out alternately in two groups, positive and negative, for adjacent paragraph lines. As long as the blank spaces in a paragraph line can completely embed the hidden fingerprint, embedding is performed. For a line of text that is too short, equal-length blank spaces (all "0" values or all "1" values) are selected for processing, as shown in Example 4.

[0083] Example 4:

[0084] 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 -> Contains fingerprint information (0011), extra-long part set to 0

[0085] 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0086] 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 -> Contains fingerprint information (1100), extra-long part set to 1

[0087] 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 -> Contains fingerprint information (0011), extra-long part set to 0

[0088] 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0089] 1 0 1 1 0 0 1 1 1 1 1 1 1 1 1 1 1 -> Contains fingerprint information (1100), extra-long part set to 1

[0090] 0 0 0 0 0 -> The paragraph line does not contain fingerprint information

[0091] Scenario 3: Embed hidden fingerprints throughout the article. Regardless of the length of a line of text, invisible fingerprint messages need to be embedded. The embedding method is carried out alternately in two groups, positive and negative, for adjacent paragraph lines. If the number of blank spaces in a paragraph line is not enough to accommodate, the embedding is interrupted. If the number of blank spaces in a paragraph line exceeds the requirement, the invisible fingerprint message is circularly embedded, as shown in Example 5.

[0092] Example 5:

[0093] 0 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 -> Contains fingerprint information (0011), with repetitions

[0094] 1 0 1 1 0 -> Contains fingerprint information (110), truncated

[0095] 0 1 0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 -> Contains fingerprint information (0011), with repetitions

[0096] 1 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 -> Contains fingerprint information (1100), with repetitions

[0097] 0 1 0 0 1 -> Contains fingerprint information (001), truncated

[0098] 1 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1 0 -> Contains fingerprint information (1100), with repetitions

[0099] 0 1 0 0 1 -> Contains fingerprint information (001), truncated

[0100] The above-described embodiments merely represent one implementation manner of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. A method for embedding invisible fingerprints using text layout, characterized in that: The steps include: Invisible fingerprint embedding stage: Step 1.1: Obtain fingerprint information and text to be typeset, generate binary messages of 0 and 1 from the fingerprint information, set the encoding rule to evenly distribute the 0 and 1 values ​​in the binary in the message, and generate two sets of positive and negative fingerprints from the same set of fingerprint information; Step 1.2: Layout and output the first set of fingerprint information and the actual displayed text. When outputting, first calculate the display blank interval width of 0 and 1. Use the first three characters of the actual displayed text to fix the width of 0 and 1 of the first set of fingerprint information. The two sets of values ​​are different. Calculate the corresponding blank width of the corresponding position of the fingerprint code to layout the text. Step 1.3: Embed the reverse code of the invisible fingerprint into the text layout according to the same rules as step 2, so that the invisible fingerprint is embedded in the carrier containing the text layout; Invisible fingerprint extraction stage: Step 2.1: Generate an image from the visualized text containing the invisible fingerprint to obtain an image containing the text information of the invisible fingerprint; Step 2.2: Perform OCR analysis on the image to find the image area containing text, and then use the image binary method to process the entire image except the text into white and the text into black, generating a black and white image containing only the text; Step 2.3: The image is further processed by scanning in both horizontal and vertical directions, and the start and end coordinates of the position occupied by the text are all processed into black, and the text becomes a pure black box; Step 2.4: Take out the first three black boxes in a row, and the width of the blank space between the two boxes corresponds to the value 0 or 1; take the blank space width of 0 and 1 as the reference value, and compare the blank space widths one by one. If the reference width is less than or equal to 0, it is 0 or 1, and if the reference width is greater than or equal to 1, it is another value, and the binary message is extracted; Step 2.5: Input the binary message extracted in step 4 into the fingerprint system for translation to obtain the real information of the hidden fingerprint.

2. The method of embedding invisible fingerprints by using text layout according to claim 1, characterized in that: In step 1.1, the encoding information can be customized, and the pre-written data can be encoded into digital information first, and then encoded into binary again.

3. The method of embedding invisible fingerprint by using text layout according to claim 1, characterized in that: In step 1.2, the width of the blank space with a value of 0 is generally short, while the width of the blank space with a value of 1 is long. The difference between the widths of the two groups of blanks can be defined by yourself. In addition, the width of the basic blank can be defined according to the typesetting requirements, taking only relative values, and not requiring absolute values.

4. The method of embedding invisible fingerprints by using text layout according to claim 1, characterized in that: In step 1.3, if a line of text is too short and there are not enough valid blank spaces, you can consider giving up the invisible fingerprint and using equal-length interval typesetting. When a line of text is too long and there are extra blank spaces after embedding, you can embed all 1 values ​​or 0 values. According to actual needs, you can also avoid cyclic embedding and only embed the invisible fingerprint in a sentence on each page to speed up typesetting.

5. The method of embedding invisible fingerprints by using text layout according to claim 1, characterized in that: In step 1.3, the carriers containing text layout include ordinary documents and publication printouts, as well as pictures and videos. The pictures and videos are embedded according to the method in step 2. Pictures can be embedded in the explanatory text. The explanatory text is horizontally laid out using normal horizontal blank intervals. If it is vertically laid out, vertical blank intervals are used. Videos can be embedded in the lower subtitles.

6. The method of embedding invisible fingerprints by using text layout according to claim 1, characterized in that: In step 2.1, the visual text containing the invisible fingerprint is generated into an image, the document can be captured by screenshot, the publication can be photographed, and the video stream can be captured by screenshot.

7. The method of embedding invisible fingerprints by using text layout according to claim 1, characterized in that: The method of embedding hidden fingerprint information includes: embedding only one set of hidden fingerprints, and you can choose a line of relatively complete explanatory text, and use the blank space between the text to complete the forward encoding or reverse encoding embedding; selective cyclic embedding, select the paragraphs that can embed fingerprint information, and find that the blank space of the entire line of text can accommodate the hidden fingerprint message for embedding, and the layout of the extra-long part can choose the equal-length blank space; embed the hidden fingerprint throughout the article, and the embedding method is to alternate between the positive and negative groups of adjacent paragraph lines.

Images