Generative steganography embedding method, extraction method, embedding system and extraction system based on manual construction

Abstract

The invention discloses a generative steganography embedding method, extraction method, embedding system and extraction system based on a chess manual structure. The embedding method comprises the following steps: making a non-repeated opening chess manual library, converting to-be-embedded information into binary information by a sender, extracting an opening chess manual with a corresponding serial number from the chess manual library according to the length of the binary information, executing and recording the opening chess manual into an encrypted chess manual; the intelligent chess player continues playing chess, the binary information is embedded into each step of playing chess according to a preset conversion rule, each step of chess piece falling operation is recorded into an encrypted chess manual, and the encrypted chess manual is sent to a receiver after all the information is embedded; and the receiver reversely extracts the embedded information from the received password-containing chess manual according to the conversion rule. According to the method, the steganography capacity is improved, the secrecy protection is enhanced, efficient and rapid embedding and extraction are realized, and meanwhile, the integrity and effectiveness of the ciphertext can be verified.

Description

Technical Field

[0001] The present invention belongs to the field of information steganography, and more specifically, relates to a generative steganography embedding method, an extraction method, an embedding system and an extraction system based on chess record construction. Background Art

[0002] With the development of information technology, the widespread use of digital information transmission and Internet media has led to an increasing demand for information security and privacy protection. In the field of basic cryptography, traditional embedding and extraction technologies can no longer meet the current high demand for information security. Therefore, steganography has emerged as a new technical means. Unlike cryptography, steganography can embed secret information in an information carrier that is not perceptually abnormal and send it to the other party, and it is not easy to arouse the suspicion of attackers, which ensures communication security to a certain extent.

[0003] With the advancement of artificial intelligence and machine learning technology, generative steganography has emerged. This technology provides a more secure means of transmitting and sharing secret information. It can not only embed information into a carrier almost unnoticeably, but also resist a certain degree of steganalysis. This is crucial to maintaining the confidentiality of information transmission and the integrity of the carrier.

[0004] Reinforcement learning technology, as a machine learning technology that learns strategies by interacting with the environment, has achieved remarkable results in the field of chess games. Take the AlphaZero technology as an example, which can combine the Monte Carlo tree search algorithm with neural networks and train high-level computer chess players through continuous self-playing, thereby generating chess records with rich strategies, which provides a new development direction for generative steganography.

[0005] The invention patent with the prior art publication number CN117745507B proposes a generative steganography method and system based on chess record construction, the method includes converting the secret text information to be hidden into a secret binary sequence; the real player drops a first type of chess piece, the machine player generates a number of candidate positions according to the position of the first type of chess piece, and calculates the score of the candidate position according to the preset scoring rule; the candidate positions exceeding the candidate score line are screened out, and the candidate positions exceeding the candidate score line are sorted according to the first sorting rule; the first n bits of the secret binary sequence are taken and put into the coding set for search, and the binary number code in the coding set that is the same as the first n bits of the secret binary sequence is matched; and the chess record containing secret image is obtained. This scheme requires the player to operate in real time, the embedding and extraction speed is slow, and the message integrity and validity cannot be verified. Summary of the invention

[0006] In order to overcome the problems in the prior art of slow embedding and extraction speed and inability to verify message integrity and validity, the present invention provides a generative steganographic embedding method, extraction method, embedding system and extraction system based on chess record construction.

[0007] The primary purpose of the present invention is to solve the above technical problems. The technical solutions of the present invention are as follows:

[0008] The first aspect of the present invention provides a generative steganalysis embedding method based on chess record construction, comprising the following steps:

[0009] Obtain a preset formulaic opening chess game database and a preset intelligent chess player, wherein each chess game in the chess game database has the same number of steps and is marked with serial numbers in sequence;

[0010] Convert the information to be embedded into binary information and calculate the length of the binary information;

[0011] Start embedding, obtain an opening chess record with a sequence number equal to the length of the binary information from the opening chess record library, start a new game, execute the opening chess record and record it into the encrypted chess record;

[0012] The intelligent chess player is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, and perform move operations according to the remaining moves after deletion and the binary information to be embedded and record them in the encrypted chess record;

[0013] Repeat the above game embedding operation until the secret information is steganographically written and the final secret chess record is obtained.

[0014] Furthermore, the preset intelligent chess player is trained by AlphaZero, and the parameters of the intelligent chess player are fixed during the embedding and extraction process. The method for constructing the preset formulaic opening chess database is: calling the preset intelligent chess player and letting it play automatically. Each game only records the first k steps of the opening. During the automatic game process, each move is randomly selected from the top n moves in probability ranking. The obtained chess records are numbered in ascending order from 1 and stored in the opening chess database.

[0015] Furthermore, it is necessary to ensure that the length of the binary information to be embedded is less than or equal to the number of opening game records in the opening game record library.

[0016] Furthermore, the specific steps of deleting some moves in the move set are:

[0017] Find the move operation with the highest probability in the move operation set and record its probability value as maxprob;

[0018] If maxprob is greater than or equal to the first preset value, only the move is retained and other moves are deleted;

[0019] If maxprob is less than the first preset value and greater than or equal to the second preset value, then delete all moves that satisfy a probability less than a third preset value, where the first preset value is greater than the second preset value and the third preset value is less than maxprob;

[0020] Delete all moves that satisfy the probability value of 0.

[0021] Furthermore, the specific steps of performing the move operation according to the remaining moves after deletion and the binary information to be embedded and recording them in the secret chess record are as follows:

[0022] If the number of remaining moves m is 1, directly execute the move and record it in the secret chess record; otherwise, sort the remaining moves by probability from large to small, and number them in ascending order from 0;

[0023] Extract the length of the binary information to be embedded. , convert it into a decimal number, select the corresponding number of moves from the currently available set of moves, execute and record it in the secret chess record, if the length of the remaining binary information to be embedded is less than Then all the remaining contents of the binary information to be embedded are intercepted.

[0024] Furthermore, if the game ends during the information embedding process, it is necessary to start the next game and embed the remaining binary information.

[0025] The second aspect of the present invention provides a generative steganalysis extraction method based on chess record construction, comprising the following steps:

[0026] Obtain a preset formulaic opening chess game database and a preset intelligent chess player, wherein each chess game in the chess game database has the same number of steps and is marked with serial numbers in sequence;

[0027] Receive a secret chess record, extract a formula opening part from the secret chess record, and search the corresponding opening chess record serial number in the formula opening chess record library to obtain the length of information. If the received secret chess record contains multiple secret records, extract them in descending order according to the size of the opening chess record serial number;

[0028] Using the intelligent chess player to start a new game and execute the opening chess record;

[0029] For the remaining moves in the chess record, the intelligent chess player is used to continuously and automatically play chess. In each step of the game process, the intelligent chess player calculates the legal move set in the current situation and determines the probability of each move, deletes some moves in the move set according to the probability of each move, extracts the moves according to the remaining moves after deletion and the moves in the secret chess record, executes the move operation in the secret chess record and writes the obtained binary number into the result file;

[0030] The game extraction process is repeated until all moves in the secret chess record are extracted.

[0031] Furthermore, the specific steps of extracting the remaining moves after deletion and the moves in the secret chess records are:

[0032] If the number of remaining moves m is 1, execute the move operation directly; otherwise, sort the remaining moves according to probability from large to small, and number them in ascending order starting from 0;

[0033] Take a move from the secret chess record, find the same move in the move set, convert its serial number into a binary number, if the length of the binary number is not enough Then add 0 at the beginning until the length is equal to Write the obtained binary number into the result file and execute the move operation of the secret chess record;

[0034] When extracting the last move of the encrypted chess record, let the binary sequence obtained after extracting this move be bin. If the length of the binary sequence in the result file plus the length of bin is less than the number of the starting chess record, it is necessary to add a number of leading 0s in bin until the length of the binary sequence in the result file plus the length of bin is equal to the number of the starting chess record. Write bin into the result file, and all the original embedded information is extracted.

[0035] The third aspect of the present invention provides a generative steganographic embedding system based on chess record construction, comprising the following modules:

[0036] A move operation module is used to execute a move according to a received move operation instruction;

[0037] The serialization module is used to record the moves into the encrypted chess records;

[0038] The formula opening chess record module is used to obtain a formula opening chess record library. Each chess record in the chess record library has the same number of steps and is marked with serial numbers in sequence. When embedding starts, the opening chess record with a serial number equal to the length of the binary information to be embedded is obtained from the opening chess record library, a new game is started, and the move operation module is notified to execute and record using the serialization module;

[0039] The intelligent chess player module is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, decide the move operation according to the remaining moves after deletion and the binary information to be embedded, notify the move operation module to execute and use the serialization module to record.

[0040] The fourth aspect of the present invention provides a generative steganalysis extraction system based on chess record construction, comprising the following modules:

[0041] The deserialization module is used to convert the chess records into corresponding move operation instructions;

[0042] A move operation module is used to execute a move according to a received move operation instruction;

[0043] The formula opening chess record module is used to obtain a formula opening chess record library. Each chess record in the chess record library has the same number of steps and is marked with serial numbers in sequence. When the extraction starts, the formula opening part is extracted from the secret chess record, and the move operation module is notified to execute the opening chess record and search for the corresponding opening chess record serial number in the formula opening chess record library to obtain the length of the information. If the received secret chess record contains multiple, the opening chess record serial numbers are sequentially extracted in descending order according to the size of the opening chess record serial numbers;

[0044] The intelligent chess player module is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, extract the remaining moves after deletion and the moves in the secret chess record, notify the move operation module to execute the moves in the secret chess record and write the obtained binary number into the result file.

[0045] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0046] The present invention utilizes the setting of the opening chess record to transmit the length information of the ciphertext. The setting of the opening chess record makes the chess game more reasonable and more confidential. The length information of the ciphertext can be used to verify the integrity and validity of information transmission and also play a role in identifying the sequence of the chess game, thereby expanding the steganographic capacity. In addition, both players are intelligent chess players based on reinforcement learning, the game quality is high, and the embedding and extraction speeds are fast. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to make the purpose and technical solution of the present invention clearer, the present invention provides the following drawings and descriptions:

[0048] Figure 1 A flow chart of a generative steganographic embedding method based on chess record construction provided by an embodiment of the present invention;

[0049] Figure 2A flow chart of a generative steganalysis extraction method based on chess record construction provided by an embodiment of the present invention;

[0050] Figure 3 An opening chess diagram provided by an embodiment of the present invention;

[0051] Figure 4 A schematic diagram of an embedded system provided by an embodiment of the present invention;

[0052] Figure 5 A schematic diagram of an extraction system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0054] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0055] The first aspect of the present invention provides a generative steganographic embedding method based on chess record construction, such as Figure 1 The figure shows a flow chart of a generative steganographic embedding method based on chess record construction. Compared with some existing constructive steganographic methods based on game behavior, such as Gobang, this embodiment uses Chinese chess, which has richer strategies and possibilities, and thus has a larger steganographic capacity. When transmitting the secret chess record, the sender and receiver of the information can arbitrarily choose to transmit the secret chess record in the form of text, image or video on the public channel. The specific steps are as follows:

[0056] S1: Obtain a preset formulaic opening chess game database and a preset intelligent chess player, wherein each chess game in the chess game database has the same number of steps and is marked with serial numbers in sequence.

[0057] More specifically, the preset method for making an opening chess record is to call an intelligent chess player with preset fixed parameters, let it play the game automatically and record the results. During the automatic game process, each move is randomly selected from the top 5 moves with the highest probability in this step, in order to ensure the rationality of the game. Each opening chess record contains the first k steps of the game, where k should be as small as possible. In a specific embodiment, the value of k is 5. Let the intelligent chess player play 1,000 games automatically. For each game, only the first 5 steps of the opening are recorded, and each step selects the top 5 moves with the highest probability among all currently feasible moves. Finally, these 1,000 games are numbered starting from 1, such as Figure 3 shown.

[0058] More specifically, the preset intelligent chess player is trained by AlphaZero, with high game quality and fast speed of generating each move. AlphaZero is a revolutionary artificial intelligence program developed by DeepMind. After being launched in 2017, it quickly surpassed the top human level in chess games such as Go, chess and shogi. At its core is a deep neural network that learns the rules of the game from scratch through reinforcement learning, that is, playing against itself, and gradually improves its strategy.

[0059] AlphaZero's technical architecture includes deep neural networks and reinforcement learning technology. It uses Monte Carlo Tree Search (MCTS) to predict possible moves and results, and evaluates game positions through neural networks to make decisions. AlphaZero's learning method is different from traditional artificial intelligence systems. It does not rely on the knowledge of human experts or pre-programmed strategies, but continuously improves through self-play and experience. The parameters of the intelligent chess player are fixed during the embedding and extraction process.

[0060] S2: Convert the information to be embedded into binary information and calculate the length of the binary information. It is necessary to ensure that the length of the binary information to be embedded is less than or equal to the number of opening chess records in the opening chess record library.

[0061] More specifically, the information sender converts the information to be embedded into binary information, takes out the opening chess record with the corresponding serial number from the chess record library according to the length of the binary information, executes the chess record and records it into the secret chess record. Thus, the serial number of the opening chess record plays the role of transmitting the length of the binary information, making the chess game more reasonable and more confidential. The length information of the ciphertext can be used to verify the integrity and validity of the information transmission.

[0062] S3: Start embedding, obtain an opening chess record with a serial number equal to the length of the binary information from the opening chess record library, start a new game, execute the opening chess record and record it into the secret chess record. In a specific embodiment, the input secret information is converted into binary as "101100011001", and its length is 12. Find the opening chess record with a serial number equal to 12 from the opening chess record library: ["two soldiers advance one", "horse 8 advances 9", "horse two advances three", "4 soldiers advance 1", "horse eight advances seven"].

[0063] S4: using the intelligent chess player to calculate the legal move set in the current situation and determine the probability of each move, deleting some moves in the move set according to the probability of each move, performing move operations according to the remaining moves after deletion and the binary information to be embedded and recording them in the secret chess record, if the chess game ends during the information embedding process, it is necessary to start the next game and embed the remaining binary information, at this time, the sequence number of the opening chess record plays the role of identifying the sequence of the chess games, the larger the value, the earlier the chess game is in the sequence, expanding the capacity of steganography, and after all the information is embedded, the secret chess record set is sent to the information receiver.

[0064] More specifically, the specific steps of deleting some moves in the move set are:

[0065] The intelligent chess player calculates the legal chess move set L = {x1, x2, ..., x n}, and calculate the probability of each move according to the degree of goodness ∈ i (1≤i≤n);

[0066] Find the move with the highest probability in L and record its probability value as maxprob;

[0067] If maxprob ≥ 0.8, only keep this move and delete the others;

[0068] If 0.3≤maxprob<0.8, delete all i <0.1 drop;

[0069] Delete all the items that satisfy ∈ i =0.

[0070] The specific steps of performing the move operation based on the remaining moves after deletion and the binary information to be embedded and recording them in the secret chess record are as follows:

[0071] If the number of remaining moves m is 1, directly execute the move and record it in the secret chess record; otherwise, sort the remaining moves by probability from large to small, and number them in ascending order from 0;

[0072] Extract the length of the binary information to be embedded. , convert it into a decimal number, select the corresponding number of moves from the currently available set of moves, execute and record it in the secret chess record, if the length of the remaining binary information to be embedded is less than Then the entire contents of the remaining binary information to be embedded are intercepted. In a specific embodiment, the intelligent chess player calculates all feasible sets of moves on the current chess board and gives probabilities respectively, as shown in Table 1.

[0073] Table 1

[0074]

[0075]

[0076] According to the rules, delete the last move "5 soldiers move 1". Now there are 8 optional moves left, which means the length of the secret information that can be hidden this time is At this time, the secret information is "010100011001", so the first 3 digits are cut off, that is, "010", and converted into a decimal number of "2". Therefore, from the remaining optional moves, the move with the second highest probability is selected. Note that the count needs to start from 0, and the move ranked second is "Rook 9 level 8". Finally, "Rook 9 level 8" is executed and the secret chess record is written. At this time, the content of the secret chess record is: "Two soldiers advance one, horse 8 advances 9, horse two advances three, 4 soldiers advance 1, horse eight advances seven, rook 9 level 8". Continue to embed the game in this way, and finally get the secret chess record "Two soldiers advance one, horse 8 advances 9, horse two advances three, 4 soldiers advance 1, horse eight advances seven, rook 9 level 8, horse seven retreat eight, 2 soldiers advance 1, rook nine advance one" and send it to the information receiver.

[0077] S5: Repeat the above game embedding operation until the secret information is steganographically written and the final secret chess record is obtained.

[0078] The second aspect of the present invention provides a generative steganalysis extraction method based on chess record construction, such as Figure 2 The figure shows a flow chart of a generative steganalysis extraction method based on chess record construction, which includes the following steps:

[0079] S1: Obtain a preset formulaic opening game database and a preset intelligent chess player, wherein each game in the database has the same number of steps and is marked with serial numbers in sequence;

[0080] S2: receiving a secret chess record, extracting a formula opening part from the secret chess record, and searching for a corresponding opening chess record serial number in the formula opening chess record library to obtain the length of the information. If the received secret chess record contains multiple secret records, extracting them in descending order according to the size of the opening chess record serial numbers;

[0081] S3: using the intelligent chess player to start a new game and execute the opening chess record;

[0082] S4: For the remaining moves in the chess record, the intelligent chess player is used to continuously and automatically play chess. In each step of the game process, the intelligent chess player calculates the legal move set in the current situation and determines the probability of each move. According to the probability of each move, some moves in the move set are deleted. If the number of remaining moves m is 1 at this time, the move operation is directly executed; otherwise, the remaining move operations are sorted from large to small according to the probability, and are numbered in ascending order starting from 0;

[0083] Take a move from the secret chess record, find the same move in the move set, convert its serial number into a binary number, if the length of the binary number is not enough Then add 0 at the beginning until the length is equal to Execute the chess move operation containing the secret record and write the resulting binary number into the result file.

[0084] S5: Repeat the game extraction process until all moves in the secret chess record are extracted. When extracting the last move of the secret chess record, set the binary sequence obtained after the extraction of this move as bin. If the length of the binary sequence in the result file plus the length of bin is less than len, it is necessary to add a number of leading 0s of bin until the length of the binary sequence in the result file plus the length of bin is equal to the sequence number of the opening chess record. Write bin into the result file, and extract all the original embedded information. In a specific embodiment, the received secret chess record is: "2 pawns advance one, horse 8 advances 9, horse 2 advances three, 4 pawns advance 1, horse 8 advances seven, chariot 9 level 8, horse 7 retreats eight, 2 pawns advance 1, chariot 9 advance one". Extract the first 5 moves, that is, "2 pawns advance one, horse 8 advances 9, horse 2 advances three, 4 pawns advance 1, horse 8 advances seven", and find this opening chess record from the opening chess record library. Its sequence number is 12, indicating that the complete length of the secret information is 12. Then, the intelligent chess player is used to continue playing the game, analyze the current situation, obtain all feasible moves, and give the probability of each move. Since the parameters of the intelligent chess player are fixed, the results in Table 1 above can still be obtained under the same situation. It is found that there are still 8 optional moves left, and the probability of this move being able to hide The first move in the secret chess record, "Rook 9, Level 8", is located at the "2nd" position of all the currently available moves. Convert it to a binary number of "10". Since this step can write a total of 3 bits of binary data, add 0 to the left of "10" until the length of 3 bits is met, and the result is "010". Write it into the extraction result, and the extraction result is "010". Continue to play and extract. When the last move of the secret chess record, "Rook 9, Advance 1", has not been extracted, 10 bits of binary data "0101000110" have been extracted. For the last step "Rook 9, Advance 1", the binary data obtained by extraction is "1", but if it is directly added to the extraction result, it will be found that the total length of the result is 11 bits, which is contrary to the known total length of the secret information of 12 bits. Therefore, it is necessary to add a 1-bit "0" before the extraction result "1" this time to make it meet the total length requirement of 12 bits, that is, the extraction result of this step should be "01". Finally, the secret information obtained is "010100011001", which is consistent with the information input when hiding the information.

[0085] The third aspect of the present invention provides a generative steganographic embedding system based on chess record construction, as shown in the schematic diagram Figure 4 As shown, it includes the following modules:

[0086] A move operation module is used to execute a move according to a received move operation instruction;

[0087] The serialization module is used to record the moves into the encrypted chess records;

[0088] The formula opening chess record module is used to obtain a formula opening chess record library. Each chess record in the chess record library has the same number of steps and is marked with serial numbers in sequence. When embedding starts, the opening chess record with a serial number equal to the length of the binary information to be embedded is obtained from the opening chess record library, a new game is started, and the move operation module is notified to execute and record using the serialization module;

[0089] The intelligent chess player module is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, decide the move operation according to the remaining moves after deletion and the binary information to be embedded, notify the move operation module to execute and use the serialization module to record.

[0090] The fourth aspect of the present invention provides a generative steganalysis extraction system based on chess record construction, as shown in the schematic diagram Figure 5 As shown, it includes the following modules:

[0091] The deserialization module is used to convert the chess records into corresponding move operation instructions;

[0092] A move operation module is used to execute a move according to a received move operation instruction;

[0093] The formula opening chess record module is used to obtain a formula opening chess record library. Each chess record in the chess record library has the same number of steps and is marked with serial numbers in sequence. When the extraction starts, the formula opening part is extracted from the secret chess record, and the move operation module is notified to execute the opening chess record and search for the corresponding opening chess record serial number in the formula opening chess record library to obtain the length of the information. If the received secret chess record contains multiple, the opening chess record serial numbers are sequentially extracted in descending order according to the size of the opening chess record serial numbers;

[0094] The intelligent chess player module is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, extract the remaining moves after deletion and the moves in the secret chess record, notify the move operation module to execute the moves in the secret chess record and write the obtained binary number into the result file.

[0095] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A generative steganalysis embedding method based on chess record construction, characterized in that: The following steps are involved: Obtain a preset formulaic opening chess game database and a preset intelligent chess player, wherein each chess game in the chess game database has the same number of steps and is marked with serial numbers in sequence; Convert the information to be embedded into binary information and calculate the length of the binary information; Start embedding, obtain an opening chess record with a sequence number equal to the length of the binary information from the opening chess record library, start a new game, execute the opening chess record and record it into the encrypted chess record; The intelligent chess player is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, and perform move operations according to the remaining moves after deletion and the binary information to be embedded and record them in the encrypted chess record; Repeat the above game embedding operation until the secret information is steganographically written and the final secret chess record is obtained.

2. According to the generative steganalysis embedding method based on chess record construction according to claim 1, it is characterized in that: The preset intelligent chess player is trained by AlphaZero, and the parameters of the intelligent chess player are fixed during the embedding and extraction process. The method for constructing the preset formulaic opening chess database is as follows: calling the preset intelligent chess player and letting it play chess automatically. Each game only records the first k steps of the opening. During the automatic game process, each move is randomly selected from the moves ranked top n in probability. The obtained chess records are numbered in ascending order from 1 and stored in the opening chess database.

3. According to the generative steganalysis embedding method based on chess record construction according to claim 1, it is characterized in that: It is necessary to ensure that the length of the binary information to be embedded is less than or equal to the number of opening game records in the opening game record library.

4. According to the generative steganalysis embedding method based on chess record construction according to claim 1, it is characterized in that: The specific steps of deleting some moves in the move set are: Find the move operation with the highest probability in the move operation set and record its probability value as maxprob; If maxprob is greater than or equal to the first preset value, only the move is retained and other moves are deleted; If maxprob is less than the first preset value and greater than or equal to the second preset value, then delete all moves that satisfy a probability less than a third preset value, where the first preset value is greater than the second preset value and the third preset value is less than maxprob; Delete all moves that satisfy the probability value of 0.

5. According to the generative steganalysis embedding method based on chess record construction according to claim 1, it is characterized in that: The specific steps of performing the move operation based on the remaining moves after deletion and the binary information to be embedded and recording them in the secret chess record are as follows: If the number of remaining moves m is 1, directly execute the move and record it in the secret chess record; otherwise, sort the remaining moves by probability from large to small, and number them in ascending order from 0; Extract the length of the binary information to be embedded. , convert it into a decimal number, select the corresponding number of moves from the currently available set of moves, execute and record it in the secret chess record, if the length of the remaining binary information to be embedded is less than Then all the remaining contents of the binary information to be embedded are intercepted.

6. According to the generative steganalysis embedding method based on chess record construction according to claim 1, it is characterized in that: If the game ends during the information embedding process, it is necessary to start the next game and embed the remaining binary information.

7. A generative steganalysis extraction method based on chess record construction, characterized in that: The following steps are involved: Obtain a preset formulaic opening chess game database and a preset intelligent chess player, wherein each chess game in the chess game database has the same number of steps and is marked with serial numbers in sequence; Receive a secret chess record, extract a formula opening part from the secret chess record, and search the corresponding opening chess record serial number in the formula opening chess record library to obtain the length of information. If the received secret chess record contains multiple secret records, extract them in descending order according to the size of the opening chess record serial number; Using the intelligent chess player to start a new game and execute the opening chess record; For the remaining moves in the chess record, the intelligent chess player is used to continuously and automatically play chess. In each step of the game process, the intelligent chess player calculates the legal move set in the current situation and determines the probability of each move, deletes some moves in the move set according to the probability of each move, extracts the moves according to the remaining moves after deletion and the moves in the secret chess record, executes the move operation in the secret chess record and writes the obtained binary number into the result file; The game extraction process is repeated until all moves in the secret chess record are extracted.

8. The generative steganalysis extraction method based on chess record construction according to claim 7 is characterized in that: The specific steps of extracting the remaining moves after deletion and the moves in the secret chess records are: If the number of remaining moves m is 1, execute the move operation directly; otherwise, sort the remaining moves according to probability from large to small, and number them in ascending order starting from 0; Take a move from the secret chess record, find the same move in the move set, convert its serial number into a binary number, if the length of the binary number is less than Then add 0 at the beginning until the length is equal to Write the obtained binary number into the result file and execute the move operation of the secret chess record; When extracting the last move of the encrypted chess record, let the binary sequence obtained after extracting this move be bin. If the length of the binary sequence in the result file plus the length of bln is less than the number of the starting chess record, it is necessary to add a number of leading 0s of bin until the length of the binary sequence in the result file plus the length of bln is equal to the number of the starting chess record. Write bin into the result file, and all the original embedded information is extracted.

9. A generative steganographic embedding system based on chess record construction, used to implement a generative steganographic embedding method based on chess record construction as described in any one of claims 1 to 6, characterized in that: Includes the following modules: A move operation module is used to execute a move according to a received move operation instruction; The serialization module is used to record the moves into the encrypted chess records; The formula opening chess record module is used to obtain a formula opening chess record library. Each chess record in the chess record library has the same number of steps and is marked with serial numbers in sequence. When embedding starts, the opening chess record with a serial number equal to the length of the binary information to be embedded is obtained from the opening chess record library, a new game is started, and the move operation module is notified to execute and record using the serialization module; The intelligent chess player module is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, decide the move operation according to the remaining moves after deletion and the binary information to be embedded, notify the move operation module to execute and use the serialization module to record.

10. A generative steganalysis extraction system based on chess record construction, used to implement a generative steganalysis extraction method based on chess record construction as claimed in claim 7 or 8, characterized in that: Includes the following modules: The deserialization module is used to convert the chess records into corresponding move operation instructions; A move operation module is used to execute a move according to a received move operation instruction; The formula opening chess record module is used to obtain a formula opening chess record library. Each chess record in the chess record library has the same number of steps and is marked with serial numbers in sequence. When the extraction starts, the formula opening part is extracted from the secret chess record, and the move operation module is notified to execute the opening chess record and search for the corresponding opening chess record serial number in the formula opening chess record library to obtain the length of the information. If the received secret chess record contains multiple, the opening chess record serial numbers are sequentially extracted in descending order according to the size of the opening chess record serial numbers; The intelligent chess player module is used to calculate the legal move set in the current situation and determine the probability of each move, delete some moves in the move set according to the probability of each move, extract the remaining moves after deletion and the moves in the secret chess record, notify the move operation module to execute the moves in the secret chess record and write the obtained binary number into the result file.

Images