CN113079274B - Encryption image reversible information hiding method of adaptive parameter binary tree mark - Google Patents
Encryption image reversible information hiding method of adaptive parameter binary tree mark Download PDFInfo
- Publication number
- CN113079274B CN113079274B CN202110368362.6A CN202110368362A CN113079274B CN 113079274 B CN113079274 B CN 113079274B CN 202110368362 A CN202110368362 A CN 202110368362A CN 113079274 B CN113079274 B CN 113079274B
- Authority
- CN
- China
- Prior art keywords
- image
- information
- parameter
- secret
- 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
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/32—Circuits or arrangements for control or supervision between transmitter and receiver or between image input and image output device, e.g. between a still-image camera and its memory or between a still-image camera and a printer device
- H04N1/32101—Display, printing, storage or transmission of additional information, e.g. ID code, date and time or title
- H04N1/32144—Display, printing, storage or transmission of additional information, e.g. ID code, date and time or title embedded in the image data, i.e. enclosed or integrated in the image, e.g. watermark, super-imposed logo or stamp
- H04N1/32149—Methods relating to embedding, encoding, decoding, detection or retrieval operations
- H04N1/32267—Methods relating to embedding, encoding, decoding, detection or retrieval operations combined with processing of the image
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/32—Circuits or arrangements for control or supervision between transmitter and receiver or between image input and image output device, e.g. between a still-image camera and its memory or between a still-image camera and a printer device
- H04N1/32101—Display, printing, storage or transmission of additional information, e.g. ID code, date and time or title
- H04N1/32144—Display, printing, storage or transmission of additional information, e.g. ID code, date and time or title embedded in the image data, i.e. enclosed or integrated in the image, e.g. watermark, super-imposed logo or stamp
- H04N1/32149—Methods relating to embedding, encoding, decoding, detection or retrieval operations
- H04N1/32267—Methods relating to embedding, encoding, decoding, detection or retrieval operations combined with processing of the image
- H04N1/32272—Encryption or ciphering
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/32—Circuits or arrangements for control or supervision between transmitter and receiver or between image input and image output device, e.g. between a still-image camera and its memory or between a still-image camera and a printer device
- H04N1/32101—Display, printing, storage or transmission of additional information, e.g. ID code, date and time or title
- H04N1/32144—Display, printing, storage or transmission of additional information, e.g. ID code, date and time or title embedded in the image data, i.e. enclosed or integrated in the image, e.g. watermark, super-imposed logo or stamp
- H04N1/32149—Methods relating to embedding, encoding, decoding, detection or retrieval operations
- H04N1/32267—Methods relating to embedding, encoding, decoding, detection or retrieval operations combined with processing of the image
- H04N1/32277—Compression
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
The invention provides a reversible information hiding method of an encrypted image marked by a self-adaptive parameter binary tree. For an image owner, firstly, carrying out vector quantization prediction on an original image to obtain a prediction error; secondly, obtaining an optimal parameter set for the original image by using a self-adaptive parameter binary tree marking method and marking the image block by using the optimal parameter set; and finally, encrypting the index value and the non-embeddable pixel of the image block and sending the encrypted index value and the non-embeddable pixel to an information hider. For the information hider, it encrypts the secret information with the encryption key and embeds the secret information into the vacated redundant space to obtain the secret-containing image. For the receiver, if the receiver has the encryption key, the original image can be obtained through decryption; if the hidden secret key is possessed by the mobile terminal, the hidden secret key can be used for decrypting and acquiring secret information; if the two keys are contained simultaneously, the original image and the secret information can be decrypted simultaneously. The scheme provided by the invention has better performance in embedding capacity.
Description
Technical Field
The invention belongs to the field of image content security of information security, and particularly relates to an encrypted image reversible information hiding method of a self-adaptive parameter binary tree mark.
Background
In a new round of scientific and technological revolution, the life style of people is greatly changed compared with the past, and the change is promoted by informatization to a great extent. The informatization not only changes the clothes and eating and living behaviors of people, but also continuously expands and meets the higher-level requirements of people. However, network viruses, trojan horses and illegal organizations pose a serious threat to the security of information while people enjoy the convenience of information-based activities. Therefore, the security of the information during transmission becomes especially important.
As an information transmission medium other than a language, images have become an indispensable part of the internet age. Image information hiding is a technology for effectively preventing information leakage, information is hidden in an image carrier by slightly modifying pixel values of an original image, and an unauthorized third party cannot detect image change. The carrier image hidden with the secret information through the image information hiding technology has certain distortion relative to the original image, but due to insensitivity of human eyes, attention of illegal stealers can be effectively avoided. Furthermore, the reversible information hiding technique is a branch of the information hiding technique, which can reconstruct the original image without loss after the receiving side acquires the secret information. Therefore, the reversible image information hiding technology is widely applied to the fields of military affairs and medical treatment. This type of information hiding technique has a drawback that the content of the original image is always exposed to the information hiding person. In order to solve the problem of exposure of the original image, reversible information hiding of the encrypted image is generated. The reversible information hiding technology for the encrypted image firstly encrypts the original image and then transmits the original image to the information hiding person, and the information hiding person can hide secret information without knowing the content of the original image. In addition, the receiver can independently restore the original image and extract the secret information.
To improve the ability of an encrypted image to carry secret information while keeping the encrypted image secure. The invention provides an encrypted image reversible information hiding method of a self-adaptive parameter binary tree mark.
Disclosure of Invention
The invention provides a method for hiding reversible information of an encrypted image based on vector quantization prediction and adaptive parameter binary tree markers. For an image owner, firstly, carrying out vector quantization prediction on an original image to obtain a prediction error; secondly, obtaining an optimal parameter set for the original image by using a self-adaptive parameter binary tree marking method and marking the image block by using the optimal parameter set; and finally, encrypting the index value and the non-embeddable pixel of the image block and sending the encrypted value to an information hider. For the information hider, it encrypts the secret information with the encryption key and embeds the secret information into the vacated redundant space to obtain the secret-containing image. For the receiver, if the receiver has the encryption key, the original image can be obtained through decryption; if the hidden secret key is possessed by the mobile terminal, the hidden secret key can be used for decrypting and acquiring secret information; if the two keys are contained simultaneously, the original image and the secret information can be decrypted simultaneously. Compared with other related work, the scheme provided by the invention has better performance in embedding capacity.
The technical scheme of the invention comprises the following steps:
a reversible information hiding method of encrypted image marked by adaptive parameter binary tree is used for secret communication among image owner, information hiding person and receiver, and comprises the following steps:
s1: an image owner predicts an original image by using vector quantization prediction to obtain a prediction error;
s2: using a self-adaptive parameter binary tree marking method to obtain an optimal parameter set, wherein the optimal parameter set comprises a plurality of different parameter pairs, and two parameter values in each pair of parameter pairs are respectively used for marking pixels with prediction errors within an acceptable range and pixels with prediction errors within an unacceptable range;
s3: performing adaptive parameter binary tree marking on each image block according to the obtained optimal parameter set to obtain a marked image, wherein a parameter pair used when any image block is marked is a parameter pair which enables the embedding capacity of the image block to be maximum in the optimal parameter set;
s4: encrypting the marked image according to the encryption key to obtain an encrypted image and sending the encrypted image to an information hiding person by an image owner;
s5: the information hiding person embeds the encrypted information into the redundant space of the encrypted image to obtain a secret-containing encrypted image and sends the secret-containing encrypted image to a receiver;
s6: the recipient extracts the secret information from the received secret-containing encrypted image and restores the original image.
Preferably, in S1, the method for the image owner to predict the original image by using vector quantization prediction to obtain the prediction error is as follows:
s11: cutting an original image I into blocks with W multiplied by W size and without overlapping with each other;
s12: carrying out vector quantization compression on the original image I to obtain an index table T I ；
S13: decompress the index Table T I Obtaining a decompressed image I VQ ；
S14: calculating the original image I and the decompressed image I pixel by pixel VQ The formula is as follows:
e(x，y)＝I(x，y)-I VQ (x，y)
wherein, I (x, y) and I VQ (x, y) respectively represent image I and image I VQ The pixel value at coordinate (x, y) and e (x, y) is the prediction error difference at the corresponding location.
Preferably, in S2, a method for obtaining an optimal parameter set by using an adaptive binary parameter tree marking method is as follows:
s21: slicing an original image I of size M × N into blocks B of size W × W that do not overlap with each other k Wherein k =1,2, \8230;, M/W × N/W;
s22: setting a parameter pair { (α, β) |1 ≦ α ≦ 7,1 ≦ β ≦ 7} for marking the prediction error of the image block, where α represents the number of binary encoding bits used to mark one pixel when the prediction error of the pixel is within the acceptable range, and β represents the number of 0-valued bits used to mark one pixel when the prediction error of the pixel is within the unacceptable range; first, set beta = alpha to optimize alpha, and obtain a parameter set composed of 4 sets of parameter pairs
All optional combinations of A, α j (t) represents the parameter alpha corresponding to the t group parameter pair in the jth combination of the selectable combinations A, wherein j is equal to the A, and 2 is equal to or larger than the alpha j (0)≤α j (1)≤α j (2)≤α j (3)≤7；
S23: for different parameter pairs (alpha) within optional combination A j (t)，α j (t)), the corresponding parameters n are calculated respectively α Wherein
S24: the embedding capacity of the image block is determined by sequentially using different parameter pairs in the parameter set: if the pixels in the block meet the condition
If the pixel is a storable pixel, otherwise, the pixel is a non-storable pixel, and the number N of the storable pixels in the block is counted em (α j (t)，α j (t)) and the number of non-storable pixels N nem (α j (t)，α j (t)), calculating an embedding capacity for the blocks of the time image using different parameters within the parameter set
And selecting the maximum embedding capacity
As the embedding capacity of the image block;
s25: obtaining a total embedding capacity of the image for all j in the optional combination A
n is the length of a single code word in the code book;
s26: according to EC = max j EC j ，
According to determined j opt Obtaining the optimal parameter set under different t values
S27: fixing alpha determined in S26 * (t) and for beta j′ (t)∈[1，α * (t)]Optimizing to obtain parameter set comprising 4 sets of parameter pairs
All optional combinations of B, β j′ (t) represents the parameter alpha, j 'epsilon B corresponding to the t group parameter pair in the j' th combination of the optional combination B;
s28: for different parameter pairs (alpha) within optional combination B * (t),β j′ (t)), calculating corresponding parameters n, respectively α In which
S29: the embedding amount of the image block is determined by sequentially using different parameter pairs in the parameter set: if the pixels in the block meet the condition
If the pixel is a storable pixel, otherwise, the pixel is a non-storable pixel, and the number N of the storable pixels in the block is counted em (α * (t),β j′ (t)) and the number of non-storable pixels N nem (α * (t),β j′ (t)), calculating the embedding capacity of the blocks of time images using different parameters within the parameter set
And selects the maximum embedding capacity
As the embedding capacity of the image block;
s210: obtaining a total embedding capacity of the image for all j's in the optional combination B
S211: according to EC = max j′ EC j′ ,
According to determined j' opt Obtaining the optimal parameter set under different t values
S212: according to the determined alpha * (t) and beta * (t) combining to obtain an optimal parameter set
Preferably, in S3, the method for performing adaptive parameter binary tree labeling on each image block according to the obtained optimal parameter set to obtain a block label image is as follows:
s31: for each image block, the MSBs of the first n pixels are reserved for recording the index value corresponding to the block;
s32: marking fixed parameter blocks with fixed parameter pairs (α, β) = (4, 3) for hiding optimal parameter sets
When marking, marking the storable pixel with 4 bits in the LSB of the pixel and marking the non-storable pixel with 3 bits until the space vacated by the image block exceeds 24 bits;
s33: for removingFor each image block except the fixed parameter block, the embedding capacity EC (t) = N is calculated by using each set of parameter pairs in the optimal parameter set respectively em ×(8-α * (t))-N nem ×β * (t) in which N em Is the number of the storable pixels in the image block, N nem Selecting the parameter pair capable of obtaining the highest block embedding quantity for the number of the non-storable pixels in the image block
Marks the image block and records t with the MSB of the n +1 th and n +2 th pixels opt Binary form (t) opt ) 2 。
Preferably, in S4, the image owner encrypts the marked image according to the encryption key as follows:
s41: encrypting a key K according to an image e Encrypting the index value of the block and recording the encrypted index value in the MSB of the first n pixels of the corresponding block;
s42: encrypting a key K according to an image e Encrypting the non-embeddable pixel, recording the LSB of beta bit and MSB of 1 bit of the original non-embeddable pixel as additional information to an embeddable space with a first priority, and recording the rest bits in the mark image;
s43: and sending the encrypted image to the information hider.
Preferably, in S5, the method for the information hiding person to embed the encrypted information into the redundant space of the encrypted image to obtain the secret-containing encrypted image and send the secret-containing encrypted image to the recipient is as follows:
s51: hiding a key K according to information h For original secret information S secret Carrying out encryption to obtain encrypted secret information;
s52: decoding is performed on (α, β) = (4, 3) using fixed parameters to obtain the optimal parameter set, and its correspondence (t) opt ) 2 ；
S53: (t) of each block is obtained from MSBs of the (n + 1) th and (n + 2) th pixels opt ) 2 To determine a redundant space for each image block, to embed encryption information in the redundant space,obtaining a secret-containing encrypted image;
s54: the secret-containing encrypted image is sent to the recipient.
Preferably, in S6, the method for the recipient to extract the secret information from the received secret-containing encrypted image and restore the original image is as follows:
s61: after a receiver receives the image containing the secret encryption, the image containing the secret encryption is cut into image blocks with W multiplied by W size and non-overlapping;
s62: decoding an image block using a fixed parameter pair (α, β) = (4, 3) to obtain an optimal parameter set, and its corresponding (t) opt ) 2 ；
S63: if the receiver has the information hidden secret key K h Determining an embeddable information space and an extra information space of the whole image according to the optimal parameter set, retrieving all embedded data, obtaining an encrypted secret stream by skipping the parameter set and the extra information, and hiding a secret key K with the information h Decryption is carried out to obtain the original secret information S secret ；
S64: if the receiver has the image encryption key K e Determining embeddable information space and extra information space of the whole image according to the optimal parameter set, recovering non-hidden pixels according to the extra information, and encrypting the secret key K by using the image e Decrypting the MSB of the first n pixels and the non-embeddable pixels, and reconstructing an original image I by using the mapped prediction error;
s65: if the receiver has the hidden secret key K h And an image encryption key K e Then the secret information S is extracted using both S63 and S64 secret And restoring the original image I.
Compared with the prior art, the invention has the advantages and beneficial effects that:
the invention provides a reversible information hiding method of an encrypted image marked by a self-adaptive parameter binary tree on the premise of ensuring the security of the encrypted image, and the reversible information hiding method has the following main beneficial effects: (1) Vector quantization compression is proposed as a prediction method; (2) The characteristics of the image block are fully exerted, and a method for marking a self-adaptive parameter binary tree is provided; (3) Compared with other existing related work, the scheme of the invention has superior performance and high hidden capacity.
Drawings
FIG. 1 is a flow chart of vector quantization compression;
FIG. 2 is a flow diagram of an adaptive parameter binary tree tagging;
FIG. 3 is a flow chart of image encryption;
FIG. 4 is a graph of the different effects of Lena and Baboon;
FIG. 5 is a histogram of the different effects of Lena and Baboon;
FIG. 6 shows that when the codebook length is 2 14 A prediction error of time;
FIG. 7 is a comparison of the present invention and related methods applied to five test images (7 bars in each test image represent, from left to right, puteaux et al,2018, puyang et al,2018, 2013, yiel at,2019, 2011, chen et al, 2019, wang et al,2020[5], wu et al,2020[6], a total of 7 methods of the present invention);
FIG. 8 is a comparison of the invention and the related methods applied to two databases (the 7 bars in each database represent, from left to right, puteaux et al,2018, puyang et al,2018, 2013, yiel at,2019, chen et al, 2019, wang et al,2020 5, wu et al,2020, 6, and the method of the invention, a total of 7 methods).
Detailed Description
The present invention will be described in further detail below with reference to the accompanying examples in order to facilitate the understanding and implementation of the present invention for those of ordinary skill in the art, and it should be understood that the embodiments described herein are only for illustration and explanation of the present invention and are not intended to limit the present invention.
Embodiments of the invention are described in further detail below with reference to the following drawings:
in a preferred embodiment of the present invention, an adaptive parameter binary tree marked encrypted image invertible information hiding method is provided, which is used for secret communication among an image owner, an information hiding person and a receiver, and comprises the following specific steps:
s1: the image owner predicts the original image using vector quantization prediction to obtain the prediction error.
In the present embodiment, the method for the image owner to predict the original image using vector quantization prediction to obtain the prediction error is as follows:
s11: cutting an original image I into blocks with W multiplied by W size and not overlapping with each other;
s12: carrying out vector quantization compression on the original image I to obtain an index table T I ；
S13: decompress the index Table T I Obtaining a decompressed image I VQ ；
S14: calculating the original image I and the decompressed image I pixel by pixel VQ The formula is as follows:
e(x，y)＝I(x，y)-I VQ (x，y)
wherein, I (x, y) and I VQ (x, y) respectively represent image I and image I VQ The pixel value at coordinate (x, y) and e (x, y) is the prediction error difference at the corresponding location.
In this embodiment, a gray-scale image with a size of 512 × 512 is selected as an original I; dividing an original image I into image blocks, wherein each image block comprises 4 multiplied by 4 pixels; vector quantization compression is actually performed by means of block coding, each block of an image being in a codebook (note: the length of the codebook needs to be more than 2) 8 And the present embodiment takes the form of 2 14 Code book of length) to find the closest code word and replace the image block with the index value corresponding to the code word to generate the index table T I (ii) a By decompressing the index table T I Replacing the corresponding image block with the code word corresponding to the index to complete image decompression and generate a decompressed image I VQ (ii) a Calculating an original calculated original image I and a decompressed image I VQ The prediction error e can be obtained from the difference values, and the vector quantization compression process is shown in fig. 1.
S2: and using an adaptive parameter binary tree marking method to obtain an optimal parameter set, wherein the optimal parameter set comprises a plurality of different parameter pairs, and two parameter values in each pair of parameter pairs are respectively used for marking pixels with prediction errors within an acceptable range and pixels with prediction errors within an unacceptable range.
The method for hiding reversible information of encrypted image based on parameter binary tree mark marks prediction error by parameter pair { (alpha, beta) |1 ≦ alpha ≦ 7,1 ≦ beta ≦ 7}, marks the pixel by alpha bit (uniquely mapped according to the prediction error) if the prediction error is in an acceptable range, and marks the pixel by beta bit (continuous 0) if the prediction error is not in the acceptable range
To mark different sub-blocks, i.e. (alpha) within the optimal parameter set * (t),β * (t)) has multiple sets, for different image blocks it is necessary to select the set of parameter pairs that best fits their characteristics to label. Optimal parameter set
Four groups of (a) * (t),β * (t)) needs to be determined through optimization, and as alpha is a determining factor determining the embedding capacity and beta has small influence on the embedding capacity, the method controls beta = alpha so as to determine the optimal alpha, and then obtains the optimal beta through further optimization by fixing alpha and fine-tuning alpha, thereby improving the hiding effect.
In this embodiment, the specific implementation steps of the method for obtaining the optimal parameter set by using the adaptive binary tree parameter tagging method are as follows in sequence:
s21: slicing an original image I of size M × N into blocks B of size W × W that do not overlap with each other k Wherein k =1,2, \8230;, M/W × N/W;
s22: the parameter pair { (α, β) |1 ≦ α ≦ 7,1 ≦ β ≦ 7} for labeling the prediction error of the image block is set, where α represents the number of binary encoding bits used to label one pixel when the prediction error of the pixel is within the acceptable range, and β represents the number of 0-valued bits used to label one pixel when the prediction error of the pixel is within the unacceptable range. For example, if the prediction error for a pixel is within an acceptable range, the pixel needs to be marked with a binary code of α bits, while if the prediction error for a pixel is within an unacceptable range, the pixel needs to be marked with 0 of β. It should be noted that the aforementioned binary codes correspond to prediction errors one-to-one, that is, one prediction error is uniquely mapped to one binary code of α bits.
Therefore, β = α is set first to optimize α first, and a parameter set consisting of 4 sets of parameter pairs is obtained
All optional combinations of (a) are designated as optional combinations A, α j (t) represents the parameter alpha corresponding to the t group parameter pair in the jth combination of the selectable combinations A, wherein j is equal to the A, and 2 is equal to or larger than the alpha j (0)≤α j (1)≤α j (2)≤α j (3)≤7；
Note: the invention is not considered because the embedding capacity is low when α =1, and α j (0)、α j (1)、α j (2) And alpha j (3) Are different and increasing values.
S23: for different parameter pairs (alpha) within optional combination A j (t),α j (t)), calculating corresponding parameters n, respectively α In which
S24: the embedding capacity of the image block is determined by sequentially using different parameter pairs in the parameter set: if the pixels in the block meet the condition
If the pixel is a storable pixel, otherwise, the pixel is a non-storable pixel, and the number N of the storable pixels in the block is counted em (α j (t)，α j (t)) and the number of non-storable pixels N nem (α j (t)，α j (t)), calculating an embedding capacity for the blocks of the time image using different parameters within the parameter set
And selecting the maximum embedding capacity
As the embedding capacity of the image block;
s25: obtaining a total embedding capacity of the image for all j in the optional combination A
n is the length of a single code word in the code book;
s26: according to EC = max j EC j ，
According to determined j opt Obtaining the optimal parameter set under different t values
S27: fixing alpha determined in S26 * (t) and p j′ (t)∈[1，α * (t)]Optimizing to obtain parameter set comprising 4 sets of parameter pairs
All optional combinations of (a) are denoted as optional combinations B, β j′ (t) represents the parameter alpha, j 'epsilon B corresponding to the t group parameter pair in the j' th combination of the optional combination B;
s28: for different parameter pairs (alpha) within optional combination B * (t)，β j′ (t)), the corresponding parameters n are calculated respectively α Wherein
S29: the embedding amount of the image block is determined by sequentially using different parameter pairs in the parameter set: if the pixels in the block are eligible
If the pixel is a storable pixel, otherwise, the pixel is a non-storable pixel, and the number N of the storable pixels in the block is counted em (α * (t),β j′ (t)) and the number of non-storable pixels N nem (α * (t),β j′ (t)), calculating an embedding capacity for the blocks of the time image using different parameters within the parameter set
And selects the maximum embedding capacity
As the embedding capacity of the image block;
s210: obtaining a total embedding capacity of images for all j' in the selectable combination B
S211: according to EC = max j′ EC j′ ,
According to determined j' opt Obtaining the optimal parameter set under different t values
S212: according to the determined alpha * (t) and beta * (t) combining to obtain an optimal parameter set
S3: and performing adaptive parameter binary tree marking on each image block according to the obtained optimal parameter set to obtain a marked image, wherein the parameter pair used when any image block is marked is the parameter pair which enables the embedding capacity of the image block to be maximum in the optimal parameter set.
In this embodiment, the method for performing adaptive binary tree labeling on each image block according to the obtained optimal parameter set to obtain a block label image is as follows:
s31: for each image block, the MSBs of the first n pixels are reserved for recording the index value corresponding to the block;
s32: marking fixed parameter blocks with fixed parameter pairs (α, β) = (4, 3) for hiding optimal parameter sets
When marking, the storable pixel is marked with 4 bits in the LSB of the pixel and the non-storable pixel is marked with 3 bits until the space vacated by the image block exceeds 24 bits (note: parameter set)
4 parameter pairs are included, each of which is represented by 3 bits, i.e. 24 bits in total, and the number of fixed parameter blocks depends on the amount of space vacated);
s33: for each image block except the fixed parameter block, the embedding capacity EC (t) = N is calculated using each set of parameter pairs within the optimal parameter set, respectively em ×(8-α * (t))-N nem ×β * (t), i.e. α for LSB of pixel * (t) bits to mark the occupiable pixels and beta * (t) bits to mark non-occupiable pixels, where N em Is the number of pixels, N, that can be stored in the image block nem Selecting the parameter pair capable of obtaining the highest block embedding amount for the number of the non-storable pixels in the image block
Marks the image block and records t with the MSB of the n +1 th and n +2 th pixels opt Binary form (t) opt ) 2 (note: if it is a storable pixel, it is advanced according to the prediction errorLine-only mapping, using alpha * (t) bit to mark the pixel, if it is an unconcealable pixel, using beta * (t) consecutive zero marks of bits).
In this embodiment, the process of marking fixed parameter blocks using fixed parameter pairs (α, β) = (4, 3) and marking image blocks using a certain parameter pair within the optimal parameter set is as in fig. 2, assuming that for Lena images, the optimal parameter set has been obtained in S2
To (3, 2), (4, 3), (5, 4) and (6, 5), which are converted to binary form of (011, 010), (100, 011), (101, 100) and (110, 101), these 24 bits will be buried into the fixed parameter block. The fixed parameter block is marked by a fixed parameter pair (alpha, beta) = (4, 3), wherein Na =14, a prediction error is obtained through image prediction, if the prediction error meets the condition that-7 is not less than e (i, j) is not less than 6, the prediction error is mapped to a binary code of 4 bits according to a mapping table and the LSB of the pixel is replaced, and if the prediction error does not meet the condition, the LSB of the pixel which cannot be embedded is replaced by 3 bits of '000'; for each image block except the fixed parameter block, the embedding amount is calculated by using the parameter pair (3, 2), (4, 3), (5, 4) and (6, 5), respectively, and it is found that the embedding amount is 60 bits, which is the maximum value when the parameter pair is (3, 2), so the parameter pair (3, 2) is the most suitable parameter pair for the block, and the parameter pair is marked with '0' at the MSB of the 15 th and 16 th bits.
S4: and encrypting the marked image according to the encryption key to obtain an encrypted image, and sending the encrypted image to the information hiding person by the image owner.
In this embodiment, the method for the image owner to encrypt the marked image according to the encryption key is as follows:
s41: encrypting a key K according to an image e Encrypting the index value of the block and recording the encrypted index value in the MSB of the first n pixels of the corresponding block;
s42: encrypting a key K according to an image e Encrypting the non-embeddable pixel and using the LSB of the beta bit and MSB of 1 bit of the original non-embeddable pixel as additional informationRecording a priority to the embeddable space, and recording the rest of bits in the marked image;
s43: and sending the encrypted image to an information hiding person.
In this embodiment, the process of image encryption is as shown in FIG. 3, the encryption key K e The index value and the non-embeddable pixel are encrypted at the same time, and furthermore, the LSB of the β bit and the MSB of the 1 bit are recorded as additional information to the embeddable space at the first priority.
S5: the information hider embeds the encrypted information into the redundant space of the encrypted image to obtain a secret-containing encrypted image, and sends the secret-containing encrypted image to a receiver.
In this embodiment, the method for the information hider to embed the encrypted information into the redundant space of the encrypted image to obtain the secret-containing encrypted image and send the secret-containing encrypted image to the receiver is as follows:
s51: hiding a key K according to information h For original secret information S secret Carrying out encryption to obtain encrypted secret information;
s52: decoding is performed using a fixed parameter pair (α, β) = (4, 3) to obtain an optimal parameter set, and its correspondence (t) opt ) 2 ；
S53: (t) of each block is obtained from MSBs of the (n + 1) th and (n + 2) th pixels opt ) 2 Determining a redundant space of each image block, and embedding encryption information in the redundant space to obtain a secret-containing encrypted image;
s54: the steganographically encrypted image is sent to the recipient.
S6: the recipient extracts the secret information from the received secret-containing encrypted image and restores the original image.
In this embodiment, the method for the recipient to extract the secret information from the received secret-containing encrypted image and to recover the original image is as follows:
s61: after a receiver receives the image containing the secret encryption, the image containing the secret encryption is cut into image blocks with W multiplied by W size and non-overlapping;
s62: decoding an image block using a fixed parameter pair (α, β) = (4, 3) to obtain an optimal parameter set, and its corresponding (t) opt ) 2 ；
S63: if the receiver has the information hiding key K h Determining an embeddable information space and an extra information space of the entire image according to the optimal parameter set, retrieving all embedded data, obtaining an encrypted secret stream by skipping the parameter set and the extra information, and hiding a secret key K with the information h Decryption is carried out to obtain the original secret information S secret ；
S64: if the receiver has the image encryption key K e Determining embeddable information space and extra information space of the whole image according to the optimal parameter set, recovering non-hidden pixels according to the extra information, and encrypting the secret key K by using the image e Decrypting the MSB of the first n pixels and the non-embeddable pixels, and reconstructing an original image I by using the mapped prediction error;
s65: if the receiver has the information hiding key K at the same time h And an image encryption key K e Then the secret information S is extracted using both S63 and S64 secret And restoring the original image I.
In this embodiment, the image owner obtains the encrypted image, the information-hiding person obtains the confidential image, and the receiver receives the confidential image and extracts the confidential information and the original image therefrom, according to the foregoing steps S1 to S6. The specific results are shown below:
i. security analysis
Fig. 4 gives the corresponding simulation results of Lena and babon as examples. Wherein fig. 4 (b) and 4 (c) are encrypted images, and fig. 4 (e) and 4 (f) are visual effect diagrams of the secrecy-containing encrypted images. Obviously, the four images can well conceal the information of the original image and the secret information, so that any useful information cannot be perceived. In addition, fig. 5 (a) and 5 (d) are histogram information of the original image, and unlike the histogram distribution thereof, the histogram distributions shown in fig. 5 (b), 5 (c), 5 (e) and 5 (f) exhibit uniform distribution, indicating that a malicious thief cannot acquire the original image information and the secret information by analyzing the encrypted image and the secret-containing image. Finally, table 1 shows the PSNR and SSIM values of the encrypted image and the secret image, and it can be seen from table 1 that both the PSNR value does not exceed 10 and the SSIM value is close to 0, which further verifies that the present invention has very high security.
TABLE 1 PSNR and SSIM for encrypted images and images containing secrets
Performance analysis
On the premise of ensuring the security of the encrypted image, the hiding capacity is one of important measures of the performance of the RDHEI method. Table 2 shows when the codebook is 2 12 、2 13 And 2 14 The amount of embedding of different test images. As the codebook becomes longer, the vector quantization prediction becomes more accurate, and FIG. 6 shows that when the length of the codebook is 2 14 The accuracy of the prediction is then vector quantized. Although a longer codebook needs to be recorded with a longer index value, it can be observed from table 2 that as the codebook grows, the embedding amount is also promoted. Finally, it can be observed that images with simpler textures enable higher embedding volumes, while images with more complex textures are lower.
Table 2 shows when the codebook length is 2 12 、2 13 And 2 14 Amount of embedding of temporally different images
iii comparison of Performance
FIGS. 7 and 8 compare the average net hiding capacity of images provided by the present invention and other related methods [1-6] in five typical grayscale images and two image databases, respectively. From fig. 7, it can be seen that the method provided by the invention is superior to other related works except "man", and the embedding amount of the invention exceeds 2bpp, which cannot be realized by other schemes. In addition, as shown in FIG. 8, the present invention can achieve embedding amounts of 3.4863bpp and 3.1205bpp for the experimental results of the databases BOWS-2 and BOSSBase, respectively.
The above comparative methods are specifically described in the following references:
[1]S.Yi and Y.C.Zhou,“Separable and reversible data hiding in encrypted images using parametric binary tree labeling,”IEEE Transactions on Multimedia,vol.21,no.1,pp.51–64,2019.
[2]P.Puteaux and W.Puech,“An efficient msb prediction-based method for high-capacity reversible data hiding in encrypted images,”IEEE Transactions on Information Forensics and Security,vol.13,no.7,pp.1670–1681,2018.
[3]Y.Puyang,Z.Yin and Z.Qian,“Reversible Data Hiding in Encrypted Images with Two-MSB Prediction,”2018IEEE International Workshop on Information Forensics and Security(WIFS),Hong Kong,Hong Kong,2018,pp.1-7,doi:10.1109/WIFS.2018.8630785.
[4]K.M Chen and C.-C.Chang,“High-capacity reversible data hiding in encrypted images based on extended run-length coding and block-based msb plane rearrangement,”Journal of Visual Communication and Image Representation,vol.58,no.2019,pp.334–344,2019.
[5]P.Wang,B.Cai,S.Xu and B.Chen,“Reversible Data Hiding Scheme Based on Adjusting Pixel Modulation and Block-Wise Compression for Encrypted Images,”IEEE Access,vol.8,pp.28902-28914,2020,doi:10.1109/ACCESS.2020.2972622.
[6]Y.Wu,Y.Xiang,Y.Guo,J.Tang and Z.Yin,“An Improved Reversible Data Hiding in Encrypted Images Using Parametric Binary Tree Labeling,”IEEE Transactions on Multimedia,vol.22,no.8,pp.1929-1938,Aug.2020.DOI:10.1109/TMM.2019.2952979.
Claims (5)
1. a reversible information hiding method of encrypted image marked by adaptive parameter binary tree is used for secret communication among image owner, information hiding person and receiver, and is characterized by comprising the following steps:
s1: an image owner predicts an original image by using vector quantization prediction to obtain a prediction error;
s2: using a self-adaptive parameter binary tree marking method to obtain an optimal parameter set, wherein the optimal parameter set comprises a plurality of different parameter pairs, and two parameter values in each pair of parameter pairs are respectively used for marking pixels with prediction errors within an acceptable range and pixels with prediction errors within an unacceptable range;
s3: performing adaptive parameter binary tree marking on each image block according to the obtained optimal parameter set to obtain a marked image, wherein a parameter pair used when any image block is marked is a parameter pair which enables the embedding capacity of the image block to be maximum in the optimal parameter set;
s4: encrypting the marked image according to the encryption key to obtain an encrypted image and sending the encrypted image to an information hiding person by an image owner;
s5: the information hiding person embeds the encrypted information into the redundant space of the encrypted image to obtain a secret-containing encrypted image and sends the secret-containing encrypted image to a receiver;
s6: the receiver extracts the secret information from the received secret-containing encrypted image and restores the original image;
in S1, the method for the image owner to predict the original image using vector quantization prediction to obtain the prediction error is as follows:
s11: cutting an original image I into blocks with W multiplied by W size and without overlapping with each other;
s12: carrying out vector quantization compression on the original image I to obtain an index table T I ；
S13: decompress the index Table T I Obtaining a decompressed image I VQ ；
S14: calculating the original image I and the decompressed image I pixel by pixel VQ The formula is as follows:
e(x,y)＝I(x,y)-I VQ (x,y)
wherein, I (x, y) and I VQ (x, y) respectively represent image I and image I VQ Pixel value at coordinate (x, y), and e (x, y) is the prediction error difference at the corresponding location;
in S2, a method for obtaining an optimal parameter set by using an adaptive binary tree parameter tagging method is as follows:
s21: will be of size M × NThe original image I is divided into W size blocks B which are not overlapped k Wherein k =1,2, \8230;, M/W × N/W;
s22: parameter pairs alpha, beta) |1 ≦ alpha ≦ 7,1 ≦ beta ≦ 7} for labeling prediction errors for image blocks are set, where alpha represents the parameter pair for labeling a pixel when the prediction error for that pixel is within an acceptable range
A binary encoding bit number, beta represents a 0-valued bit number used for marking a pixel when the prediction error of the pixel is in an unacceptable range; first, set beta = alpha to optimize alpha, and obtain a parameter set composed of 4 sets of parameter pairs
All optional combinations of A, α j (t) represents the parameter alpha corresponding to the t-th group of parameter pairs in the j-th combination of the selectable combination A, wherein j belongs to A, and 2 is larger than or equal to alpha j (0)≤α j (1)≤α j (2)≤α i (3)≤7；
S23: for different parameter pairs (alpha) within optional combination A j (t),α j (t)), calculating corresponding parameters n, respectively α In which
S24: the embedding capacity of the image block is determined by using different parameter pairs in the parameter set in turn: if the pixels in the block meet the condition
If the pixel is a storable pixel, otherwise, the pixel is a non-storable pixel, and the number N of the storable pixels in the block is counted em (α j (t),α j (t)) and the number of non-storable pixels N nem (α j (t),α j (t)), calculating the embedding capacity of the blocks of time images using different parameters within the parameter set
And selects the maximum embedding capacity
As the embedding capacity of the image block;
s25: obtaining a total embedding capacity of the image for all j in the optional combination A
n is the length of a single code word in the code book;
s26: according to
According to determined j opt Obtaining the optimal parameter set under different t values
S27: fixing alpha determined in S26 * (t) and p j′ (t)∈[1,α * (t)]Optimizing to obtain parameter set comprising 4 sets of parameter pairs
All optional combinations of B, β j′ (t) represents the parameter alpha, j 'epsilon B corresponding to the t group parameter pair in the j' th combination of the selectable combination B;
s28: for different parameter pairs (alpha) within optional combination B * (t),β j′ (t)), calculating corresponding parameters n, respectively α Wherein n is α ＝2 α*(t) -1；
S29: the embedding amount of the image block is determined by sequentially using different parameter pairs in the parameter set: if the pixels in the block are eligible
If the pixel is a storable pixel, otherwise, the pixel is a non-storable pixel, and the number N of the storable pixels in the block is counted em (α * (t),β j′ (t)) and the number of non-storable pixels N nem (α * (t),β j′ (t)), calculating the embedding capacity of the blocks of time images using different parameters within the parameter set
And selecting the maximum embedding capacity
As the embedding capacity of the image block;
s210: obtaining a total embedding capacity of the image for all j's in the optional combination B
S211: according to
According to determined j' opt Obtaining the optimal parameter set under different t values
S212: according to the determined alpha * (t) and beta * (t) combining to obtain an optimal parameter set
2. The method according to claim 1, wherein in S3, the method for adaptively labeling the binary tree of parameters for each image block according to the obtained optimal parameter set to obtain the image with the block label comprises:
s31: for each image block, the MSBs of the first n pixels are reserved for recording the index value corresponding to the block;
s32: marking fixed parameter blocks with fixed parameter pairs (α, β) = (4, 3) for hiding optimal parameter sets
When marking, marking the storable pixel with 4 bits in the LSB of the pixel and marking the non-storable pixel with 3 bits until the space vacated by the image block exceeds 24 bits;
s33: for each image block except the fixed parameter block, the embedding capacity EC (t) = N is calculated using each set of parameter pairs within the optimal parameter set, respectively em ×(8-α * (t))-N nem ×β * (t) wherein N em Is the number of the storable pixels in the image block, N nem Selecting the parameter pair capable of obtaining the highest block embedding quantity for the number of the non-storable pixels in the image block
Marks the image block and records t with the MSB of the n +1 th and n +2 th pixels opt Binary form (t) opt ) 2 。
3. The method according to claim 2, wherein in S4, the image owner encrypts the marked image according to the encryption key as follows:
s41: encrypting a key K according to an image e Encrypting the index value of the block and recording the encrypted index value in the MSB of the first n pixels of the corresponding block;
s42: encrypting a key K according to an image e Encrypting non-embeddable pixels and rendering the original non-embeddable pixelsThe LSB of the beta bit and the MSB of the 1 bit are recorded as additional information to the embeddable space with a first priority, and the remaining bits are recorded within the mark image;
s43: and sending the encrypted image to an information hiding person.
4. The method according to claim 3, wherein in step S5, the information hider embeds the encrypted information into the redundant space of the encrypted image to obtain a secret encrypted image, and sends the secret encrypted image to the recipient as follows:
s51: hiding a key K according to information h For original secret information S secret Carrying out encryption to obtain encrypted secret information;
s52: decoding is performed using a fixed parameter pair (α, β) = (4, 3) to obtain an optimal parameter set, and its correspondence (t) opt ) 2 ；
S53: (t) of each block is obtained from MSBs of the (n + 1) th and (n + 2) th pixels opt ) 2 Determining a redundant space of each image block, and embedding encryption information in the redundant space to obtain a secret-containing encrypted image;
s54: the steganographically encrypted image is sent to the recipient.
5. The method according to claim 4, wherein in S6, the method for the receiver to extract the secret information from the received secret-containing encrypted image and recover the original image is as follows:
s61: after a receiver receives the image containing the secret encryption, the image containing the secret encryption is cut into image blocks with W multiplied by W size and non-overlapping;
s62: decoding an image block using a fixed parameter pair (α, β) = (4, 3) to obtain an optimal parameter set, and its corresponding (t) opt ) 2 ；
S63: if the receiver has the information hiding key K h Determining an embeddable information space and an extra information space of the entire image based on the optimal parameter set, retrieving all embeddedData, obtaining an encrypted secret stream by skipping parameter sets and additional information, and hiding a secret key K with the information h Decryption is carried out to obtain the original secret information S secret ；
S64: if the receiver has the image encryption key K e Determining an embeddable information space and an extra information space of the whole image according to the optimal parameter set, recovering the non-hidden pixels according to the extra information, and encrypting a secret key K by using the image e Decrypting the MSB of the first n pixels and the non-embeddable pixels, and reconstructing an original image I by using the mapped prediction error;
s65: if the receiver has the information hiding key K at the same time h And an image encryption key K e Then the secret information S is extracted using both S63 and S64 secret And restoring the original image I.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110368362.6A CN113079274B (en) | 2021-04-06 | 2021-04-06 | Encryption image reversible information hiding method of adaptive parameter binary tree mark |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110368362.6A CN113079274B (en) | 2021-04-06 | 2021-04-06 | Encryption image reversible information hiding method of adaptive parameter binary tree mark |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113079274A CN113079274A (en) | 2021-07-06 |
| CN113079274B true CN113079274B (en) | 2023-03-07 |
Family
ID=76615164
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110368362.6A Active CN113079274B (en) | 2021-04-06 | 2021-04-06 | Encryption image reversible information hiding method of adaptive parameter binary tree mark |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113079274B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115134474A (en) * | 2022-06-30 | 2022-09-30 | 西安理工大学 | Pixel prediction-based parameter binary tree reversible data hiding method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110337000A (en) * | 2019-07-11 | 2019-10-15 | 南京信息工程大学 | Encrypted domain reversible information hidden method based on double binary tree extension and public key encryption |
| CN111311474A (en) * | 2020-02-04 | 2020-06-19 | 太原科技大学 | Large-capacity image encryption domain information hiding method |
| CN112311954A (en) * | 2020-10-29 | 2021-02-02 | 杭州电子科技大学 | Ciphertext domain reversible information hiding method based on complementary code mark and bitmap embedding |
-
2021
- 2021-04-06 CN CN202110368362.6A patent/CN113079274B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110337000A (en) * | 2019-07-11 | 2019-10-15 | 南京信息工程大学 | Encrypted domain reversible information hidden method based on double binary tree extension and public key encryption |
| CN111311474A (en) * | 2020-02-04 | 2020-06-19 | 太原科技大学 | Large-capacity image encryption domain information hiding method |
| CN112311954A (en) * | 2020-10-29 | 2021-02-02 | 杭州电子科技大学 | Ciphertext domain reversible information hiding method based on complementary code mark and bitmap embedding |
Non-Patent Citations (2)
| Title |
|---|
| An Improved Reversible Data Hiding in Encrypted Images Using Parametric Binary Tree Labeling;Youqing Wu Et.al;《IEEE TRANSACTIONS ON MULTIMEDIA》;20200831;第1929页倒数第2段-第1933页最后1段和附图1 * |
| 基于二叉树标记与码分复用的密文域图像;刘蒙蒙等;《科学技术与工程》;20200909;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113079274A (en) | 2021-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Multi-MSB compression based reversible data hiding scheme in encrypted images | |
| Qiu et al. | Reversible data hiding in encrypted images using adaptive reversible integer transformation | |
| Puteaux et al. | An efficient MSB prediction-based method for high-capacity reversible data hiding in encrypted images | |
| He et al. | A novel high-capacity reversible data hiding scheme for encrypted JPEG bitstreams | |
| Chen et al. | Compression-unimpaired batch-image encryption combining vector quantization and index compression | |
| Chang et al. | A novel VQ-based reversible data hiding scheme by using hybrid encoding strategies | |
| Kumar et al. | Reversible data hiding: A contemporary survey of state-of-the-art, opportunities and challenges | |
| CN107578452B (en) | JPEG image encryption method with compatible format and invariable size | |
| Wu et al. | High-capacity reversible data hiding in encrypted images by bit plane partition and MSB prediction | |
| Xu et al. | A high-capacity reversible data hiding scheme for encrypted images employing vector quantization prediction | |
| Fu et al. | Adaptive embedding combining LBE and IBBE for high-capacity reversible data hiding in encrypted images | |
| CN113382128B (en) | Bit plane compressed encrypted image reversible information hiding method | |
| CN113079274B (en) | Encryption image reversible information hiding method of adaptive parameter binary tree mark | |
| Panchikkil et al. | A pseudo-random pixel mapping with weighted mesh graph approach for reversible data hiding in encrypted image | |
| Shie et al. | Data hiding based on compressed VQ indices of images | |
| Deshmukh et al. | High capacity reversible data hiding in encrypted images using multi-MSB data hiding mechanism with elliptic curve cryptography | |
| Panchikkil et al. | A prediction error based reversible data hiding scheme in encrypted image using block marking and cover image pre-processing | |
| Wang et al. | High capacity reversible data hiding in encrypted images based on prediction error and block classification | |
| Cao et al. | Low-complexity reversible data hiding in encrypted image via MSB hierarchical coding and LSB compression | |
| Wu et al. | Reversible data hiding in encrypted images based on Bit-plane compression of prediction error | |
| Wang et al. | [Retracted] High‐Capacity Reversible Data Hiding in Encrypted Images Based on Prediction Error Compression and Block Selection | |
| Teotia et al. | Enhancing audio and video steganography technique using hybrid algorithm | |
| Chen et al. | Reversible data hiding in encrypted images based on reversible integer transformation and quadtree-based partition | |
| CN115379066B (en) | Encryption image reversible data encryption and decryption method based on self-adaptive compression coding | |
| Yu et al. | Reversible data hiding with pixel prediction and additive homomorphism for encrypted image |
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 |