TW452732B - Methods of digital steganography for multimedia data - Google Patents

Abstract

A lossless steganographic encoding method for secure transmission or storage of multimedia data. Primary data, such as text, image, video, audio or other digital data, is utilised in a steganographic process to encode secondary data, such as text, image, video, audio or other digital data. The primary data includes a plurality of first data elements and the secondary data includes a plurality of second data elements. For each second data element an operation is performed with a first data element so as to generate a key element as a result of the operation. The key elements may then be securely transmitted and/or stored. In preferred embodiments of the method, the primary data may be rearranged according to a predefined or random manner, or it may be resized so as to match the size of the secondary data. A complementary decoding method is disclosed, and a method generating a pseudo-random number sequence, which may be used in the steganographic encoding and decoding methods, is also disclosed.

Description

452732 V. Description of the invention d) Inventive method i The present invention relates to the method of hidden data supplementary decoding method (steganogr This encoder image or video This example application application and method of using the column can be invented Background about the general use for secure transmission or Steganographic method for information storage. The present invention also interacts with it to generate a quasi-random number sequence for any digital file. This quasi-random number sequence can be used for hidden coding aphic or decoding methods. The method is particularly suitable for digitally disguising or hiding confidential information such as text, audio, and data, which will facilitate the description of the method. However, it is known that this encoding method is intended to be used more broadly. Similarly, generating a quasi-random number sequence can be used for applications other than the hidden method application. The phenomenal growth of multimedia products and services provided via the Internet and Digital Data Storage Media (DSM) has led to the need for copyright identification and integrity of data protection. Over the past few years, some digital watermarking technologies have been developed to address the legal use of copyright information on the Internet and DSM. Some digital watermarking technologies have recently been patented. These examples include US Patent No. 5,636,292 issued to Rhoads in 1997 and US Patent No. 5,6 5 9,7 2 6 issued to Stanford and Handel in 1997. Rh o a d s discloses a method of embossing an identification code on a carrier wave, such as an electronic data signal or a physical medium, to a certain extent allowing the identification code to be identified later and the carrier wave determined accordingly. Stanford and Handel disclose a way to embed subsidiary information into master data such as photos, television signals, fax transmissions, or identification cards. This method uses the noise component of the master data to operate according to the subsidiary information.

452 73 2 V. Description of invention ¢ 2). Many previous techniques of digital watermarking, including the aforementioned technology disclosed in the US patent, can only hide limited information such as a small number of logical bits (ie, π 1 and " OM) or a few letters (such as "A1 2") In the master data. However, for copyrights that exist in the master product such as Singapore satellite imagery to record all the details. Right information, the entire information or sentence may need to be hidden or linked to the master data. For example, regarding the ownership of the master product, the sentence "Singapore" The digital image is owned by Mr. Tan 'John, dated on February 16, 1997, π can provide more conclusive proof than having to rely on a simple password to determine copyright infringement. Therefore, there is still a need for a hidden encoding (stega η 〇graphic) method, which can allow the use of primary data (such as text, video, audio or video data) to encode a relatively long secondary data string without reducing the amount of primary data. (Such as text 'image * audio or video data). In addition to the Internet applications mentioned above, many potential consumers, business and service applications can benefit from the use of digital steganography, including for copyright protection and signature verification purposes and for Information security transmission. These applications include guaranteeing the hidden encoding of text, video, audio or video data (stega η 〇graphic), which contains the ownership identification or characteristic information related to digital still photography photos or video cameras, and the copyright protection of popular music records in the music industry With royalty tracking. The business and services sector may benefit from the secure transmission and reception of confidential information and related confidential documents and digital signatures of electronic transactions that can be encoded in normal data strings transmitted via open channels.

/ -N 〇 ^ V. Description of the invention (3). A pseudo-random number generator is an algorithm or device that gives a fixed random number sequence when the seeds are the same. This seed may be a number, a bit string, a digital file, or any other form of data. Typical random number generators use chaotic functions, such as SH A (Secure Hybrid. Chaotic Algorithm), which was awarded to Microsoft Corporation (Micr 〇s 〇ft 〇〇 ^ 〇 『31: 1〇11) (1998) U.S. Patent (118 & VII 6111;) No. 5,787,179 'and U.S. Patent No. 5,732,138 issued to Silicon Graphics Inc. (1998). SUMMARY OF THE INVENTION In one aspect, the present invention provides a method for generating a quasi-random number sequence, which includes the following steps: Provide source data including arranged data elements, and the content of each data element is represented by a number group; Read into an array such that each position of the array contains one of the numbers; select a starting position within the number array; and recombine the number to form a new group of numbers about the starting position such that each new group of numbers A group represents a quasi-random number and a continuous new group of numbers represents the quasi-random number sequence. In one embodiment, the data elements of the source data are represented by a binary notation, and the content of each data element is better represented by a single byte (that is, 8 bits). In this embodiment, each bit of each 8-bit byte group constitutes a number, which may be read into a one-bit array such that each position of the array contains one bit

Page 8 R Ο ^ £ — 5. Description of the invention (4) Yuan. The starting position may be selected randomly, quasi-randomly or by a previously defined method. Based on this starting position, the bits are reassembled into new and better octet groups, and each new group forms a new information byte. In this way, each new byte represents a quasi-random number, which has no numerical relationship with the value of the data element of the source data. The terms “pre-defined” used throughout this manual refer to those that are defined or can be defined by users or programs. This source material can be obtained from digital files that can be used in the public domain, private databases, or any digital storage media (DSM). This file can represent a text sequence, image, audio sequence, video sequence, coordinate map, computer program, or any available digital data. Unlike the aforementioned prior art random number generators for chaotic functions, the present invention uses all or part of a digital file. The contents of the digital file can be considered random according to the position of the selected starting position and how the bits are combined. Therefore, the same digital file with different starting positions and combining methods will produce completely different pseudo-random digital sequences. Different digital files with the same starting position and the same combination method will also generate completely different pseudo-random number sequences. This has the obvious advantage that as long as you use the same digital file, the same starting position, and the same combination method, you can regenerate the same quasi-random number sequence. Since this method is not based on any mathematical formula, there is no way to obtain the same random number sequence except knowing the source file, starting position, and combination method. Advantageously, this pseudo-random number sequence is stored for hiding

Page 9 5. Description of the invention (5) (s t e g a η o g r a p h i c) data encoding or decoding method, cryptographic coding or decoding method, or any other intention that requires a random number sequence. In another aspect, the present invention provides an encoding method including the following steps: providing primary data including a first data element arranged; providing secondary data including a second data element; and for each second data element (i) Perform an operation with the first data element, and (ii) generate a golden key element due to the operation. Before performing the operation, in one embodiment, the encoding method includes a step of reconfiguring the first data element of the main data. The technique for reconfiguring the first data element is available and the choice can be determined by this technique. This selection may be random or quasi-random, or determined by a user. This first data element can be reconfigured according to a previously defined or random method. A similar reconfiguration step may be performed selectively or additionally on the second data element of the secondary data. In one embodiment, the primary data is represented in the form of a primary data array containing a first data element and the secondary data is represented in the form of a secondary data array containing a second data element. This encoding method may include re-estimating the primary data array size to match the secondary data array size. If the secondary data array is smaller than the primary data array, the primary data array can be truncated to match the size of the secondary data array. If the secondary data array is larger than the primary data array, the first data element of the primary data array can be repeated to increase the size of the first data array to match the size of the secondary data array. Included in one

Page 10 4 52 73 2 V. Description of the invention (6) In the embodiment of the configuration step and the re-estimation step, the repeated first data element may be based on a technology other than selecting and reconfiguring the first group of the first data element. Technology reconfiguration. In other words, although the first data element of the main data can be multiplied, each multiplied first data element group need not necessarily be reconfigured according to the same technology as the first data element first group. In addition, each repeating group can be reconfigured according to different technologies. The operation performed between the first and second data elements may include a mathematical operation, a logical operation, a corresponding function, or any other operation suitable for generating a key element due to this operation. Better yet, operations are available and choices can be determined through this operation. This choice can be random or pseudo-random, or it can be determined by a user. This encoding method can generate a key element string related to the corresponding second data element string. The unique key data generated from the primary and secondary data can be stored using a complementary decoding method, as described below. It is even better to store this key element in a key file, which can be transferred or archived for future use. More beneficially, information about encoding methods, such as performing calculations, reconfiguration techniques, etc., are also stored in the key file. This information may be stored in the header or attribute block of the key file. An attribute block can be placed anywhere in the key file, and does not need to be at the beginning. The sources mentioned above, primary, secondary and golden spoon data can be represented in digital binary form. However, any data representation or notation used for any convenient collection of symbols can be used, such as alphabetic and numeric symbols, integers

Page 11 452 73 2 V. Description of the invention (7) Numbers and so on. The main data can represent or come from a still image, dynamic video, audio, text or other information style. Similarly, secondary data can represent a still image, dynamic video, audio, text, or other information. In a preferred form of the invention, the secondary data includes a text message and each secondary data element may include an alphabetic and numeric symbol. However, each secondary data element may include a symbol from another symbol set. This letter and number symbol can be used to form a text message. In a typical application of the present invention, the text message may include confidential information about an image, video, and audio sequence contained in the main data. In one embodiment, this text message may include one or more of the following: a title, a master, a copyright holder, a royalty-paid group, and a universal deadline distributed by the publisher. In other embodiments, the text message may be a confidential message, an image representation, an audio sequence representation, or a combination thereof. The main data can represent text messages, still images, audio sequences, dynamic video clips, general multimedia data, coordinate files, complete programs, or any other available digital data, which can be retrieved from public areas, such as Internet sites, Private database, computer's random access memory or buffer memory, or any digital storage medium. The first profile element of the master profile can be reconfigured in an array. Each first data element may define a characteristic related to a still image element. The first data element may be obtained via a data string representing a digitized still image. Images can be obtained from Internet sites, digital cameras, computer games, computer software or other sources. It can be a grayscale or color shadow

Page 12 ^ 52732 5. · Explanation of the invention (8) Images (such as each of them _ _ points) and can be stored in any number of pixels to define a gray level or color into TIFF or JPEG. '° Known formats, such as BMP, GIF, each first data element can be related to the characteristics of the element. Section — $, ground or additionally defines a data string obtained with dynamic video. : The data element can be broadcasted via a digital video station (VC ^) f. Positioned video can be obtained through Internet game ‘computer software, digital multi =’ & laser disc (LD) player ’computer source, and can be stored in a functional video disc (DVD) player or other AVI. Any known format, such as MPEG or each first W ^ i |-7Γ -Γ signal sampling-related characteristics :, number: f optionally or additionally defined-data string with digital audio or music J Bit_sampling can be obtained through a digital audio and video player's digital audio tape (DAT) player, Lei and Ke, ° ° video player (VCD) player or other sources, etc. It is stored in any known format, such as WAV, AIFF, MIDI * Besa * The embodiment digital audio samples can be obtained via two data strings for stereo re-binary digitized sound channels. In the preferred embodiment of the slave code method, the main data includes a static or random image mentioned in the previous three paragraphs. Dynamic video or leather can be used to generate a method based on the above description. Source data to be randomly listed. The: ^ column based on the original image 'video or audio data can be used as the main data of the encoding method of the present invention. In the selected embodiment, the main data can be obtained through a traditional random number.

Page 13 V. Description of the invention (9) Obtained from a word generator or other suitable source. In other aspects, the present invention provides a secondary data decoding method including a second data element, the secondary data is encoded in a key element so that each key element is generated by performing an operation with the first data element of the respective primary data, The method includes the steps of: providing the main data including the first data element arranged; providing the golden key element; and for each golden key element, generating the second data element by performing the inverse operation. Compared with the existing hidden (stega η graphic) or digital watermarking technology, the present invention has obvious advantages in long text * any form of large amount of data, such as video, audio, video, or any binary file, can be secretly encoded And subsequently decoded. For any kind of data, such as video, audio, video, binary file, digital style, the integrity of the main data has never been affected or endangered by any method. For its part, the main data can be transmitted through any device such as mail , Email, phone, fax, file transfer protocol, hyper file transfer protocol, dial-up network, local area network, broadband local area network, internet, internal internet, external internet, or any other electronic device To transmit. Data can also be retrieved from any storage medium, such as hard disk, floppy disk, compressed disk, CD-ROM, digital audio tape (DAT), video CD (VCD), digital versatile video disk (DVD). In a preferred method, since the main data is never modified, there is no need to re-transmit the main data for each message. Only golden spoon data needs to be passed. Therefore, when compared to any existing hidden

Page 14 52 73 2 V. Description of the invention (ίο) (s t e g a η 〇 g r a p h i c) or watermarking technology gains this day ’This method leads to the use of lower bandwidth and faster transmission via communication channels. In alternative embodiments, when opening or storing data, such as the Internet, optical discs, video compact discs (V CDs), digital versatile video discs (DVDs), etc., the use is restricted or restricted to the receiving end of the transmission channel, The main or source data (in whole or in part) can also be transmitted as part of the key file. This embodiment of the invention provides a low level of protection but may be preferred by some users due to its convenience. In order to improve the protection measures of this embodiment, a password or other protection can be implemented in combination with the present invention. This embodiment of the present invention may form part of a large system for transmitting confidential information. In the modified version of the following embodiments, the main or source data (in whole or in part) can be passed as a separate file suitable for identification. BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included and constitute a part of the description of the invention, illustrate embodiments of the invention and serve to explain its principles. However, it should be understood that the drawings and the detailed descriptions below are definitions which are produced for illustrative purposes only and are not intended to be a limitation of the present invention. In the drawings: FIG. 1 shows a background diagram for displaying the exemplary application of the present invention for confidential data transmission; FIG. 2 shows a flowchart of a preferred embodiment of the present invention including a two-part steganographic encoding method; An example of reconfiguration of the main data file by a steganographic encoding method;

Page 15 V. Description of the invention (11) Figure 4 shows an example of a mathematical operation; Figure 5 shows an example of a logical XOR operation between primary and secondary data; Figure 6 shows an example of a 1: 1 correspondence operation; And FIG. 7 shows an example of a steganographic encoding method represented in a password. Description of the Preferred Embodiment The preferred embodiment of the present invention uses source or primary data, such as still images, dynamic video, audio, text, or other data, to steganographically encode secondary data, such as data files containing confidential information . Confidential information can also include still images, motion video, audio, text, or other types of data. This encoding method produces unique key data representing secondary data in encoded form. The detailed description of the embodiment of the present invention includes two main methods as follows. The first main method uses source data, such as still images, dynamic video, audio, text, or other data to generate an array containing a sequence of quasi-random numbers. Quasi-random number arrays are used as the primary data for the second primary method to steganographically encode secondary data. Source data can be provided via a file containing images, video, audio, text or other data. For ease of description, this archive can be classified as a container archive. Similarly, secondary information may be provided for convenience of description by an archive classified as a confidential archive. Key data can also be stored in a file classified as a key file. Referring now to FIG. 1, it is shown that the present invention is used for confidential information via the disclosure

Page 16 4 52 7S ^ _________ Transmission channel, a preferred embodiment of full transmission. The transmitting end 10 executes a steganographic encoding method η in the confidential case 12 in order to generate a unique golden key file 13 which can be transmitted securely via an open communication channel 4. The receiving end 15 of the golden key case 13 executes a complementary decoding method U on the retrieved confidential file 12A. In order to steganographically encode the confidential file 12, either the transmitting end 10 or the encoding method 1 1 must be selected from the Internet 18 8-7 to be downloaded 19 The data file is used as the container file in the encoding method 11. After executing the coding method 11 and generating the gold mark case 13, the transmitting end 10 can transmit the golden key file 13 to the receiving end 15 through the public channel 14. The receiver 15 can then send a request 20 to the Internet 18 to download 21 the same container file at its end and execute the decoding method 16 on the golden key file 13. The transmitting end 10 and the receiving end 15 may have reached an agreement on a special Internet broadcast as a container in the encoding and decoding methods. Golden Key Case 13 can optionally convey information when it finds Internet files for the sender. The container files and confidential files mentioned in the following paragraphs can contain any type of = $: Therefore, we can choose to encode the video file using audio files: sub-encoded image targets, or any other combination. The invention does not restrict the user from using any particular combination. For example, referring to FIG. 2, it is a flowchart showing a more detailed description of the preferred embodiment of the present invention. 30 ~ < ^ partially hidden (31: ^ 311〇21'31) 11 < :) encoding method. Step 2 of this step is related to generating a quasi-random number array based on the source data (container case). Steps 3 3- 3 7 are related to using a quasi-random number array.

Page 17 ^ 52732 V. Description of the invention (13) The second main method of generating key data (gold key file) as the main data to hide (steganographically) encoded secondary data (secret building case). Main method 1 This method generates a quasi-random number array based on a source file containing digital data. In step 30, the digital source file (container file) containing the data bytes is read into a bit array. The source file can be any file type, which contains any type of data, such as audio, video, image, text, etc. In step 31, one of the elements of the bit array is selected as the starting position. This choice can be determined randomly, quasi-randomly, or by a previously defined method. In step 32, the elements of the bit array are reassembled into a new byte (8-bit) group starting from the starting position. In this method, the new group generated represents a pseudo-random number in a sequence that can be stored in the array. This method should be perceptible for digital systems other than binary (ie, binary) based. That is, the digital information transmitted in the source data does not need to be converted into bits. If the information is converted to a decimal system or a hexadecimal number system, the same principle can be applied to generate a new random number. The recombination step performed in step 32 does not require a new group that often reassembles the bits into groups of eight. If a binary system is used and the bit array is recombined into bytes, the random numbers generated will range from 0 to 2 5 5. If reassembled to half of the bits (4 bits), the range will be narrower (0-1 5). For larger ranges, groups can become larger. Correct

Page 18 4 52732 V. Description of the invention (14) For other digital base systems, the size of the selection group can be modified as well. Because this method is not limited to any particular media, users have a very large number of files to choose from and use as container functions. Even when using the same file, there are many possibilities for picking the starting position, and the flexibility of this method allows the user to generate many possible random number arrays. It can therefore provide an effective tool for formatting source data files before they are encoded steganographically. In other words, the above method is a preferred preliminary method to implement before the main method 2 described below. Primary Method 2 This method uses primary data (such as a pseudo-random number array obtained from Step 3 2 of Primary Method 1) to generate golden key data (gold key file) to hide (secondary file) steganographically encoded secondary data. The main file can optionally be obtained through a traditional random number generator or through an image, video, text or other digital data file. In steps 33 of Figure 2, the random number array of the main data is reconfigured to increase the difficulty of cracking the password. Users can reconfigure the random number array to provide a wide choice of technologies to further increase the difficulty of cracking. The choice of reconfiguration technology can be determined randomly. For example, a password can be used as a seed to generate a pseudo-random number (such as using the RAND () function in C) to choose the reconfiguration technique. The user can optionally allow the definition or selection of reconfiguration techniques to be implemented. Reconfiguration techniques can use previously defined or quasi-random methods. Examples include: reverse order configuration, zigzag method, or spiral method to scan in columns and rows. Figure 3 shows a reconfiguration using reverse order configuration 9 9

Page 19

4f ^ 273L V. Description of the invention (15) Example of a typical data string from the container file 38. According to a further example, the spiral method requires that the element is first at the X position, then the element is at the X + 1 position, then the element is at the (X-1) position, then the (X + 2) position, then the (X-2) position, and And so on. If we feel that the application of a random number generator to the source data file is sufficient, the degree of unpredictability is sufficient. The reconfiguration step is unnecessary and can be omitted. In this preferred embodiment, the random number array is reconfigured to produce a higher degree of unpredictability. In step 34, the random data array of primary data can be re-estimated to match the size of the secondary data array of the second data element contained in the confidential file. The random number array can be larger or smaller than the secondary data array. Therefore the random number array is either truncated or repeated to match the size of the secondary data array. Therefore, this step is required regardless of the size of the array or the type of operation that has been performed or will be performed in subsequent steps of this method. In the case where the secondary data array is larger than the random number array, all or part of the random number array is repeated and the repeated random numbers can be reconfigured according to different technologies (at step 3 3). In this method, more random numbers can be provided by the subsequent operations in step 35, as described below. In step 35, at least one operation is performed between the random number array element and the secondary data array element contained in the confidential file. This results in an array of golden keys containing the results of the operation. Because each operation is between at least one random number and at least one secondary data element, the results obtained are different even for similar secondary data elements. For example, given an array of random numbers [3, 5, 2, ...]

Page 20 452? 3 ^ V. Description of the invention (16) and the second data element array [1, 3, 1, ...], and if the selected operation is a random number minus the second data element, the obtained The golden spoon array will be [2, 2, 1, ...]. Because random numbers are used in the encoding method, the first and third elements of the secondary data array are the same but produce different key elements. This is an important advantage of the present invention because it makes cracking passwords more difficult. In addition, the present invention does not restrict the user to select the operation to be performed, which makes cracking even more difficult. Different types of operations can be performed, including the following: (i) Mathematical operations, such as subtraction. An example of this operation is shown in Fig. 4, where the first data element 41 of the random number array is subtracted from the second data element 40 of the confidential data file to generate a golden key element 42. Other mathematical operations can include addition, multiplication, and so on. (ii) Logical operations, such as XOR operations. Such an operation is shown in FIG. 5, where each bit of each second data element 50 and a corresponding bit of each first data element 51 perform an XOR operation to generate a result bit of each key element 52. (i i i) 1: 1 corresponds to the function. An example of such a function is illustrated in Figure 6, where the correspondence is based on the index position specified by the second data element. For example, if the content of the second data element 60 has a value "2", then "2" is used as an index to point 2 of the random number 61. The position 2 of the random number 61 has a value 18 "and this is stored as gold The value of the corresponding key element 6 2 of the key array. The choice of the operation to be performed can be determined randomly. For example, a password can be used as a seed for generating a pseudo-random number (such as using the R A N D () function in C)

P.21 4 5 ο 5. Description of the invention (17) to select the operation to be performed. The user can optionally be allowed to define or choose to perform operations. Referring to steps 35 of FIG. 2 again, the operation result is stored in the golden key array. In step 36, the information about the encoding method is stored in the header or attribute file, which is combined with the golden key array from the golden key file in step 37. The information header and attribute blocks of the golden key file contain all the information needed to perform the complementary decoding method 1f1. Such information may include the physical location of the container file, the starting position of the pseudo-random number generation method, the technology and device for reconfiguring the random number array, and the operations performed. The encoding method can optionally include a cryptographic feature to increase security. The sender can provide a password that is also put into the encoding method. At the other end, the receiving end can prompt for a password and perform the decoding of the encoded password provided by the transmitting end. Only when the decoded password matches the password provided by the sender, the decoding method will start to regenerate the confidential file. This method is illustrated in Fig. 7. One of the password arrays 70 including the password "He 1 loWor Id" is represented by American Standard Code for Information Interchange (ASCII) 72, 101, 108, and the like. The random number 71 is subtracted from these ASCI I codes to create the golden key element 72. These golden key elements are stored in the attribute block of the golden key file. It should be understood that the data transmission application shown in FIG. 1 may or may not include the two-part encoding method shown in FIG. 2. For example, the first main method for generating a pseudorandom number array of container files can be omitted. In this case, the container file can be used as the main data in the encoding method to replace the random number array. Moreover, it should be understood that in the coding method, the main method 2,

Page 22 452732 V. Description of the invention (18) The reconfiguration and re-estimation steps are unnecessary and can be omitted. Complementary decoding methods are self-evident to the information presented by those who are skilled in the technology from the information that exists here. Therefore, the decoding method need not be described in detail. Clearly, the key part of the decoding method is to perform the inverse operation performed in the encoding method. If the reconfiguration and resizing of the main data (ie, random number array) has been performed in the encoding method, the details must be stored in the attribute block of the key file or elsewhere so that the inverse operation can be performed in the decoding method. Similarly, if a random number array has been generated from the source data file used for Main Method 1, the same random number array must be regenerated from the source data file used to decode the key file. Advantages of the invention (A) The size of the secondary data is not limited compared to the existing hidden (stega η graphic) or watermark technology. The present invention has different advantages, that is, long sentences, any type of data such as images, audio, video, etc. A large amount of information in the carry file can be encoded (masked) and subsequently secretly decoded. (B) No distortion of primary or secondary data For any type of data such as video, audio, video, binary file, digital mode, etc., the integrity of the primary or secondary data has never been affected or Endangered. In other words, this decoding technique is lossless. The main information can be freely used in any type such as via mail, telephone, e-mail, fax, file transfer protocol, hyper file transfer protocol, dial-up network, broadband LAN, Internet, internal Internet, external

Page 23 452732 V. Description of Invention (19) Internet, or any other electronic device. Data can also be transferred via any storage medium such as hard disk, floppy disk, compressed disk, digital audio cassette (DAT), optical disc, video disc (VCD), laser disc (LD), digital versatile disc (DVD) Capture. The invention has a significant advantage over traditional methods, such as the least significant bit (LSB) decoding, which adds distortion to the burden on the data, so the entire container file must be transmitted. In addition, LSB decoding allows only high-bit depth container files to be used, so this is not suitable for most multimedia materials. (C) Lower bandwidth usage and faster transmission In the better method, since the main data is never modified, there is no need to send or resend the main data of each message. Only golden key files need to be sent. This results in reduced storage space compared to traditional methods that require the entire container file to be transferred. Therefore, compared with any existing hidden (ste g a η 〇 g r a p h i c) or watermarking technology, this method results in lower bandwidth usage and faster transmission under the communication channel. (D) Unrestricted primary and secondary data types. Existing steganographic and watermarking techniques often have problems with low-bit-depth bitmaps (such as black and white images), low-bit-depth audio, and video files. This is usually because changing the least significant bit of the low bit depth will change the problem of too much original information. This restricts the existing hiding (ste g a η 〇 g r a p h i c) or watermarking techniques that are only applicable to large-bit depth slots, such as 24-bit bitmaps. However, since the present invention maintains the integrity of both the primary and secondary data, it does not suffer from this problem and can be used with any primary or secondary data type.

Page 24 452732 V. Description of the invention (20) (E) The only golden key data generated The invention disclosed above has other advantages, that is, the golden key data generated even under the same primary and secondary data Often different and unique. This makes it almost impossible for any illegal user to use the key data generated by the analysis to crack the password. (F) Different reconfiguration techniques Many reconfiguration techniques can be implemented in the present invention. This means that unscrupulous users must try all reconfiguration techniques in order to crack the password. It is assumed that cracking a single technology is already a very difficult task, and cracking a password is practically impossible. (G) Unlimited amount of primary data available For the phenomenal growth of Internet communications, the number of primary data files available on the Internet is virtually unlimited. Therefore, legitimate users can choose images, audio, video, or any digital binary file on the Internet as the main data. Therefore, without the knowledge of the main data, illegal users must try an infinite number of images, audio and video files before they can start the cracking task. (Ii) Password protection and garbage-in-garbage-out system The present invention includes a passwordless protection system for useless input and useless output. The password can be used to generate the starting position of the random reconfiguration method and / or primary data and / or secondary data. Because this is designed as a useless input and useless output system, it does not need to give any clues as to whether the password is invalid or the master data is invalid. Therefore, even if the unscrupulous user manages to get the master data file

Page 25 V. Description of the invention (21), this is already very difficult. Continuously failing to crack the key file and obtaining different passwords may finally lead the illegal user to think that this main data file is not the correct file. (I) Generate new container files. The only major data file known to legitimate users can be easily generated. Examples could be digital images of legitimate users, audio voices of legitimate users, and video folders of legitimate users. Typical Application of the Invention In one embodiment, the present invention can be used for confidential data communication. In the preferred method, the key data generated by the main data can be determined in advance. Route, internal internet, external internet, or any other electronic device to legitimate users. We found that the size of the key file to be transmitted is almost equal to the size of the actual message, usually within the upper limit of 10 bytes. In another embodiment, the present invention can be implemented as a plug for Internet website browsing Device, email program, graphics program, document program or any other computer program so that confidential files can be hidden and only sent to legitimate users. In another embodiment, software developers who want to protect their data can also apply the disclosed invention. For example, in Microsoft Word, the program can use the password and the file itself to hide the original data. Therefore, even if other programs can open Microsoft Word documents, the opened files will still be displayed as unknown data. In the same way, this embodiment can be extended to other processes

Page 26 452? 3c V. Description of invention (22), such as e-mail programs such as E X c h n n eTM or drawing software such as AutoCAD®. In another embodiment, the present invention can be used to transmit message modification detection as a data verifier. When the digital identification mark on the transmitting end can be regarded as the secondary data, the transmitting message in this example can be regarded as the primary message, and vice versa. On the received message, the receiver can decode it to detect whether the actual sender transmitted it and check if the message has been modified. In another embodiment, the confidential information or identification code may be stored on a credit card, passport, ID card, debit card, or any device where both primary and secondary data are present. For example, in the case of a credit card, when the information about the owner, his / her account number or identification code can be regarded as secondary information, the biometrics (such as photos, fingerprints, sounds, etc.) of the credit card owner can be used as the main information, and vice versa The same is true. In such cases, if an attempt is made to change the biometrics of the credit card owner, decoding the confidential information or authentication code will not match. In another embodiment, this technique can be used to generate a digital watermark of any digital image, text, audio, video or other digital data. When the digital watermark can be considered as secondary data (confidential file), this image, text, audio, video or digital data can be considered as primary data (container file). In the encoding method, a golden key file will be generated based on this disclosed invention. The legitimate user will have a unique golden key file and he can use it to decode the digital watermark from the main data and verify the source of the main data from there. In another embodiment, part of the current invention (main method 1 described above) can be used in the field of cryptography. In cryptography, container files are not used

Page 27 45 ^ 732 ^ Case No. 88119232, "Month and Month Amendment Year / Month / Day Amendment / Correction / Supplement" V. Description of Invention (23) Steganography case. Instead, a scramble function is used to decode the encrypted message. The chaos function can be a cipher string known only to the transmitting end and the receiving end or a very large prime number. Therefore, the quasi-random number sequence generated using the main method 1 can be used instead of any chaotic function. Furthermore, given that many possible digital files can be used in public and secret areas, and it is easy to create new digital files, it would be extremely difficult if not impossible to break the pseudo-random number array. In another embodiment, the current invention can also be applied complementaryly in the field of cryptography. Using the current invention, either the chaotic function or the encrypted message can be encoded and subsequently decoded for added security. The golden key file generated using the current invention can optionally be encrypted before being transmitted to the transmitting end and then decrypted before being steganographically decoded. It is expected that the present invention will be shaped and implemented in software on a general-purpose computer platform. The present invention may optionally be implemented using hard-wired circuit systems, central processing units (CPUs), and digital signal processing (DSPs) to incorporate one or more application specific integrated circuits (ICs). Furthermore, it is expected that the present invention can be embedded in a facsimile, a telephone, a digital camera, a portable radiotelephone or other electronic messaging devices to give the ability to encode and decode confidential information. Finally, these technical capabilities will be able to detect changes in different adaptability and the modifications of the preferred embodiment just described can be assembled without departing from the scope and spirit of the invention. Therefore, what needs to be understood is that within the scope of the additional patent application, the present invention can be practiced except as specifically described herein. Schematic component symbol description I 0 Transmitter II Hidden coding method

61155.ptc Page 28 2001.06. 27.028 々52 7 Case No. 88119232 € Month Day Amendment V. Description of Invention (2¾) 12, 1 2A 13 1 4 15 16 17 18 19 20 2 1 30-37 38 39 40 41 42 50 5 1 52 60 6 1 62 70 7 1 72 Confidential file key file public communication channel Receiver decoding method selection Internet to download data file Request download step Container reverse order Second first key Second second key Second Random Key Password Random Key File Configuration Tribute Element Data Element Data Element Data Element Data Element Data Element Digital Element Array Digital Element

61155.ptc Page 28a 2001.06.27.029

Claims (1)

2. Application for Patent Fanyuan 1. First, the method of generating a random number sequence includes the following steps: mention: Source data of multiple arranged data elements, each data TL element inner valley is represented by a number group Read the group of numbers _ _ 1 Λ — in the car so that each position of the array contains one of the # numbers; select within the number array — phase ^ Λ Α starting position: and ρ heavy numbers to A new group of numbers constituting the reference starting position, such that :::: number subgroups represent a quasi-random number and consecutive new groups of numbers represent the quasi-random number sequence. In addition, it includes the steps of storing the proposed data element in which each data element is binary, and each new number group, where the starting position can be followed, and the starting position can be used. 2. Such as the method of applying for the scope of patent application. 3. If the method of the first scope of patent application is applied, the marking method is not used. 4. The method for applying for item 3 of the patent scope includes eight binary numbers. 5. If the method of applying for the scope of the first item of the patent scope is selected on the spot or randomly. 6. Select the method as defined in the first patent application scope. 7. — A coding method generates a quasi-random number sequence using a method such as the first item in the scope of patent application. 8. — A decoding method uses a method such as the sphere item in the patent application to generate a pseudo-random number sequence. 9. An encoding method comprising the following steps: O: \ 61 \ 61155.PTD Page 29..1 52 73 2 VI. Patent application scope Provide source data including arranged data elements: Provide secondary data including second data elements; and for each second data Element (i) performs an operation with the first data element, and (ii) the operation generates a gold element. 10. If the coding method of item 9 of the scope of patent application, before performing the operation, includes the following steps: Reconfigure the first data element of the main data. 11. The coding method as described in item 10 of the patent application scope, wherein a technique for reconfiguring the first data element is available and at least one choice can be determined by these techniques. 12. As for the coding method of item 11 in the scope of patent application, one or each of these choices can be randomly determined to be randomly determined. 13 3. As for the coding method of item 11 in the scope of patent application, one or each of the selections can be determined by a previously defined method. 14. The encoding method according to item 11 of the scope of patent application, including the following steps: storing the golden key in the golden key file; and storing the attribute block of one or each selected relocation technology in the golden key file. 1 5. The coding method of item 10 in the scope of patent application, wherein the first data element can be reconfigured by a previously defined method. 16. The coding method according to item 10 of the patent application range, wherein the first data element can be reconfigured by a random or quasi-random method. Page 30 4 52 73 6. Scope of patent application 1 7. If the coding method of item 9 of the scope of patent application is applied, the following steps are performed before performing the operation: Reconfiguring the second data element of the secondary data. 1 8. The coding method according to item 9 of the scope of patent application, wherein the main data is in the form of a primary data array containing a first data element and the secondary data is in the form of a secondary data array containing a second data element To represent, and includes the following steps: re-evaluate the size of the primary data array to match the secondary data array, and estimate the items in the French code that are in the new weight. The next column includes the fruit package, such as the main column, the short column, the main column, and the double column Wh. The items that are coded in French code are 9 11th. Fan Li specially invited the first technique in the first line and the first line in the field. It is necessary to have the master in addition to the new or heavy skills. One by one, the data should be rooted and included before, and the elements should be stepped on. These elements of the item number 9 of the first round of the second round of the standard replenishment and special allocation of new items, please reiterate the technical support 1 I 2 And France with a first resource in the first two packet count comprises per its PBC ^. And the calculation of the number of characters in the Yundai school is based on the element material data item 9 of the table. With 1 inclusive calculation " * 11 Page 31 452732 6. Scope of patent application Logical operation between data elements. 2 3. The coding method according to item 9 of the scope of patent application, wherein the first and second data elements are represented by numbers and each operation is a corresponding function. 2 4. The coding method of item 9 in the scope of the patent application, wherein the first and second data elements are represented by numbers and each operation is a 1: 1 correspondence function, wherein the content of each second data element is used to select the first An index of a data element and the content of each selected first data element is designated as a corresponding key element. Among them, the operation can be made 1 where the selection can be made 1 where the selection can include the storage of golden keys 1 including the storage of 2 5. As for the encoding method used in the 9th patent application, the selection can be determined by the operation. 26. The coding method as described in item 25 of the scope of patent application is randomly or quasi-randomly determined. 27. The encoding method as described in item 25 of the patent application scope is determined by the previously defined method. 2 8. The encoding method as described in item 9 of the patent application is the same as the key file file. 29. The coding method according to item 28 of the scope of patent application. The steps of coding method information in the attribute block of the key file. 30. The coding method of item 29 in the scope of patent application, wherein the information stored in the attribute block includes the operation or the operation performed. 3 1. The coding method according to item 28 of the patent application scope, including the step of storing the main data in a key file. 3 2. The coding method according to item 9 of the scope of patent application, wherein the main information includes a quasi-random number sequence generated by the method like item 1 of the scope of patent application Page 32 45273 £ 6. Scope of patent application. 3 3. The coding method according to item 9 of the scope of patent application, wherein the main data includes a random number sequence generated by a random number generator. 3 4. The coding method according to item 9 of the scope of patent application, in which the main information is provided by files obtained from the Internet. 35. The encoding method according to item 34 of the patent application scope, including the following steps: storing the key element in the key file; and storing information about the Internet file in the attribute block of the key file. 36. The coding method according to item 9 of the scope of patent application, wherein the secondary data includes a text message and each second data element includes a symbol from a symbol set. 37. The coding method according to item 9 of the scope of patent application, wherein the first data element is arranged in an array and each first data element represents a characteristic of a relative digital audio sample. 38. The coding method according to item 9 of the scope of patent application, wherein the first data element is arranged in an array and each first data element represents a characteristic of a relatively still image element. 39. The coding method according to item 9 of the scope of patent application, wherein the first data element is arranged in an array and each first data element represents a characteristic of a relatively dynamic video element. 4 0. — A method for decoding secondary data, including a second data element, the secondary data is encoded in a key element so that each key element can be related to 45273t on page 33 6. The main data of the patent application scope is generated by the operation performed on the first data element. The method includes the following steps: providing the main data including the arranged first data element; providing the key element, 'for Each golden key element generates a corresponding second data element by performing the inverse operation. 41. The method of claim 40 in the scope of patent application, wherein during the compilation of the secondary data, before performing this operation, the first data element is reconfigured according to a defined technique, and before the second data element is generated, the The method includes the following steps: The first data element of the master data is reconfigured according to the defined technology. 42. The method according to item 41 of the scope of patent application, wherein the key element is provided by a key file with an attribute block, and this attribute block contains information about the data element to be used for the first data element during the encoding of the secondary data. Information of a defined technology, the method comprising reading information from an attribute block to determine the defined technology. 4 3. The method according to item 40 of the scope of patent application, wherein during the encoding of the secondary data, the size of the primary data is re-estimated to match the size of the secondary data. The re-estimated size performed during encoding is used to re-estimate the primary data size. 4 4. The method according to item 43 of the scope of patent application, in which the primary data is re-estimated by shortening the primary data during the encoding of the secondary data. , The method includes the following steps before generating the second data: Page 34 ^ 273 VI. Scope of patent application If the main data is shortened during the coding period, the main data will be shortened according to the truncation instruction executed during the coding of the secondary data; if the main data is repeated during the coding period, the main data will be coded according to the secondary data. Repeated instructions executed during the period repeat the main data. 4 5. The method according to item 44 of the scope of patent application, wherein during the encoding of the secondary data, the first data element of the primary data is reconfigured according to the first technology and according to the first technology or in addition to the first The non-technical technology repeats the first data element. Before generating the second data element, the method includes reconfiguring the first data element of the main data according to the first technology and repeating the first data element according to the first technology or the other technology Steps for the first data element. 46. The method of claim 40, in which the key element is provided in a key file with an attribute block and the attribute block contains information about operations performed during the encoding of the secondary data, the method includes from this attribute The block reads this information to determine the inverse operation for each key element. 4 7. The method according to item 40 of the scope of patent application, wherein during the encoding of the secondary data, the main data is provided by a file obtained from the Internet, and the key element is provided by The key file of the attribute block is provided, and the method includes reading the information from the attribute block to retrieve the Internet file. 48 8. The method as claimed in item 40 of the scope of patent application, wherein the main information includes a quasi-random number sequence generated as in item 1 of the scope of patent application. Page 35