CN112541854B - Selective encryption method based on target detection - Google Patents
Selective encryption method based on target detection Download PDFInfo
- Publication number
- CN112541854B CN112541854B CN202011432547.0A CN202011432547A CN112541854B CN 112541854 B CN112541854 B CN 112541854B CN 202011432547 A CN202011432547 A CN 202011432547A CN 112541854 B CN112541854 B CN 112541854B
- Authority
- CN
- China
- Prior art keywords
- bits
- channel
- image
- network
- frames
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 title claims abstract description 21
- 230000000739 chaotic effect Effects 0.000 claims abstract description 11
- 230000001174 ascending effect Effects 0.000 claims description 4
- 230000010339 dilation Effects 0.000 claims description 4
- 230000001629 suppression Effects 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 2
- 238000000926 separation method Methods 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims 1
- 238000013507 mapping Methods 0.000 abstract description 9
- 230000015556 catabolic process Effects 0.000 abstract description 3
- 238000006731 degradation reaction Methods 0.000 abstract description 3
- 230000008034 disappearance Effects 0.000 abstract description 2
- 238000004880 explosion Methods 0.000 abstract description 2
- 238000012545 processing Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 238000009792 diffusion process Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 1
- 241000282326 Felis catus Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001680 brushing effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/0021—Image watermarking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/51—Indexing; Data structures therefor; Storage structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/5838—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/60—Protecting data
- G06F21/602—Providing cryptographic facilities or services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2201/00—General purpose image data processing
- G06T2201/005—Image watermarking
- G06T2201/0065—Extraction of an embedded watermark; Reliable detection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Library & Information Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Bioethics (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Computer Security & Cryptography (AREA)
- Image Analysis (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention belongs to the field of image processing and discloses a selective encryption method based on target detection. In order to increase the key space of an encryption system, two rounds of logistic mapping are used, and the absolute value of the sequence generated by two rounds of chaotic mapping is taken by using the difference of the sequence, so that a new chaotic sequence is obtained. The image is encrypted using the new chaotic sequence. Meanwhile, the problem of network degradation is solved by utilizing a residual network, and the original nonlinear mapping is approximately converted into an identity mapping by connecting a shortcut, so that the problems of gradient disappearance or gradient explosion and the like which occur along with the increase of network layers are avoided.
Description
Technical Field
The invention belongs to the field of image processing, and relates to a selective encryption method based on target detection.
Background
Today, when information explodes greatly, the science brings convenience, personal information also faces the risk of leakage, many people like to take a photo along with him, and the people's own sights are released to the internet, at this time, the information of other people may be unintentionally exposed, sometimes the author does not realize the problem, and in this case, it is necessary to hide some information.
Face detection is a relatively important branch in the field of artificial intelligence, is also a basic step for face research, is widely applied to the development of science and technology today, represents the arrival of the age of face brushing, and has a plurality of places related to face detection in life, such as identity verification of an access control system, and a face-beautifying camera and the like, all have figures of the face detection. The face detection technology mainly extracts the characteristics of the picture through a network and then classifies the characteristics to realize the positioning of the face. More common detection networks are Faster R-CNN, mask R-CNN, YOLO, SSD, etc.
Image encryption is a method for hiding information, and the main operation is scrambling and diffusing the image. The scrambling is to change the original pixel position of the image to destroy the regularity of the original space and the correlation among local pixels, so that the image is disordered and can not be identified, thereby achieving the purpose of hiding information. Common permutation methods are cat mapping, magic square transformation, s-box based scrambling, etc. The diffusion is to change the pixel value of the original image to make the image 'look new', so that the original information is covered, and the image can be diffused by using a chaotic system and a transformation domain. The invention provides a selective encryption algorithm which comprehensively considers a target detection technology and an image encryption technology.
Many image encryption algorithms employ a single level of permutation and diffusion operations, have a small key space and poor resistance to brute force attacks, while a good encryption algorithm should require a large enough key space to resist exhaustive attacks. Meanwhile, when a detection network such as fast R-CNN, mask R-CNN, YOLO, SSD is used to detect a target, a problem of network degradation occurs with an increase in network depth, that is, network accuracy is saturated or even degraded when the network depth increases.
Disclosure of Invention
The invention provides a selective encryption method based on target detection aiming at the problem of information leakage in the modern society and the problems of key space and encryption performance of an encryption algorithm. In order to increase the key space of an encryption system, two rounds of logistic mapping are used, and the absolute value of the sequence generated by two rounds of chaotic mapping is taken by using the difference of the sequence, so that a new chaotic sequence is obtained. The image is encrypted using the new chaotic sequence. Meanwhile, the problem of network degradation is solved by utilizing a residual network, and the original nonlinear mapping is approximately converted into an identity mapping by connecting a shortcut, so that the problems of gradient disappearance or gradient explosion and the like which occur along with the increase of network layers are avoided. The structure of the residual block is shown in fig. 1.
The invention uses ResNet152 as a backbone network to extract the characteristics of the image, and simultaneously carries out one-time cavity convolution and up-sampling operation on the extracted characteristics, and finally carries out the selection of a prediction frame by utilizing non-maximum suppression.
In the encryption process, bits of different channels are interchanged, the high four bits of the R channel and the low four bits of the B channel are combined, the high four bits of the G channel and the low four bits of the R channel are combined, the high four bits of the B channel and the low four bits of the G channel are combined, and the encryption effect is effectively enhanced.
The specific technical scheme of the invention is as follows:
(1) Reading a three-channel color image, transmitting the image to a detection network shown in fig. 2, extracting features through the detection network shown in fig. 2, and performing convolution operation on the extracted feature matrix to obtain the offset of the prior frame;
(2) Regression and classification are carried out on the prior frames by utilizing the prior frames preset by the network and the offset of the prior frames obtained in the step (1) to obtain the predicted frames, non-maximum suppression is carried out on the predicted frames to screen out repeated predicted frames, the repeated predicted frames are deleted, and finally the reserved predicted frames are the predicted frames containing the human faces (the non-maximum suppression is to remove redundant human face detection frames in fact, and the most correct human face detection frames are reserved);
(3) Obtaining the left upper corner coordinate and the right lower corner coordinate of a predicted frame containing a human face through the step (2), intercepting an image in the predicted frame, and performing separation operation on image channels to obtain three channels respectively: r, G, B dividing the three channels into bit planes, wherein each channel is composed of eight bits and is divided into a high four bits and a low four bits;
(4) The bits of different channels are interchanged, the upper four bits of the R channel and the lower four bits of the B channel are combined to form eight bits of R1, the upper four bits of the G channel and the lower four bits of the R channel are combined to form eight bits of G1, and the upper four bits of the B channel and the lower four bits of the G channel are combined to form eight bits of B1.
(5) Converting the 24 bit planes into R, G and B three-channel planes; generating chaotic sequences K1, K2, K= ((fabs (K1-K2)) 10 by using initial values (x 0, u 0) and (x 1, u 1) and chaotic system logistic 14 ) mod 256), where fabs represent absolute values, mod is a modulo operation;
(6) Ascending order of the chaos sequence K is arranged, and ascending order index is obtained;
(7) Scrambling the pixel positions of the image with index values, (e.g., index [0] =5, then the first pixel value changes to the fifth position)
(8) Diffusing the pixel value with c (i) =p (i)/(k (i-1)); wherein ∈ represents an exclusive or operation; p (i) represents a pixel value scrambled with index; k (i-1) and K (i) obtained by step (5), K (1), K (2), K (3), -K (i), -kn belongs to the element in K, n is equal to the width of the image; c (i-1) is the previous term for c (i), wherein c (1) =p (1);
(9) Repeating operations (7) - (8);
(10) Obtaining a final image.
Further, as shown in fig. 2, the detection network framework in the step (1) has six original feature layers (of_1 to of_6) forming a first layer of the network and six enhanced feature layers (ef_1 to ef_6) forming a second layer of the network, wherein the enhanced feature layers are obtained by the original feature layers through the enhancement module, and the relationship between the enhanced features and the original features is as follows:
ec(i,j,l)＝f concat (f dilation (nc(i,j,l))
nc(i,j,l)＝f prod (oc(i,j,l),f up (oc(i,j,l+1))
where ec (i, j, l) represents the (i, j) position of the first layer of the enhancement feature layer and nc (i, j, l) represents the (i, j) position of the first layer of the original feature layer. f (f) prod Representing dot product operations between corresponding elements between matrices, representing upsampling, f dilation Representing the convolution of the holes, f concat Representing the connection operation.
The method has the beneficial effects that the method is provided with a selective encryption method based on target detection, in order to improve the encryption effect, a bit replacement strategy among channels is introduced, and experiments prove that the method has good effect in the aspect of information hiding.
Drawings
Figure 1 is a block diagram of a residual network.
Fig. 2 target detection network.
Fig. 3 selective encryption based on object detection.
Fig. 4 selective decryption based on object detection.
FIG. 5 is a bar graph; (a) R channel, (B) G channel, and (c) B channel.
Fig. 6 illustrates a diagram.
Detailed Description
Taking fig. 6 as an example, a detailed operation of the one-pass algorithm is performed. Initial values (x 0, u 0), (x 1, u 1) are (3.7,0.6) and (3.9,0.8), respectively.
1. Taking correlation coefficient as an example to analyze the algorithm
TABLE 1 correlation coefficient of adjacent pixels of scrambled image
| R | G | B | |
| Horizontal correlation | -0.000229 | 0.001069 | -0.000584 |
| Vertical correlation | -0.000861 | 0.000838 | 0.002111 |
| Diagonal correlation | 0.000375 | 0.000058 | -0.0001992 |
From the correlation coefficient, the three channels are very close to 0 in terms of horizontal correlation, vertical correlation and diagonal correlation, the relation between the connected pixels is very small, and the logicalness between the pixels is destroyed, so that the proven algorithm effect is better.
2. Taking a histogram as an example to analyze the algorithm
From the histogram of fig. 4, the pixel values of the three channels are all relatively uniform, with no significant peaks.
3 taking information entropy as an example to analyze the algorithm
Table 2 scrambling diagram information entropy analysis
| R channel | G channel | B channel | |
| Information entropy | 7.999788450622844 | 7.999818402206316 | 7.99979199338572 |
From the information entropy, the information entropy of the three channels is close to 8, reflecting good encryption effect
The decryption process is the reverse process of the decryption process, in the decryption process, the encrypted image is firstly subjected to the diffusion reverse operation and then subjected to the replacement reverse operation, the steps are repeated to obtain the image subjected to bit replacement among channels, and then the reverse operation of bit replacement in the encryption process is performed to obtain the original image. The flow chart of the encryption process is shown in fig. 2, and the flow chart of the decryption process is shown in fig. 3.
Claims (2)
1. The selective encryption method based on target detection is characterized by comprising the following steps:
(1) Reading a three-channel color image, transmitting the image to a detection network, and extracting features through the detection network; performing convolution operation on the extracted feature matrix to obtain the offset of the prior frame;
(2) Regression and classification are carried out on the prior frames by utilizing the prior frames preset by the network and the offset of the prior frames obtained in the step (1) to obtain the predicted frames, non-maximum suppression screening is carried out on the predicted frames to select repeated predicted frames, the repeated predicted frames are deleted, and finally the reserved predicted frames are the predicted frames containing the human faces;
(3) Obtaining the left upper corner coordinate and the right lower corner coordinate of a predicted frame containing a human face through the step (2), intercepting an image in the predicted frame, and performing separation operation on image channels to obtain three channels respectively: r, G, B dividing the three channels into bit planes, wherein each channel is composed of eight bits and is divided into a high four bits and a low four bits;
(4) The bits of different channels are interchanged, the upper four bits of the R channel and the lower four bits of the B channel are combined to form eight bits of R1, the upper four bits of the G channel and the lower four bits of the R channel are combined to form eight bits of G1, and the upper four bits of the B channel and the lower four bits of the G channel are combined to form eight bits of B1;
(5) Converting the 24 bit planes into R, G and B three-channel planes; generating chaotic sequences K1, K2, K= ((fabs (K1-K2)) 10 by using initial values (x 0, u 0) and (x 1, u 1) and chaotic system logistic 14 ) mod 256), where fabs represent absolute values, mod is a modulo operation;
(6) Ascending order of the chaos sequence K is arranged, and ascending order index is obtained;
(7) Scrambling pixel locations of the image using the index values;
(8) Diffusing the pixel value with c (i) =p (i)/(k (i-1)); wherein ∈ represents an exclusive or operation; p (i) represents a pixel value scrambled with index; k (i-1) and K (i) obtained by step (5), K (1), K (2), K (3), -K (i), -kn belongs to the element in K, n is equal to the width of the image; c (i-1) is the previous term for c (i), wherein c (1) =p (1);
(9) Repeating operations (7) - (8);
(10) Obtaining a final image.
2. The selective encryption method based on object detection according to claim 1, wherein the network frame for detection in step (1) has a first layer with six original feature layers forming a network and a second layer with six enhanced feature layers forming a network, wherein the enhanced feature layers are obtained by the original feature layers through an enhancement module, and the relationship between the enhanced features and the original features is as follows:
ec(i,j,l)＝f concat (f dilation (nc(i,j,l))
nc(i,j,l)＝f prod (oc(i,j,l),f up (oc(i,j,l+1))
where ec (i, j, l) represents the (i, j) position of the first layer of the enhancement feature layer, nc (i, j, l) represents the (i, j) position of the first layer of the original feature layer, f prod Representing dot product operations between corresponding elements between matrices, representing upsampling, f dilation Representing the convolution of the holes, f concat Representing the connection operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011432547.0A CN112541854B (en) | 2020-12-10 | 2020-12-10 | Selective encryption method based on target detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011432547.0A CN112541854B (en) | 2020-12-10 | 2020-12-10 | Selective encryption method based on target detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112541854A CN112541854A (en) | 2021-03-23 |
| CN112541854B true CN112541854B (en) | 2023-09-29 |
Family
ID=75019736
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011432547.0A Active CN112541854B (en) | 2020-12-10 | 2020-12-10 | Selective encryption method based on target detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112541854B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6064738A (en) * | 1996-12-10 | 2000-05-16 | The Research Foundation Of State University Of New York | Method for encrypting and decrypting data using chaotic maps |
| CN101610408A (en) * | 2008-06-16 | 2009-12-23 | 北京智安邦科技有限公司 | Video protection disorder method and structure |
| CN106296560A (en) * | 2016-07-29 | 2017-01-04 | 广东工业大学 | A kind of multicolor image encryption algorithm of complex cascade chaos |
| CN106910156A (en) * | 2017-02-27 | 2017-06-30 | 东北大学 | A kind of New chaotic image encryption method with plaintext association key stream generting machanism |
| CN106952213A (en) * | 2017-03-16 | 2017-07-14 | 郑州轻工业学院 | New chaotic image encryption method based on bit permutation with dynamic DNA encoding |
| CN110889877A (en) * | 2019-11-26 | 2020-03-17 | 齐齐哈尔大学 | Image encryption method based on L-F cascade chaos and 3D bit scrambling |
-
2020
- 2020-12-10 CN CN202011432547.0A patent/CN112541854B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6064738A (en) * | 1996-12-10 | 2000-05-16 | The Research Foundation Of State University Of New York | Method for encrypting and decrypting data using chaotic maps |
| CN101610408A (en) * | 2008-06-16 | 2009-12-23 | 北京智安邦科技有限公司 | Video protection disorder method and structure |
| CN106296560A (en) * | 2016-07-29 | 2017-01-04 | 广东工业大学 | A kind of multicolor image encryption algorithm of complex cascade chaos |
| CN106910156A (en) * | 2017-02-27 | 2017-06-30 | 东北大学 | A kind of New chaotic image encryption method with plaintext association key stream generting machanism |
| CN106952213A (en) * | 2017-03-16 | 2017-07-14 | 郑州轻工业学院 | New chaotic image encryption method based on bit permutation with dynamic DNA encoding |
| CN110889877A (en) * | 2019-11-26 | 2020-03-17 | 齐齐哈尔大学 | Image encryption method based on L-F cascade chaos and 3D bit scrambling |
Non-Patent Citations (3)
| Title |
|---|
| Liu Bo等."Research of image encryption algorithm base on chaos theory".《Proceedings of 2011 6th International Forum on Strategic Technology》.2011,1-2. * |
| 吴杰."基于混沌的置乱—置换图像加密算法研究".《中国优秀硕士学位论文全文数据库 信息科技辑》.2018,(第2期),I138-1576. * |
| 张晓博 等."基于复合混沌序列的图像加密方法".《冶金自动化》.2010,第36卷(第11期),3396-3400. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112541854A (en) | 2021-03-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kamal et al. | A new image encryption algorithm for grey and color medical images | |
| CN111105339B (en) | Image encryption method based on multidimensional chaotic system and Joseph scrambling | |
| Zhu et al. | A chaos-based symmetric image encryption scheme using a bit-level permutation | |
| CN113129196A (en) | Image encryption method based on DNA sequence and memristor chaos | |
| CN112422269A (en) | Combined chaotic pseudo-random number generator and digital image encryption method thereof | |
| Wang et al. | TPE-ISE: approximate thumbnail preserving encryption based on multilevel DWT information self-embedding | |
| CN113992810B (en) | Agile image encryption method based on deep learning | |
| Abdulkadhim et al. | Audio steganography based on least significant bits algorithm with 4D grid multi-wing hyper-chaotic system | |
| CN111177746A (en) | Efficient visual secret sharing method with core participants | |
| Siahaan | High Complexity Bit-Plane Security Enchancement in BPCS Steganography | |
| Ghoul et al. | A review on security techniques in image steganography | |
| CN112541854B (en) | Selective encryption method based on target detection | |
| Thajeel et al. | An Improve Image Encryption Algorithm Based on Multi-level of Chaotic Maps and Lagrange Interpolation. | |
| Cao et al. | Low-complexity reversible data hiding in encrypted image via MSB hierarchical coding and LSB compression | |
| CN110069901B (en) | Space domain multi-image encryption algorithm based on chaotic system | |
| Laffont et al. | Enhanced pixel value modification based on modulus function for RGB image steganography | |
| Bartwal et al. | Lossless and Reversible Data Hiding in Encrypted Images With Public Key Cryptography. | |
| Hamano et al. | Image classification using vision transformer for EtC images | |
| CN115348361B (en) | Encryption image reversible data encryption and decryption method based on multi-MSB block coding | |
| CN109920015B (en) | Image visual meaningful encryption method based on fractal graph generation | |
| CN114760391B (en) | Reversible data hiding method with high embedding rate based on double-layer embedding | |
| CN115037845B (en) | Color image reversible hiding method based on Hachimoji DNA and Julia fractal | |
| Dhannoon | An indirect MSB data hiding technique | |
| CN118101845A (en) | Three-level protection method for image | |
| CN110278067B (en) | Image encryption method based on low-dimensional fractional order chaotic system |
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 |