CN110248193B - Reversible information hiding method based on improved difference expansion - Google Patents
Reversible information hiding method based on improved difference expansion Download PDFInfo
- Publication number
- CN110248193B CN110248193B CN201910634241.4A CN201910634241A CN110248193B CN 110248193 B CN110248193 B CN 110248193B CN 201910634241 A CN201910634241 A CN 201910634241A CN 110248193 B CN110248193 B CN 110248193B
- Authority
- CN
- China
- Prior art keywords
- coefficient
- information
- embedding
- embedded
- block
- 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.)
- Active
Links
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
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/83—Generation or processing of protective or descriptive data associated with content; Content structuring
- H04N21/835—Generation of protective data, e.g. certificates
- H04N21/8358—Generation of protective data, e.g. certificates involving watermark
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The invention discloses a reversible information hiding method based on improved difference expansion, which is characterized in that embedding is carried out on brightness residual QDCT coefficients of an I frame, every two adjacent QDCT coefficients in each 4 multiplied by 4 block form a pair and extract coefficient difference values, and the coefficient difference value pair of information to be embedded is modified, so that the difference values are expanded and the aim of embedding secret information is achieved. Due to the application of an intra-frame prediction mode in H.264video coding, modification of residual coefficients causes intra-frame distortion drift. The amplitude of the double expansion of the difference value during embedding can be controlled through the improved reversible algorithm, and the modified amplitude of the coefficient is reduced; and by utilizing the characteristic that the DCT coefficient in video coding is smaller in medium-high frequency coefficient value after quantization, the embedded distortion is reduced, and higher video quality is obtained.
Description
Technical Field
The invention relates to the field of video information hiding, in particular to a reversible information hiding method based on improved difference expansion.
Background
The information hiding technology is an important technology for solving the security problems of multimedia copyright protection, authentication, tracing and the like. The conventional information hiding technology emphasizes that secret information is embedded into a carrier to achieve the purpose of hidden communication, and the embedding process is often irreversible, so that permanent distortion is brought to carrier data. Although the conventional information hiding technology requires that the usability of the carrier file be preserved, even minor distortions are not allowed as the technology is put into practical use, especially in application scenarios where data content authentication is extremely demanding, such as military, medical and judicial. In the background of such demands, the concept of reversible information hiding is proposed and gradually developed and paid attention to.
The development of the reversible information hiding technology of the video is slower than that of the image due to the characteristics of the coding characteristic and the complex data structure of the video, but the development of the video compression technology is focused more and more. In the document "A robust reversible data hiding scheme for h.264with out motion drift" (Liu Y, ju L, hu M, ma X, zhao H, neuroompuing, vol.151, pp.1053-1062,2015), a one-dimensional histogram translation method based on h.264 is proposed, which is different from a common one-dimensional histogram translation method, in which a selected value of the same absolute value is a marker post, the translation is performed along two directions in which the absolute value increases, 0 and 1 bit data are respectively embedded at both sides, and the information to be embedded is preprocessed by BCH encoding before the information is embedded, and the processed secret information is embedded in a carrier video, thereby providing a robust reversible video information hiding scheme. The document "A novel two-dimensional histogram modification for reversible data embedding into stereo H.264video" (Zhao Juan, li Zhi-tan, feng Bing, multimedia Tools and Applications, vol.75, no.10, pp.5959-5980,2016) is based on a histogram translation technique, and proposes a reversible information hiding method suitable for the H.264video coding standard, which is a new two-dimensional histogram translation method and has good performance in terms of embedding capacity and distortion reduction. Document "Two-dimensional reversible data hiding-based approach for intra-frame error concealment in H.264/AVC." (Xu D, wang R, signal Processing: image Communication, vol.47, pp.369-379,2016) proposes an H.264/AVC intra-frame error concealment scheme based on Two-dimensional reversible information concealment. In the prior art, a new two-dimensional histogram translation reversible information hiding method is provided by utilizing the Laplacian distribution rule of QDCT coefficients, namely combining the characteristic of more zero quantized DCT coefficients, and the motion vector generated in the inter-frame prediction process in the encoding process is used as secret information to be embedded into the zero quantized DCT coefficients of adjacent blocks as an error concealment means, and when the blocks are lost due to noise or other man-made tampering in the transmission process, the lost macro blocks can be recovered according to the extracted motion vector information. However, if two macro blocks are lost in one macro block group, there is a problem that the motion vector extraction fails, thereby affecting the error concealment effect of the video.
Disclosure of Invention
The invention aims to provide a reversible information hiding method based on improved difference expansion, which reduces the modification amount of embedding to a carrier by using the improved difference expansion method, ensures the quality of a carrier video, and simultaneously ensures that the original video data can be completely restored after information is extracted.
The technical scheme for realizing the purpose of the invention is as follows:
a reversible information hiding method based on improved difference expansion selects quantized 4X 4 brightness residual DCT coefficients extracted on I frames, adjacent QDCT coefficients in each 4X 4 block are formed into a pair by pairs and coefficient difference values are extracted, and the coefficient pairs to be embedded with information are modified to expand the difference values and achieve the purpose of embedding secret information; extracting the brightness 4 x 4DCT coefficient block of the embedded information at the decoding end, extracting the embedded information and recovering the original carrier data; mainly comprises the following steps:
(1) Reversible information hiding embedded block selection and information embedding based on improved difference expansion
The method comprises the steps that information embedding is carried out on a brightness residual QDCT coefficient selected as an I frame, and in the intra-frame prediction process of H.264/AVC coding, an embedding modification quantity on a residual coefficient block currently carrying out embedding operation is spread to other residual coefficient blocks to generate a distortion drift phenomenon; to suppress distortion drift, the coding is performed based on the coded sub-block B i,j And adjacent subblock B to be encoded i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Setting M in the positional relationship of (2) and 4×4 block prediction mode i,j+1 、M i+1,j+1 、M i+1,j 、M i+1,j-1 Respectively represent B i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Two screening conditions are proposed for intra prediction modes of (a):
1)CON1：M i,j+1 ∈{0,3,7} 4×4 U{0} 16×16 ；
2)CON2：M i+1,j-1 ∈{0,1,2,4,5,6,8} 4×4 U{0,1,2,3} 16×16 and M is i+1,j ∈{1,8} 4×4 U{1} 16×16 ；
Reading a code stream after H.264 entropy coding, extracting a QDCT brightness residual coefficient of an I frame and intra-frame prediction information of each brightness sub-block through entropy decoding, screening each 4X 4 sub-block through a prediction mode, and implementing embedding operation on the sub-blocks meeting CON1 and CON2 simultaneously; extracting medium-high frequency coefficients from each 4×4 sub-block coefficient capable of embedding information, dividing the coefficients into two-by-two and one group of coefficient pairs, performing expansion operation on the difference value of each coefficient pair, and embedding secret information to obtain a secret-carrying video;
(2) Extraction of embedded information and carrier recovery
The extraction of embedded data is the inverse of the embedding process, firstly, the quantized DCT coefficient after entropy decoding and the intra-frame prediction information of each brightness sub-block are obtained, 4 multiplied by 4 sub-blocks embedded with secret information are selected through a prediction mode, and the processes of information extraction and carrier recovery are carried out.
Further, the middle-high frequency part of the DCT coefficient in the steps (1) and (2) is obtained by selecting after performing inverse Z-shaped transformation and inverse run-length coding on the quantized DCT coefficient.
Further, the direct current coefficient in the 44 block selected to be embedded is a non-zero coefficient.
The steps in the actual implementation can be further described in terms of the following equivalents:
and before embedding, screening the embedded 4 multiplied by 4 blocks, inhibiting intra-frame distortion drift, and improving the quality of the loaded video. The information embedding mainly comprises the following steps:
(1) Screening of embedded blocks
In the H.264/AVC codec standard, intra-prediction is used as an important coding technique for eliminating video spatial redundancy, and H.264/AVC coding is based on blocks and uses different prediction directions as corresponding coding modes. For a luminance block, h.264 performs intra prediction using two block modes, 4×4 and 16×16, and generates a prediction residual by subtracting the current coding block from a prediction block of the current coding block, which is predicted from a reconstructed macroblock of a previous coded macroblock. Assuming that information is embedded in the quantized residual coefficient, that is, the residual QDCT coefficient is modified, a difference is necessarily generated between the quantized residual coefficient and the original reconstructed reference block in the reconstruction process at the decoding end, and this difference directly affects the intra-frame prediction process of other macro blocks in the later decoding process, so that a prediction result different from that obtained when data is not embedded originally is obtained in the decoding process. The modification is directly carried out on the QDCT coefficient of the prediction residual to achieve the effect of information embedding, the modification of the current QDCT coefficient is likely to be diffused into other macro blocks and sub blocks through intra-frame prediction in the following intra-frame prediction process, and the intra-frame prediction mode is a key factor influencing intra-frame distortion drift, so that two screening conditions are proposed according to the intra-frame prediction mode to screen the embedded block; meanwhile, the embedded block must satisfy that the direct current coefficient is a non-zero coefficient.
(2) Information embedding and extraction and carrier data recovery
Extracting brightness residual coefficients of 4 multiplied by 4 sub-blocks capable of embedding information, and scanning according to Zigzag scanning order to obtain a one-dimensional sequence Y= { Y k |k∈[0,15]}. Because the absolute value of the low-frequency coefficient is larger, and the distortion caused by modifying the low-frequency coefficient to the video is more obvious, the method selects the coefficient pairs which divide the medium-frequency coefficient into two groups of high-frequency coefficients, such as { (Y) 4 ,Y 5 ),(Y 6 ,Y 7 ),(Y 8 ,Y 9 ),(Y 10 ,Y 11 ),(Y 12 ,Y 13 ),(Y 14 ,Y 15 ) }. Calculating the difference value of each coefficient pair, judging whether the coefficient pair is used for embedding information according to the difference value, and modifying the coefficient pair according to the improved difference value expansion method provided by the invention if the embedding condition is met, so as to achieve the aim of embedding information; if the embedding condition is not satisfied, translating the coefficient pair difference value along the increasing direction of the absolute value so as to ensure the lossless recovery of the carrier data.
The extraction and recovery process is performed at the decoding end of the encoder, which is the inverse of the embedding process. Firstly, selecting sub-blocks of the carrier 4X 4QDCT coefficients according to the screening conditions of the embedded blocks, extracting the medium-high frequency coefficients in the sub-blocks, and carrying out coefficient modification and information extraction on each group of coefficient pairs of each carrier sub-block according to the inverse process of the embedding method. Firstly judging whether the coefficient pairs contain secret information bits according to the difference value of each group of coefficient pairs, if so, extracting secret information and carrying out numerical recovery on the carrier coefficient pairs; if not, only the carrier coefficients are subjected to the restoration operation.
Compared with the existing reversible video information hiding method, the method improves the difference expansion based on the difference expansion principle, and reduces the total modification quantity of the coefficient pairs during information embedding; the problem of motion vector extraction failure caused by the loss of a plurality of macro blocks is solved; meanwhile, the coefficient pairs are selectively embedded, so that the expansion amplitude of the coefficient difference value when the coefficient difference value is doubled is effectively controlled, the embedding distortion is effectively reduced, and the quality of the video is improved.
Drawings
Fig. 1 is a positional relationship of adjacent sub-blocks at the time of intra prediction of a 4×4 luminance sub-block.
Fig. 2 is a process of information embedding.
Fig. 3 shows the capacity test results (units bits) under different quantization parameters QP.
Fig. 4 and fig. 5 are respectively an I-frame image 2 of 7 original video sequences and a dense video frame image corresponding to the I-frame image 2 in experimental simulation.
Detailed Description
The technical scheme of the invention is further described below with reference to the accompanying drawings.
The method comprises two parts of information embedding, information extracting and video carrier recovering.
1. The information embedding part mainly comprises the following steps:
at the coding end of the H.264/AVC video coder, the image frame is divided into non-overlapped 16X 16 macro block MB, then the operations of prediction coding, integer DCT transformation, quantization, reordering and entropy coding are carried out on the prediction residual in turn, and then the operation is carried out in the form of code stream through NAL, which is aimed at the coding operation of one MB, when all MB on one frame of image is coded, the coding operation of one frame of image is finished, the method provided by the invention is to implement information embedding on the quantized brightness DCT coefficient block. After entropy encoding, the QDCT brightness residual coefficient of the I frame and the intra-frame prediction information of each brightness sub-block are extracted through entropy decoding, 4 multiplied by 4 sub-blocks which can be embedded are screened out according to the block selection condition for inhibiting intra-frame distortion drift, the middle and high frequency coefficients in the sub-blocks are extracted to be divided into two-by-two coefficient pairs, the difference value of the coefficient pairs is expanded, secret information is embedded at the same time, and the modified carrier data is subjected to entropy encoding again, so that final carrier-secret video data are obtained. The method comprises the following specific steps:
(1) Selection of an embedded block:
the coded sub-block B during the coding is given in FIG. 1 i,j And adjacent subblock B to be encoded i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Positional relationship between according to B i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Prediction mode (set M i,j+1 、M i+1,j+1 、M i+1,j 、M i+1,j-1 Respectively represent B i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Intra prediction mode of (a) and each mode corresponds to B to be referred to i,j The residual coefficient d, h, l, m, n, o, p in (2) proposes two conditions:
1)CON1：M i,j+1 ∈{0,3,7} 4×4 U{0} 16×16 。
2)CON2：M i+1,j-1 ∈{0,1,2,4,5,6,8} 4×4 U{0,1,2,3} 16×16 and M is i+1,j ∈{1,8} 4×4 U{1} 16×16 。
When the embedded sub-block meets the condition 1, the modification on the rightmost column residual coefficient d, h, l, p of the embedded sub-block will not affect the coding process of the adjacent sub-block on the right; when the embedded sub-block satisfies condition 2, the residual coefficient m, n, o, p of its lowest row will not affect the subsequent encoding of neighboring lower left and right neighboring sub-blocks. By utilizing the two conditions, the embedded block is screened in the information embedding process to determine whether the current block can be used for information hiding, so that distortion drift can be effectively restrained, and video quality degradation caused by information embedding is controlled.
The DC coefficient in each 4×4QDCT sub-block concentrates most of the energy of the whole sub-block, and if the DC coefficient is modified on the sub-block with zero, the overall energy of the whole sub-block is obviously changed, which affects the video quality, so that the 4×4 sub-block with the DC coefficient being a non-zero coefficient and meeting the two conditions of CON1 and CON2 is selected for information embedding.
(2) The embedding process comprises the following steps:
since DC coefficients and low frequency coefficients have most energy, embedding data at these locations results in a more significant reduction in video quality and increase in bit rate. Based on this, the present patent chooses to embed in the medium-high frequency DCT coefficients. Fig. 2 is a diagram showing an information embedding process of the present invention, in which a code stream after h.264 entropy encoding is read, a QDCT luminance residual coefficient of an I frame and intra-frame prediction information of each luminance sub-block are extracted through entropy decoding, each 4×4 sub-block is filtered through a prediction mode, an embedding operation is performed on the filtered sub-block, and after the completion, the same operation steps are performed on the next 4×4 sub-block.
After entropy decoding, extracting the luminance residual coefficient of the 4×4 sub-block capable of embedding information, and scanning according to the Zigzag scanning order to obtain a one-dimensional sequence Y= { Y k |k∈[0,15]}. The patent has the selection number of [4,15 ]]The DCT coefficients between are embedded in the data. Dividing the mid-high frequency partial coefficients of the read one-dimensional residual coefficient sequence Y into pairs of coefficients, e.g., { (Y) 4 ,Y 5 ),(Y 6 ,Y 7 ),(Y 8 ,Y 9 ),(Y 10 ,Y 11 ),(Y 12 ,Y 13 ),(Y 14 ,Y 15 ) And calculating the difference value h=x-y of each group of coefficient pairs (x, y), and setting the information bit m epsilon {0,1} to be embedded.
The coefficient with the difference value larger than-2 is selected to be embedded, namely when h is more than or equal to-1, the embedding process is as follows:
in order to ensure the reversibility of the algorithm and ensure that secret information can be accurately extracted, other difference coefficient pairs (i.e. h < -1) which do not meet the embedding requirement are required to be subjected to the following operation, and the process is to carry out histogram translation on the difference values of the residual coefficient pairs.
x′＝x-1,y′＝y+1
The difference h of the transformed residual coefficient pair (x ', y') * When the value is changed to 2h-m, secret information can be accurately extracted according to the following formula during decoding.
m＝(x′-y′)mod2
2. Information extraction and h.264 compressed video carrier recovery:
extraction and recovery of algorithmsThe process is performed at the decoding end of the encoder, and the extraction process is the inverse of the embedding process. According to the selection requirement of the embedded block, the intra-frame prediction mode is extracted at the decoding end in the same way, 4 multiplied by 4 sub-blocks embedded with secret information are selected, and the coefficient pair of each sub-block is { (Y) 4 ,Y 5 ),(Y 6 ,Y 7 ),(Y 8 ,Y 9 ),(Y 10 ,Y 11 ),(Y 12 ,Y 13 ),(Y 14 ,Y 15 ) -calculating the difference h of each set of coefficient pairs (x ', y').
If h * And (3) accurately extracting the secret information m according to the following formula.
m＝(x′-y′)mod2
The coefficients are then modified according to the following operation to recover the carrier data.
If h * And (3) indicating that the coefficient pair does not contain secret information, and recovering the carrier coefficient according to the following formula.
x＝x′+1,y＝y′-1
The effect of the method of the present invention can be verified by the following performance analysis.
1. Embedded capacity
To test the performance of the method, the method was successfully applied to the H.264/AVC test platform JM12.0 (http:// iphome. Hhi. De/suehring/tml/download/old_jm /). The main parameter settings are as in table 1, with the exception of the parameters mentioned, the other parameters remain default. In experimental simulations, 7 standard video sequences "Akiyo", "Salesman", "Hall", "Foreman", "car", "News" and "Mobile" with a resolution of 176×144 were used as test sequences, and the embedding capacity corresponding to each video test sequence is shown in table 2. Fig. 3 shows the variation of the present method for different video embedding capacities at different QPs, and it can be seen from fig. 3 that the embedding capacity gradually increases as the QP decreases.
2. Imperceptibility
Fig. 4 and 5 are respectively the 2 nd I-frame before and after the 7 video sequences embedded with information, and it is apparent that the encrypted video frame is visually imperceptible from the original video frame.
The judged video quality is compared by peak signal to noise ratio (PSNR). PSNR represents the difference in quality of an image after compression compared to the original image, and can be calculated from the Mean Square Error (MSE) as shown in the following equation:
wherein f' (x i ,y j ) Representing the pixel value of the reconstructed image in row j, f (x) i ,y j ) Representing the pixel value of the original image in row j and column i. The calculation of PSNR is shown in the following formula:
where MAX represents the maximum number of pixels, typically 255 in the video sequence.
The test result of the peak value-to-noise ratio PSNR value of the loaded video is shown in table 3, and the PSNR change average value before and after embedding the secret information is 0.32dB, so that the video reversible information hiding algorithm has good visual imperceptibility.
3. Bit rate variation
Considering the practicability of the video coding standard, the code rate control becomes an important technology of the video coding standard, and the code rate is also an important index reflecting the coding performance of the encoder. The Bit Rate Increase (BRI) reflects the change in data size of the encrypted video relative to the original video file. The calculation formula of the bit rate increase rate BRI is as follows, wherein BR marked Is the carrier video bit rate, BR org Is the bit rate of the original video.
Tables 4 and 5 are the values of the change in the encoder encoding bit rate before and after embedding information, respectively, and it is seen from the table that the bit rate change is lower than 0.5kbps and the bit growth amplitude is between 0.114% and 0.813%, so that the influence of the embedded information on the encoder is very small.
TABLE 1 JM12.0 parameter configuration
Table 2 embedded capacity test results
TABLE 3 imperceptibility test results
Table 4 bit rate comparison of video sequences before and after embedding information
Table 5 bit rate of video sequence after embedding information
Claims (3)
1. A reversible information hiding method based on improved difference expansion selects quantized 4X 4 brightness residual DCT coefficients extracted on I frames, adjacent QDCT coefficients in each 4X 4 block are formed into a pair by pairs and coefficient difference values are extracted, and the coefficient pairs to be embedded with information are modified to expand the difference values and achieve the purpose of embedding secret information; extracting the brightness 4 x 4DCT coefficient block of the embedded information at the decoding end, extracting the embedded information and recovering the original carrier data; the method comprises the following steps:
(1) Reversible information hiding embedded block selection and information embedding based on improved difference expansion
The brightness residual error QDCT coefficient of the I frame is selected to implement information embedding; to suppress distortion drift, the coding is performed based on the coded sub-block B i,j And adjacent subblock B to be encoded i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Setting M in the positional relationship of (2) and 4×4 block prediction mode i,j+1 、M i+1,j+1 、M i+1,j 、M i+1,j-1 Respectively represent B i,j+1 、B i+1,j+1 、B i+1,j 、B i+1,j-1 Two screening conditions are proposed for intra prediction modes of (a):
1)CON1：M i,j+1 ∈{0,3,7} 4×4 ∪{0} 16×16 ；
2)CON2：M i+1,j-1 ∈{0,1,2,4,5,6,8} 4×4 ∪{0,1,2,3} 16×16 and M is i+1,j ∈{1,8} 4×4 ∪{1} 16×16 ；
Reading a code stream after H.264 entropy coding, extracting a QDCT brightness residual coefficient of an I frame and intra-frame prediction information of each brightness sub-block through entropy decoding, screening each 4X 4 sub-block through a prediction mode, and implementing embedding operation on the sub-blocks meeting CON1 and CON2 simultaneously; extracting medium-high frequency coefficients from each 4×4 sub-block coefficient capable of embedding information, dividing the coefficients into two-by-two and one group of coefficient pairs, performing expansion operation on the difference value of each coefficient pair, and embedding secret information to obtain a secret-carrying video; extracting the brightness residual coefficient of the 4 multiplied by 4 sub-block capable of embedding information, and scanning according to the Zigzag scanning order to obtain a one-dimensional sequence Y= { Y k |k∈[0,15]-a }; the selection sequence number is [4,15 ]]DCT coefficient embedded data between the two, dividing the middle and high frequency part coefficients of the read one-dimensional residual coefficient sequence Y into two groupsThe difference h=x-y of each set of coefficient pairs (x, y) is calculated, the information bit m e 0,1 to be embedded is set,
the coefficient with the difference value larger than-2 is selected to be embedded, namely when h is more than or equal to-1, the embedding process is as follows:
in order to ensure the reversibility of the algorithm and ensure that secret information can be accurately extracted, other difference coefficient pairs which do not meet the embedding requirement, namely h < -1, are required to be subjected to the following operation, and the process is to carry out histogram translation on the difference value of the residual coefficient pairs:
x′＝x-1,y′＝y+1
the difference h of the transformed residual coefficient pair (x ', y') * Changing the time to 2h-m;
(2) Extraction of embedded information and carrier recovery
The extraction of embedded data is the inverse of the embedding process, and the quantized DCT coefficients after entropy decoding and the intra-prediction information of each luminance sub-block are obtained first, 4×4 sub-blocks embedded with secret information are selected by intra-prediction mode, and the coefficient pair { (Y) of each sub-block is obtained 4 ,Y 5 ),(Y 6 ,Y 7 ),(Y 8 ,Y 9 ),(Y 10 ,Y 11 ),(Y 12 ,Y 13 ),(Y 14 ,Y 15 ) -calculating the difference h of each set of coefficient pairs (x ', y') } * ；
If h * More than or equal to-3, the secret information m is accurately extracted according to the following formula,
m＝(x′-y′)mod2
the coefficients are then modified, according to the following operation, to recover the carrier data,
if h * And < -3 >, indicating that the coefficient pair does not contain secret information, recovering the carrier coefficient according to the following formula
x＝x′+1,y＝y′-1。
2. The improved difference extension-based reversible information hiding method according to claim 1, wherein intermediate and high frequency parts of said DCT coefficients of steps (1) and (2) are obtained by inverse zigzag transforming and inverse run-length coding the quantized DCT coefficients.
3. The improved difference extension based reversible information hiding method of claim 1, wherein the dc coefficient in the embedded 4 x 4 block is selected to be a non-zero coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910634241.4A CN110248193B (en) | 2019-07-12 | 2019-07-12 | Reversible information hiding method based on improved difference expansion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910634241.4A CN110248193B (en) | 2019-07-12 | 2019-07-12 | Reversible information hiding method based on improved difference expansion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110248193A CN110248193A (en) | 2019-09-17 |
| CN110248193B true CN110248193B (en) | 2023-06-27 |
Family
ID=67892224
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910634241.4A Active CN110248193B (en) | 2019-07-12 | 2019-07-12 | Reversible information hiding method based on improved difference expansion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110248193B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111918072A (en) * | 2020-08-12 | 2020-11-10 | 中国人民武装警察部队工程大学 | Method, system and device for modifying video steganography by combining trailing coefficients of N-dimensional cubic coding |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102110281A (en) * | 2011-02-18 | 2011-06-29 | 南京师范大学 | Method for embedding and extracting watermarks by adopting difference expansion aiming at shape point map layer data |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4762938B2 (en) * | 2007-03-06 | 2011-08-31 | 三菱電機株式会社 | Data embedding device, data extracting device, data embedding method, and data extracting method |
| WO2015028098A1 (en) * | 2013-09-02 | 2015-03-05 | Centum Research & Technology S.L.U. | Method and system for embedding information and authenticating a h.264 video using a digital watermark |
| CN105227962B (en) * | 2015-09-29 | 2018-01-05 | 西安空间无线电技术研究所 | A kind of lossless information concealing method based on data difference |
| CN105681803B (en) * | 2016-01-21 | 2018-08-31 | 宁波大学 | A kind of HEVC video information hiding methods of large capacity |
| CN105979268A (en) * | 2016-05-05 | 2016-09-28 | 北京智捷伟讯科技有限公司 | Safe information transmission method based on lossless watermark embedding and safe video hiding |
| WO2018177522A1 (en) * | 2017-03-29 | 2018-10-04 | Fundació Per A La Universitat Oberta De Catalunya | Video watermarking |
| CN107040787B (en) * | 2017-03-30 | 2019-08-02 | 宁波大学 | A kind of 3D-HEVC inter-frame information hidden method of view-based access control model perception |
| 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 |
-
2019
- 2019-07-12 CN CN201910634241.4A patent/CN110248193B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102110281A (en) * | 2011-02-18 | 2011-06-29 | 南京师范大学 | Method for embedding and extracting watermarks by adopting difference expansion aiming at shape point map layer data |
Non-Patent Citations (1)
| Title |
|---|
| 可逆信息隐藏算法综述;奚吉 等;《信息化研究》(第3期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110248193A (en) | 2019-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109427032B (en) | Scalable video watermark embedding and extracting method based on compressed domain | |
| JP2004518373A (en) | Watermarking of compressed information signal | |
| CN104581176B (en) | The insertion of H.264/AVC compression domain robust video watermark and extracting method without frame in error drift | |
| JP2014519215A (en) | Method and apparatus for lossy compression encoding data and corresponding method and apparatus for reconstructing data | |
| CN108366259A (en) | A kind of reversible steganography method of three coefficient of H.264/AVC video of high-fidelity | |
| Wang et al. | An Information Hiding Algorithm for HEVC Based on Angle Differences of Intra Prediction Mode. | |
| CN108769700A (en) | Reduce the robustness video steganography method of H.264 interframe drift distortion | |
| Zhou et al. | An intra-drift-free robust watermarking algorithm in high efficiency video coding compressed domain | |
| CN110248193B (en) | Reversible information hiding method based on improved difference expansion | |
| CN113329229A (en) | High-capacity hiding method for H.265 video information with high-efficiency fidelity | |
| Zhu et al. | Information hiding algorithm for H. 264 based on the motion estimation of quarter-pixel | |
| CN110072109B (en) | H.264/AVC video information hiding method for macroblock-level minimized distortion | |
| CN108683921B (en) | Video reversible information hiding method based on zero quantization DCT coefficient group | |
| KR101529903B1 (en) | Block-based depth map coding method and apparatus and 3D video coding method using the method | |
| CN109361926B (en) | Lossless reversible information hiding method for H.264/AVC video visual quality | |
| Hsu et al. | Blind video watermarking for H. 264 | |
| CN101534441A (en) | AVS video watermarking method based on gray theory and uniform spectrum theory | |
| CN114598887A (en) | Anti-recompression video watermarking method for controlling bit rate increase | |
| Bouchama et al. | Reversible data hiding scheme for the H. 264/AVC codec | |
| KR102242723B1 (en) | A method and an apparatus for decoding a video signal | |
| Xu et al. | Low complexity video watermarking algorithm by exploiting CAVLC in H. 264/AVC | |
| Ellinas | Reversible watermarking on stereo image sequences | |
| Farrugia | Reversible visible watermarking for H. 264/AVC encoded video | |
| CN114998081B (en) | Video blind watermark embedding and extracting method based on H.265/HEVC | |
| Joan et al. | Data Embedding with Dynamic Slice Quantisation in Versatile Video Coding |
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 |