CN111432217B - Information hiding embedding and extracting method based on two-dimensional histogram translation - Google Patents
Information hiding embedding and extracting method based on two-dimensional histogram translation Download PDFInfo
- Publication number
- CN111432217B CN111432217B CN202010225600.3A CN202010225600A CN111432217B CN 111432217 B CN111432217 B CN 111432217B CN 202010225600 A CN202010225600 A CN 202010225600A CN 111432217 B CN111432217 B CN 111432217B
- Authority
- CN
- China
- Prior art keywords
- embedding
- information
- coefficient
- block
- embedded
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000013519 translation Methods 0.000 title claims abstract description 33
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000000605 extraction Methods 0.000 claims description 42
- 238000013139 quantization Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 9
- 239000000203 mixture Substances 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 4
- 239000003550 marker Substances 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000004091 panning Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000002441 reversible effect Effects 0.000 abstract description 2
- 239000000284 extract Substances 0.000 description 10
- 239000013598 vector Substances 0.000 description 9
- 230000000007 visual effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012805 post-processing Methods 0.000 description 2
- 102100029469 WD repeat and HMG-box DNA-binding protein 1 Human genes 0.000 description 1
- 101710097421 WD repeat and HMG-box DNA-binding protein 1 Proteins 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000012634 fragment Substances 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/46—Embedding additional information in the video signal during the compression process
- H04N19/467—Embedding additional information in the video signal during the compression process characterised by the embedded information being invisible, e.g. watermarking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
Abstract
The invention provides an information hiding, embedding and extracting method based on two-dimensional histogram translation. The information hiding and embedding method comprises the following steps: selecting from the embedding carrierTaking a 4 multiplied by 4 coefficient embedding block to be embedded with secret information; randomly selecting two coefficients Y1 and Y2 from the coefficients of the embedded block to form an embedded coefficient pair { Y 1 ,Y 2 H, according to the embedding coefficient pair { Y } 1 ,Y 2 Dividing the coefficient pairs into 19 sets which are not crossed with each other by the value of { Y }, respectively defining a two-dimensional histogram translation algorithm corresponding to each set of the 19 sets, and embedding the coefficient pairs { Y } according to the secret information to be embedded and the current embedding coefficient pair 1 ,Y 2 And embedding the secret information into the 4 x 4 coefficient embedding block by using a histogram translation algorithm corresponding to the set. And extracting the information by adopting the reverse operation of information hiding and embedding. The two-dimensional histogram translation method provided by the invention has the advantages of large embedding capacity, small embedding error, high embedding efficiency, high suitability for high-resolution videos and the like.
Description
Technical Field
The invention belongs to the technical field of information processing, and particularly relates to an information hiding, embedding and extracting method based on two-dimensional histogram translation.
Background
A two-dimensional histogram translation algorithm is an extremely efficient method in the field of video information hiding, and the embedding and extraction of secret information are completed by mainly using widely distributed video information such as QDCT coefficients and motion vectors as carriers. Furthermore, the two-dimensional histogram shift algorithm mainly uses the QDCT coefficient or the peak point of the motion vector to embed data, and completes the data embedding by adding 1 or subtracting 1 to the peak point value, so that the number of coefficients or motion vectors at the peak point largely determines the embedding capacity performance of the two-dimensional histogram shift. Since video has more redundant space, QDCT coefficients or motion vector histogram distribution is sharper, the two-dimensional histogram shift algorithm has better embedding performance compared to image with video as carrier, and has become a hot spot of video information hiding research.
Among the conventional histogram shift algorithms, the histogram shift algorithm based on the QDCT coefficient is a highly effective information hiding method. According to the statistics of the prior art, the QDCT coefficient with a value of "0" occupies 82.87% to 99.39% of all QDCT coefficients, and thus it can be seen that the distribution of QDCT coefficients is very sharp, and thus the QDCT coefficient has excellent embedding performance as histogram shift. In the information hiding algorithm proposed in the prior art, the embedding of the secret information is typically implemented by using one-dimensional histogram translation. The embedding of the secret information is completed by peak points 0 and-1, and the absolute values of other non-peak points are added with 1 to realize the translation to the left and the right. The algorithm is simple and efficient to implement, but due to the fact that one bit of secret information is embedded, one bit of QDCT coefficient needs to be modified, embedding efficiency is not high, and the challenge of poor embedding capacity exists. Two-dimensional histogram translation is the combined use of two one-dimensional histograms, another prior art proposes a novel two-dimensional histogram translation algorithm, which forms a coefficient pair by randomly selecting two high-frequency coefficients in a 4 × 4 small block, and completes the embedding of secret information by correspondingly modifying the current coefficient pair according to the set to which the existing coefficient pair belongs. The algorithm has small embedding error and high embedding efficiency, but the algorithm is only applied to 3D video taking H.264/AVC as a coding mode and cannot be directly applied to common 2D video files occupying mainstream on a network. In addition to this, two-dimensional histogram shift algorithms based on motion vectors are also of great interest. Another prior art proposes that the horizontal and vertical components of a motion vector are used as candidate embedding positions, and the embedding and extraction operations of secret information are completed according to a two-dimensional histogram set to which the two components of the motion vector belong. The method has extremely large embedding capacity, but the embedding error caused by embedding secret information, namely visual distortion, is also large, so the method is not suitable for application scenes with high visual quality requirements such as high-definition videos and medical images.
Disclosure of Invention
Aiming at least one defect or improvement requirement in the prior art, the invention provides an information hiding, embedding and extracting method based on two-dimensional histogram translation, which has the advantages of large embedding capacity, small embedding error, higher embedding efficiency, particular suitability for high-resolution videos and the like.
To achieve the above object, according to a first aspect of the present invention, there is provided an information hiding and embedding method based on two-dimensional histogram shifting, including the steps of:
selecting an embedded block for information embedding from an embedded carrier;
randomly selecting two coefficients Y1 and Y2 from the coefficients of the embedded block to form embedded blockCoefficient pair { Y 1 ,Y 2 The embedding coefficient pair { Y } 1 ,Y 2 The secret information to be embedded is received, and the embedding coefficient pair { Y is adopted 1 ,Y 2 Embedding the secret information into the embedded block by a histogram translation algorithm corresponding to the set to which the secret information belongs;
wherein the 19 non-intersecting sets include set A 1 -the secret information is denoted m { (0,0) } i m i+1 m i+2 m i+3 Defining set A in advance 1 The histogram shift algorithm of (1) is: if the current coefficient pair { Y 1 ,Y 2 }∈A 1 And updating the coefficient pair embedded with the secret information to be { Y 1 ′,Y 2 ′}，
Preferably, the 19 non-intersecting sets further include sets: a. the 2 ＝{(Y 1 ,0)|Y 1 >0}
A 3 ＝{(Y 1 ,0)|Y 1 <0}
A 4 ＝{(0,1)}
A 5 ＝{(0,-1)}
A 6 ＝{(0,Y 2 )|Y 2 >1}
A 7 ＝{(0,Y 2 )|Y 2 <-1}
A 8 ＝{(-1,Y 2 )|Y 2 >1}
A 9 ＝{(1,Y 2 )|Y 2 >1}
A 10 ＝{(1,Y 2 )|Y 2 <-1}
A 11 ＝{(-1,Y 2 )|Y 2 <-1}
A 12 ＝{(Y 1 ,1)|Y 1 <0}
A 13 ＝{(Y 1 ,Y 2 )|Y 1 <-1,Y 2 >1}
A 14 ＝{(Y 1 ,1)|Y 1 >0}
A 15 ＝{(Y 1 ,Y 2 )|Y 1 >1,Y 2 >1}
A 16 ＝{(Y 1 ,-1)|Y 1 <0}
A 17 ＝{(Y 1 ,Y 2 )|Y 1 <-1,Y 2 <-1}
A 18 ＝{(Y 1 ,-1)|Y 1 >0}
A 19 ＝{(Y 1 ,Y 2 )|Y 1 >1,Y 2 <-1}，
The histogram translation algorithm corresponding to each set in the 19 non-intersecting sets is predefined as follows: if the embedding coefficient pair { Y 1 ,Y 2 }∈A 1 ,A 1 ,A 3 ,A 4 ,A 5 ,A 8 ,A 9 ,A 10 ,A 11 Embedding the secret information into the embedding coefficient pair, if the embedding coefficient pair { Y } 1 ,Y 2 }∈A 6 ,A 7 ,A 12 ,A 13 ,A 14 ,A 15 ,A 16 ,A 17 ,A 18 ,A 19 Then histogram shifting is performed on the pair of embedded coefficients.
Preferably, the embedded block is a 4 × 4 transform quantized residual coefficient block, the pair of embedded coefficients { Y } 1 ,Y 2 And the coefficients are high-frequency coefficients of residual coefficient blocks after intra-frame and inter-frame 4 multiplied by 4 transform quantization.
Preferably, the pair of embedding coefficients { Y } 1 ,Y 2 Is the QDST coefficient.
Preferably, the selecting of the embedded block for information embedding specifically includes:
predefined Random function seed Random (seed1) and block threshold T block Selecting a product satisfying Random (seed) 1 )≤T block The 4 × 4QDST small blocks as embedded blocks.
Preferably, before embedding the secret information into the embedding block, preprocessing the secret information, the preprocessing including the steps of:
encrypting the secret information into a ciphertext by using a public key in an RSA encryption mode;
carrying out binarization processing on the ciphertext to obtain a binarization bit sequence;
and adding a start marker and an end identifier to the binary bit sequence to obtain the bit sequence.
According to a second aspect of the present invention, there is provided an information extraction method based on two-dimensional histogram translation, comprising the steps of:
receiving information embedding parameters, and selecting an extraction block for information extraction from an embedded carrier by using the embedding parameters, wherein the embedded carrier is embedded with secret information;
randomly selecting two coefficients Y from the coefficients of the extraction block using the embedding parameters 1 ′，Y 2 ' composition includes pairs of secret coefficients { Y 1 ′,Y 2 ′}；
Extracting secret information by performing an inverse operation of secret information embedding on the pair of secret-containing coefficients by using the embedding parameter;
wherein, if Y 1 ′∈[-1,1]And Y is 2 ′∈[-1,1]Then the extracted secret information m is:
according to a third aspect of the invention, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs any of the methods described above.
In general, compared with the prior art, the invention has the following beneficial effects:
(1) the invention makes full use of the coefficient peak point '0' to embed more secret information into the coefficient pair position, namely, modifying one-bit coefficient to embed 4-bit secret information at most. Through the novel two-dimensional histogram translation method, the information hiding and embedding capacity is greatly improved, and compared with the existing related information hiding technology, the embedding efficiency is higher.
(2) The invention uses the high-frequency coefficient of the 4 x 4 small block of the H.265/HEVC inter-frame prediction block as an embedding position. In common compressed video files, inter-frame prediction is more common than intra-frame prediction; secondly, the positions of the 4 × 4 small blocks are all positions where the motion of the video content is severe, and the high-frequency coefficients after transform quantization contain less video information than the rest of the low-frequency coefficients, so that the positions are selected as candidate positions for embedding information, and the embedding performance of video information hiding can be further improved, namely, compared with the embedding capacity and the video distortion.
(3) The invention fully utilizes the QDST coefficient after the transformation and quantization, and the QDST coefficient occupies most of the video code stream file, thereby the safety is higher compared with other methods.
Drawings
FIG. 1 is a schematic diagram of an information embedding process according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of information-embedded secret information preprocessing according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of an information-embedding histogram shift algorithm according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of an information extraction flow according to an embodiment of the present invention;
FIG. 5 is a high frequency coefficient diagram of information extraction according to an embodiment of the present invention;
fig. 6 is a schematic diagram of secret information restoration proposed by the information of the embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples, so that the objects, aspects and advantages of the present invention will become more apparent. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The invention provides an information hiding, embedding and extracting method based on two-dimensional histogram translation, which is respectively executed at a secret information embedding end and a secret information extracting end.
When information embedding is carried out, a two-dimensional histogram translation algorithm is defined in advance, wherein the coefficient peak value point '0' is fully utilized, so that more secret information can be embedded into the coefficient pair position. Preferably, the embedded block is a 4 × 4 small block, i.e., a 4 × 4 block of transform quantized residual coefficients. Embedding coefficient pair of embedded block { Y 1 ,Y 2 And the coefficients are high-frequency coefficients of residual coefficient blocks after intra-frame and inter-frame 4 multiplied by 4 transform quantization. Preferably, the pair of embedding coefficients of the embedded block { Y } 1 ,Y 2 Is the QDST coefficient.
When information is extracted, a two-dimensional histogram translation algorithm similar to that during embedding is adopted. Preferably, the extracted block is a 4 × 4 small block, i.e., a 4 × 4 block of transform quantized residual coefficients, containing the pair of secret coefficients { Y } 1 ,Y 2 And the coefficients are high-frequency coefficients of residual coefficient blocks after intra-frame and inter-frame 4 multiplied by 4 transform quantization. Preferably, the pair of secret coefficients { Y } of the block is extracted 1 ,Y 2 Is the QDST coefficient.
The information embedding method based on H.265 histogram translation of the embodiment of the invention comprises four parts of preprocessing of secret information, selection of an embedded block position, embedding of the secret information, two-dimensional histogram translation and entropy coding into a code stream file, as shown in figure 1.
And S11, preprocessing the secret information to be embedded.
The preprocessing part of the secret information comprises the steps of encryption, binarization, adding start and end bits of the secret information, and the like, as shown in fig. 2. The main purpose of the preprocessing is to encrypt the secret information to strengthen the security of the secret information, namely to ensure that the secret information is not read and known maliciously when the secret information is attacked maliciously; secondly, binarization is carried out so that secret information can be embedded into a subsequent carrier video, namely a preceding condition for embedding the secret information subsequently; finally, the start bit and the end bit of the secret information are added to ensure that the secret information can be accurately extracted at the extraction end, and irrelevant information is removed, namely, when information hiding begins and when information hiding ends in the carrier video are accurately known.
Specifically, the pretreatment comprises the steps of:
and S111, encrypting. And the RSA encryption mode is adopted, and the secret information to be embedded is encrypted into a ciphertext by using the public key so as to ensure the safety of the secret information. And the receiving end correspondingly decrypts the obtained ciphertext by using the private key.
And S112, binarization. And sequentially binarizing the ciphertexts encrypted in the step S111 to obtain a binary bit sequence. The purpose of binarization is the necessary pre-processing for embedding information in the embedding module.
And S113, adding a start bit and an end bit. In order to accurately locate the start and end positions of the secret information, the binary bit sequence obtained in step S112 needs to be added with the start identifier 0x000000 and the end identifier 0x000001 in the bit string sequence.
S12, an embedded block for embedding information is selected from the embedded carrier.
The selection of the position of the embedded block has two main purposes: firstly, in order to strengthen the safety of information hiding, namely improve the anti-steganography analysis capability of the information hiding, namely randomly selecting an embeddable block for embedding secret information; and secondly, in order to improve the embedding performance of information hiding, namely, the high visual quality of the carrier video is maintained from the perspective of visual effect by selecting an area with complex video background texture characteristics and violent motion for embedding information while maximizing the embedding capacity.
Specifically, the method comprises the following steps:
and S121, predefining a random seed. A Random function Random (seed1) is used to decide whether the current 4 × 4 block is a candidate embedded block in step S122. The random function seed is required to be consistent with the extraction end so as to correctly screen the embedded block and extract the secret information.
S122, selecting candidate embedded blocks. In particular, the selection of candidate blocks involves two aspects.
First, a block threshold value T is predefined block According to the current Random function Random (seed) 1 ) Whether e (0,1) is less than threshold T block E (0,1) to determine whether to treat the current 4 x 4QDST tile as an embedded block. If Random (seed) 1 )≤T block If not, the current 4 × 4QDST block is not taken as a candidate embedded block, and the next 4 × 4QDST block is determined.
The second is the selection of candidate embedded block locations. The technical scheme is suitable for I, B, P frames, and particularly depends on the application scene of information hiding. For a video sequence with I frames, P frames or B frames, if the last frame of a GOP (group of picture) sequence is selected as an embedding position, the inter-frame distortion drift caused by embedded information can be effectively prevented, and the video sequence has a good effect on scenes with high requirements on video application visual quality, such as medical images and the like. Meanwhile, if the embedded information is in the I frame, distortion drift caused by embedding does not affect other I frames, but for all subsequent P frames or B frames which need to refer to the current I frame, inter-frame distortion drift still exists, but the position has larger embedding capacity, so that a video frame in which the embedded block is located is specifically dependent on which frame of the GOP, and can be flexibly adjusted according to actual needs, that is, the scheme has certain expandability.
S13, the embedded secret information is shifted with the two-dimensional histogram. That is, randomly selecting two coefficients Y1 and Y2 from the coefficients of the embedded block to form an embedded coefficient pair { Y 1 ,Y 2 R, embedding coefficient pair Y 1 ,Y 2 And the histogram translation algorithm corresponding to each of the 19 non-crossed sets is predefined.
Randomly selecting two coefficient composition coefficient pairs { Y ] from the selected 4 x 4QDST coefficients 1 ,Y 2 }. Embedding random seed in this step 2 The information extraction device needs to be consistent with the extraction terminal so as to correctly extract the information. Selecting coefficient pair { Y 1 ,Y 2 The steps of (c) include:
s131, first, defining a coefficient threshold T of a 4 × 4 embedded block embed Judging the DC coefficient Y in the current 4X 4QDST coefficient block 0,0 Whether or not it is greater than coefficient threshold T embed . If Y is 0,0 ≥T embed Then two coefficient composition coefficient pairs { Y } are selected from the high frequency coefficients in the current 4 x 4 tile according to step S132 1 ,Y 2 }; if Y is 0,0 < T embed If so, the next coefficient block is directly traversed without performing subsequent operations such as embedding the current 4 × 4QDST coefficient block, and the process proceeds to step S12.
S132, establishing a simple random function f (seed) 2 ) E.g. (0,1), sequentially traversing the high-frequency coefficients, and obtaining a random function value f (seed) 2 ) More than or equal to 0.5, selecting corresponding coefficient to put in coefficient pair { Y ≧ 0.5 1 ,Y 2 }; if the random function value f (seed) 2 )<0.5, the coefficient of the current position is not taken as the embedding position.
S133, dividing the 4 × 4QDST coefficient pairs into 19 sets that do not intersect with each other, and defining a two-dimensional histogram shift algorithm corresponding to each set in advance.
According to the current coefficient pair { Y 1 ,Y 2 And (4) embedding the secret information in the set to which the secret information belongs according to a histogram translation algorithm of the corresponding set. Wherein the coefficient pair { Y 1 ,Y 2 Different sets can lead to different embedded bits of the secret information, that is, the technical scheme embeds more secret information by using peak points as much as possible, and maximally improves the embedding capacity on the premise of ensuring the visual quality of the carrier video.
Preferably, the two-dimensional histogram shift algorithm is as shown in fig. 3:
current coefficient pair Y 1 ,Y 2 The set to which it belongs is denoted A, then 19 non-intersecting sets are denoted as follows, where the subset A is 1 ,A 2 ,A 3 ,A 4 ,A 5 ,A 8 ,A 9 ,A 10 ,A 11 Actually for information hiding and others for translation of coefficients.
A 1 ＝{(0,0)}
A 2 ＝{(Y 1 ,0)|Y 1 >0}
A 3 ＝{(Y 1 ,0)|Y 1 <0}
A 4 ＝{(0,1)}
A 5 ＝{(0,-1)}
A 6 ＝{(0,Y 2 )|Y 2 >1}
A 7 ＝{(0,Y 2 )|Y 2 <-1}
A 8 ＝{(-1,Y 2 )|Y 2 >1}
A 9 ＝{(1,Y 2 )|Y 2 >1}
A 10 ＝{(1,Y 2 )|Y 2 <-1}
A 11 ＝{(-1,Y 2 )|Y 2 <-1}
A 12 ＝{(Y 1 ,1)|Y 1 <0}
A 13 ＝{(Y 1 ,Y 2 )|Y 1 <-1,Y 2 >1}
A 14 ＝{(Y 1 ,1)|Y 1 >0}
A 15 ＝{(Y 1 ,Y 2 )|Y 1 >1,Y 2 >1}
A 16 ＝{(Y 1 ,-1)|Y 1 <0}
A 17 ＝{(Y 1 ,Y 2 )|Y 1 <-1,Y 2 <-1}
A 18 ＝{(Y 1 ,-1)|Y 1 >0}
A 19 ＝{(Y 1 ,Y 2 )|Y 1 >1,Y 2 <-1}
The embedding rule depends on the current coefficient pair Y 1 ,Y 2 Different sets are adopted, different embedding modes are adopted to improve the embedding capacity as much as possible, and the current coefficient pair after information is embedded is assumed to be { Y } 1 ′,Y 2 ' }, the secret information bit string to be embedded is m i m i+1 m i+2 m i+3 Then the specific embedding rule is as follows:
if the current coefficient pair Y 1 ,Y 2 }∈A 1 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 2 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 3 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 4 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 5 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 8 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 9 After embedding the informationThe coefficients of (a) are:
if the current coefficient pair Y 1 ,Y 2 }∈A 10 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 11 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 6 ,A 7 ,A 12 ,A 13 ,A 14 ,A 15 ,A 16 ,A 17 ,A 18 ,A 19 Then, the current coefficient is not embedded with secret information, only the histogram is shifted, and the shifted coefficient pair { Y } 1 ′,Y 2 ' } is:
the embedding process is illustrated as one example. If the current coefficient pair Y 1 ,Y 2 With {0,0}, then the current coefficient pair {0,0} can be extended to {0,0}, {1,0}, {0,1}, { -1,0}, {0, -1}, {1,1}, { -1,1}, { -1, -1}, {1, -1}, and {1, -1}, with the corresponding embedded secret information being 000, 001,010,011,100,101,110,1110,1111, respectively. From the above example, it can be seen that, in the optimal case, the scheme proposed herein can achieve the purpose of embedding 3 bits of secret information at most and modifying one bit coefficient only, i.e. adding 1 or subtracting 1, to achieve the purpose of information hiding.
After the current 4 × 4QDST coefficient completes embedding the secret information, the discrimination and embedding of the next 4 × 4QDST coefficient are continued. I.e., proceeds to step S12 until the secret information is embedded in the stop bit.
And S14, entropy coding into a code stream file.
The main function of this section is to embed the modified QDST coefficients { Y } after the secret information is embedded at step S13 1 ′,Y 2 ' } and other syntactic elements such as prediction modes, motion vectors, partition modes and the like are subjected to entropy coding one by one to form a code stream file, so as to form a compressed video format which can be finally transmitted by a network. The processing mode of the part is not related to a specific information hiding algorithm, but is a necessary basic condition for video files at a network transmission and information hiding extraction end.
Another embodiment of the present invention provides an information extraction method based on h.265 histogram panning, as shown in fig. 4, the extraction end mainly includes four main processes of entropy decoding, extracting information according to two-dimensional histogram panning rules, decoding reconstructed video files, and post-processing secret information, and corresponds to the embedded end.
S21, carrying out entropy decoding on the embedded carrier and selecting an extraction block.
The main function of the part is to obtain the QDST coefficient { Y after network transmission 1 ′,Y 2 '}. This part contains two steps, entropy decoding and selection of the extraction block location.
S211 entropy decoding. The extraction end firstly obtains carrier video containing secret information, namely video file with suffix of bin from the network as input of the extraction end, and QDST coefficient { Y is decoded by the entropy decoding module 1 ′,Y 2 ' decoding video coding parameters and residual data such as a segmentation mode, a prediction mode, a motion vector, a reference frame index and the like, and taking the decoded video coding parameters and residual data as a necessary basis for extracting subsequent secret information.
S212 extracts the selection of the block position. Receiving information embedding parameters including seed of random function used in embedding 1 Random seed 2 And a block threshold T block And selecting an extraction block for information extraction from the embedded carrier by using the information embedding parameters. Seed of random function 1 Need to be embedded withThe input ends are kept consistent, and the block threshold T of the block is extracted at the same time block Should also be consistent with the embedded end. Whether the entropy-decoded current 4 × 4QDST block is an extraction block depends on a Random function Random (seed) 1 ) Whether or not it is less than or equal to block threshold T block If Random (seed) 1 ) Less than or equal to T block If the current 4 × 4QDST block is used as the extraction block of the secret information, entering the next part to extract the information according to the two-dimensional histogram translation rule; if Random (seed) 1 ) Greater than T block If the current 4 × 4QDST block is not used as the extraction block, the next 4 × 4QDST block is traversed according to the entropy decoding flow.
S22, randomly selecting two coefficients Y from the coefficients of the selected extraction block 1 ′，Y 2 ' Pair of component extraction block coefficients { Y 1 ′,Y 2 ' }, for the extraction block coefficient pair { Y 1 ′,Y 2 ' } perform the inverse operation of the secret information embedding to extract the secret information.
The main function of this part is to derive the QDST coefficients Y from the secret information contained 1 ′,Y 2 ' } secret information is extracted. Mainly comprising two steps, the QDST coefficient pair { Y 1 ′,Y 2 ' } and extracting secret information according to an extraction rule.
S221, secret QDST coefficient pair { Y } 1 ′,Y 2 ' } construction of the DNA fragment.
Step1, the extraction end firstly defines a coefficient threshold T of an extraction block Extract Here coefficient threshold T Extract Needs to be connected with an embedded end T embed And the consistency is maintained. Discriminating DC coefficient Y 0,0 And coefficient threshold T Extract To determine whether the current 4 x 4 block has secret information embedded at the embedding end. If Y is 0,0 ≥T Extract If yes, the current 4 × 4QDST block is taken as an extraction block and is transferred to Step 2; if Y is 0,0 <T Extract If the current 4 × 4QDST block is not used as the extraction block of the secret information, the discrimination of the next 4 × 4QDST block is continued. Here by determining the low frequency coefficient Y 0,0 And coefficient threshold T Extract Is used for screening the embedded and extracted blocks, mainly selecting a relatively complex area of the video background as informationHidden carrier region due to low frequency coefficient Y 0,0 Is the average of the video residual energy of the current 4 x 4 block, larger Y 0,0 Meaning that the video content at that location changes more and the background is more complex. The operation of hiding information in the area has more visual secrecy.
Step2. for the current 4 x 4QDST coefficient block, a random seed is first defined 2 Seed here 2 Should also be consistent with the embedded end to extract the secret information normally. Sequentially traversing the high-frequency coefficients in the current 4 x 4QDST coefficient block, and using a random function f (seed) for each high-frequency coefficient 2 ) E (0,1), if the random function value f (seed) 2 ) Less than or equal to 0.5, the current stage coefficient is taken as the coefficient pair { Y 1 ′,Y 2 ' } up to coefficient pair Y 1 ′,Y 2 ' } finish forming; if the random function value f (seed) 2 )>0.5, then the current coefficient is not taken as the coefficient pair { Y 1 ′,Y 2 ' } composition. The reason why the high-frequency coefficient is selected as the information hiding carrier is that the high-frequency coefficient has smaller video energy, and the change of the high-frequency coefficient has little influence on the video, so that the information hiding effect is better than the visual effect of changing the low-frequency coefficient.
The high-frequency coefficient is a coefficient at the lower right position of the current 4 × 4QDST small block, as shown in fig. 5, the green labeled coefficient is the high-frequency coefficient, and the yellow labeled coefficient at the upper left corner is the low-frequency coefficient.
S221, extracts the secret information.
Extracting the secret information according to the result of the embedded end after the two-dimensional histogram translation (Y) 1 ′,Y 2 ' } perform the inverse operation to get the embedded information. Assuming that the extracted secret information is m, the specific extraction manner is as follows:
if Y is 1 ′∈[-1,1]And Y is 2 ′∈[-1,1]Then the extracted secret information is determined by:
if Y is 1 ' epsilon [2, + ∞) and Y 2 ′∈[-1,1]Or Y 1 ′∈(-∞,-2]And Y is 2 ′∈[-1,1]Then the extracted secret information is determined by:
if Y is 1 ′∈[-1,1]And | Y 2 ' | 2, then the extracted secret information is determined by:
if | Y 1 ′|∈[1,2]And | Y 2 ′|>2, the extracted secret information is determined by:
if { Y 1 ′,Y 2 ' } does not belong to any of the above cases, the extraction of secret information is not performed, because the embedding of secret information is not performed at the embedding end, only the shifting operation of the histogram is performed, and additionally, the shifting can be used for the application of reversible information hiding and the like.
And S23, decoding the reconstructed video file.
The main function of this part is to derive the QDST coefficients Y from the secret information contained 1 ′,Y 2 ' extracting the secret information, and subjecting the syntax element and the QDST coefficient decoded according to the first part of entropy to inverse transformation inverse quantization, intra-frame inter-frame prediction and other processing mechanisms to obtain a decoded yuv format video file which can be used for normal video playing and other functions.
S24, secret information post-processing.
The main function of this part is to extract the secret information bit string obtained after the operation from step S22, remove the start bit and the end bit of the secret information bit string, binarize, decrypt, and so on, to obtain the original secret information, as shown in fig. 6, including the steps of:
s241, remove the start bit and the end bit. The bit string sequence obtained by the second part extraction operation needs to be added with a start marker 0x000000 and an end marker 0x000001 in the bit string sequence, so as to obtain secret information with accurate positioning.
And S242, inverse binarization. And (4) performing inverse binarization on the secret information bit string obtained by removing the start bit and the end bit in the step (S241) to obtain an encrypted secret information sequence.
And S243, decrypting. And (4) decrypting the encrypted secret information sequence obtained in the step (S242) by using the private key of the extraction terminal, so as to obtain the original ciphertext.
The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the technical solution of any one of the above embedding or extracting method embodiments. The implementation principle and technical effect are similar to those of the above method, and are not described herein again.
It must be noted that in any of the above embodiments, the methods are not necessarily executed in order of sequence number, and as long as it cannot be assumed from the execution logic that they are necessarily executed in a certain order, it means that they can be executed in any other possible order.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (5)
1. An information hiding and embedding method based on H.265 two-dimensional histogram translation is characterized by comprising the following steps of:
selecting an embedded block for information embedding from an embedded carrier;
randomly selecting two coefficients Y1 and Y2 from the coefficients of the embedded block to form an embedded coefficient pair { Y 1 ,Y 2 The embedding coefficient pair { Y } 1 ,Y 2 Belong to 19 non-intersecting pairsOne set in the sets, predefining a histogram translation algorithm corresponding to each set in the 19 sets which are not intersected with each other, receiving secret information to be embedded, and adopting the embedding coefficient pair { Y 1 ,Y 2 Embedding the secret information into the embedded block by a histogram translation algorithm corresponding to the set to which the secret information belongs;
wherein the 19 non-intersecting sets include set A 1 -the secret information is denoted m { (0,0) } i m i+ 1 m i+2 m i+3 Defining set A in advance 1 The histogram shift algorithm of (1) is: if the current coefficient pair { Y 1 ,Y 2 }∈A 1 Then the coefficient pair embedded with the secret information is updated to { Y } 1 ′, 2 ′}，
Wherein the embedded block is a residual coefficient block after 4 × 4 transform quantization, and the embedded coefficient pair is a QDST high-frequency coefficient of the residual coefficient block after 4 × 4 transform quantization within h.265 frames and between frames;
the 19 non-intersecting sets further include sets:
A 2 ＝{(Y 1 ,0)|Y 1 >0}
A 3 ＝{(Y 1 ,0)|Y 1 <0}
A 4 ＝{(0,1)}
A 5 ＝{(0,-1)}
A 6 ＝{(0,Y 2 )|Y 2 >1}
A 7 ＝{(0,Y 2 )|Y 2 <-1}
A 8 ＝{(-1,Y 2 )|Y 2 >1}
A 9 ＝{(1,Y 2 )|Y 2 >1}
A 10 ＝{(1,Y 2 )|Y 2 <-1}
A 11 ＝{(-1,Y 2 )|Y 2 <-1}
A 12 ＝{(Y 1 ,1)|Y 1 <0}
A 13 ＝{(Y 1 ,Y 2 )|Y 1 <-1,Y 2 >1}
A 14 ＝{(Y 1 ,1)|Y 1 >0}
A 15 ＝{(Y 1 ,Y 2 )|Y 1 >1,Y 2 >1}
A 16 ＝{(Y 1 ,-1)|Y 1 <0}
A 17 ＝{(Y 1 ,Y 2 )|Y 1 <-1,Y 2 <-1}
A 18 ＝{(Y 1 ,-1)|Y 1 >0}
A 19 ＝{(Y 1 ,Y 2 )|Y 1 >1,Y 2 <-1}，
the histogram translation algorithm corresponding to each set in the 19 non-intersecting sets is predefined as follows: if the embedding coefficient pair { Y 1 ,Y 2 }∈A 1 ,A 2 ,A 3 ,A 4 ,A 5 ,A 8 ,A 9 ,A 10 ,A 11 Embedding the secret information into the embedding coefficient pair, if the embedding coefficient pair { Y } 1 ,Y 2 }∈A 6 ,A 7 ,A 12 ,A 13 ,A 14 ,A 15 ,A 16 ,A 17 ,A 18 ,A 19 Then performing histogram translation on the embedding coefficient pair;
if the current coefficient pair Y 1 ,Y 2 }∈A 2 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 3 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 4 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 5 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 8 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 9 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 10 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 11 The coefficients after embedding the information are:
if the current coefficient pair Y 1 ,Y 2 }∈A 6 ,A 7 ,A 12 ,A 13 ,A 14 ,A 15 ,A 16 ,A 17 ,A 18 ,A 19 Then the current coefficient is not embedded with secret information, only the histogram is shifted, and the shifted coefficient pair { Y } 1 ′, 2 ' } is:
2. the information hiding and embedding method based on h.265 two-dimensional histogram panning as claimed in claim 1, wherein said selecting an embedded block for information embedding specifically is:
predefined Random function seed Random (seed1) and block threshold T block Selecting a product satisfying Random (seed) 1 )≤T block The 4 × 4QDST small blocks as embedded blocks.
3. The information hiding and embedding method based on h.265 two-dimensional histogram shift as claimed in claim 1, wherein the secret information is preprocessed before being embedded into the embedding block, the preprocessing includes steps of:
encrypting the secret information into a ciphertext by using a public key in an RSA encryption mode;
carrying out binarization processing on the ciphertext to obtain a binarization bit sequence;
and adding a start marker and an end identifier to the binary bit sequence to obtain the bit sequence.
4. An information extraction method based on H.265 two-dimensional histogram translation is characterized by comprising the following steps:
receiving information embedding parameters, and selecting an extraction block for information extraction from an embedded carrier by using the embedding parameters, wherein the embedded carrier is embedded with secret information;
randomly selecting two coefficients Y from the coefficients of the extraction block using the embedding parameters 1 ′，Y 2 ' composition includes pairs of secret coefficients { Y 1 ′, 2 ′}；
Extracting secret information by performing an inverse operation of secret information embedding on the pair of secret-containing coefficients by using the embedding parameter;
wherein, if Y 1 ′∈[-1,1]And Y is 2 ′∈[-1,1]Then the extracted secret information m is:
the extraction block is a residual coefficient block after 4 x 4 transformation quantization, and the secret coefficient pair { Y 1 ′, 2 ' } is the QDST high frequency coefficient of the residual coefficient block after 4 × 4 transform quantization between or within H.265 frames;
if Y is 1 ' epsilon [2, + ∞) and Y 2 ′∈[-1,1]Or Y 1 ′∈(-∞,-2]And Y is 2 ′∈[-1,1]Then the extracted secret information is determined by:
if Y is 1 ′∈[-1,1]And | Y 2 ' | 2, then the extracted secret information is determined by:
if | Y 1 ′|∈[1,2]And | Y 2 ′|>2, the extracted secret information is determined by:
if { Y 1 ′,Y 2 ' } does not belong to any of the above cases, and extraction of secret information is not performed.
5. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010225600.3A CN111432217B (en) | 2020-03-26 | 2020-03-26 | Information hiding embedding and extracting method based on two-dimensional histogram translation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010225600.3A CN111432217B (en) | 2020-03-26 | 2020-03-26 | Information hiding embedding and extracting method based on two-dimensional histogram translation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111432217A CN111432217A (en) | 2020-07-17 |
| CN111432217B true CN111432217B (en) | 2022-08-05 |
Family
ID=71548911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010225600.3A Expired - Fee Related CN111432217B (en) | 2020-03-26 | 2020-03-26 | Information hiding embedding and extracting method based on two-dimensional histogram translation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111432217B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2002307765A1 (en) * | 2001-04-11 | 2002-10-28 | Nice Systems Ltd. | Digital video protection for authenticity verification |
| CN107333139B (en) * | 2017-06-27 | 2019-12-13 | 郑州师范学院 | H.265 semi-reversible video steganography information method based on intra-frame prediction mode |
| CN108366259B (en) * | 2018-01-15 | 2020-04-07 | 华南理工大学 | High-fidelity H.264/AVC video three-coefficient reversible steganography method |
-
2020
- 2020-03-26 CN CN202010225600.3A patent/CN111432217B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN111432217A (en) | 2020-07-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Long et al. | Separable reversible data hiding and encryption for HEVC video | |
| Qin et al. | Reversible data hiding in encrypted image with separable capability and high embedding capacity | |
| Celik et al. | Lossless generalized-LSB data embedding | |
| Chen et al. | A virtual image cryptosystem based upon vector quantization | |
| Chang et al. | A reversible compression code hiding using SOC and SMVQ indices | |
| CN111917558B (en) | Video frame data double-authentication and hierarchical encryption method based on block chain | |
| CN107333139B (en) | H.265 semi-reversible video steganography information method based on intra-frame prediction mode | |
| Li et al. | A HEVC Video Steganalysis Algorithm Based on PU Partition Modes. | |
| Kumar et al. | Reversible data hiding: A contemporary survey of state-of-the-art, opportunities and challenges | |
| Choudry et al. | A survey paper on video steganography | |
| El‐Shafai et al. | Efficient hybrid watermarking schemes for robust and secure 3D‐MVC communication | |
| Taneja et al. | Chaos based partial encryption of spiht compressed images | |
| Farfoura et al. | Low complexity semi-fragile watermarking scheme for H. 264/AVC authentication | |
| CN111432217B (en) | Information hiding embedding and extracting method based on two-dimensional histogram translation | |
| Vivek et al. | Video steganography using chaos encryption algorithm with high efficiency video coding for data hiding | |
| CN114598887A (en) | Anti-recompression video watermarking method for controlling bit rate increase | |
| Sakib et al. | A robust DWT-based compressed domain video watermarking technique | |
| Vinay et al. | A novel secure data hiding technique into video sequences using RVIHS | |
| CN108200435B (en) | Reversible information hiding method for fractal image | |
| Kumar et al. | Image Hiding in a Video-based on DWT & LSB Algorithm | |
| Sivadasan | A survey paper on various reversible data hiding techniques in encrypted images | |
| Kaur et al. | Segmentation and block based image steganography using optimal pixel adjustment process and identical approach | |
| Kunte | Framework for reversible data hiding using cost-effective encoding system for video steganography. | |
| CN116320471B (en) | Video information hiding method, system, equipment and video information extracting method | |
| Zhao et al. | Reversible Data Hiding Based on Two-Dimensional Histogram Shifting |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220805 |