CN111104872A - GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing - Google Patents
GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing Download PDFInfo
- Publication number
- CN111104872A CN111104872A CN201911210657.XA CN201911210657A CN111104872A CN 111104872 A CN111104872 A CN 111104872A CN 201911210657 A CN201911210657 A CN 201911210657A CN 111104872 A CN111104872 A CN 111104872A
- Authority
- CN
- China
- Prior art keywords
- image
- hash sequence
- svd
- perceptual
- integrity
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2133—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on naturality criteria, e.g. with non-negative factorisation or negative correlation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
Abstract
The invention discloses a GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing, which comprises the following steps: generating an image perception hash sequence and authenticating the integrity of image content, wherein the generation of the image hash sequence comprises the following steps: after the image is subjected to blocking processing, the perceptual hash sequences of the sub-blocks are constructed by using an SIFT operator and an SVD method, and the perceptual hash sequences of all the sub-blocks are further connected in series to generate the perceptual hash sequence of the whole image. In the image integrity authentication process, the detection of the image integrity is realized by measuring the difference degree between the Euclidean distance between an original image and a perception hash sequence of an image to be authenticated and an initial threshold value, if the Euclidean distance between the original image and the perception hash sequence of the image to be authenticated is smaller than the initial threshold value, the image content is judged to be complete, if not, the image is detected to be tampered, and then the positioning of a local tampered area is completed by measuring the difference of the perception hash between sub-blocks. The method has good robustness and is a GF-2 image content integrity authentication method with high practicability.
Description
Technical Field
The invention belongs to the technical field of cartography and geographic information systems, and relates to a GF-2 image integrity authentication method using SIFT and SVD perceptual hashing.
Background
The high-resolution binary (GF-2) satellite transmitted in 2014 in China is a domestic high-resolution image, the spatial resolution of the image is higher than 1m, and the appearance of the image marks that the domestic remote sensing image breaks through the sub-meter level for the first time and reaches the world advanced level. At present, GF-2 images are widely applied in the fields of national defense and military construction, modern construction, scientific research and the like. However, GF-2 images are often subject to various types of attacks and malicious tampering during transmission, use and storage, most commonly, such as modification or deletion of features. To ensure the integrity of the video content is important in practical applications, it is necessary to check the integrity of the video content before using the data to ensure the reliability of the data. However, the high resolution image contains a large amount of information, and the integrity of the image content cannot be detected and judged manually. Therefore, the high-efficiency and reliable inspection of the content integrity of the image is the guarantee and the foundation for ensuring the practical application value of the GF-2 image.
The traditional data authentication method mainly comprises an authentication technology based on traditional cryptography and a digital watermarking technology. The data authentication technology based on the traditional cryptography is quite sensitive to data transformation, and an avalanche effect exists in the authentication process, so that the data authentication technology has a good utilization value in the field of text data authentication, but is not suitable for image data. The digital watermarking technology embeds fragile watermarking information into original data in an authentication process, and realizes the authentication of data content by judging the integrity of the extracted watermarking information. In this way, the watermark information is embedded in the original data, so the precision of the image to the data itself is improved to a certain extent, and the GF-2 image has a higher requirement on the precision of the image in the actual use process due to the improvement of the spatial resolution, so the digital watermarking technology has an influence on the actual use value of the image.
The perceptual hash function has been widely applied in the field of multimedia data as an effective method for data content authentication. The perceptual hash function is based on the information processing theory of cognitive psychology, and is a unidirectional mapping from a multimedia data set to a multimedia perceptual abstract set, so that multimedia digital representations with the same perceptual content are uniquely mapped into a section of digital abstract, and the perceptual robustness requirement is met. Therefore, the data authentication technology based on the perceptual hash function has better practical significance in the data authentication process. However, the existing data authentication technology based on the perceptual hash function is mostly used for multimedia data or common medium-low resolution remote sensing images and multispectral remote sensing images, and the method can not be directly used for authentication of GF-2 influencing content integrity.
The SIFT operator is used as a stable point feature extraction method, has scale, rotation, simulation, affine, visual angle and illumination invariance, has good robustness on operations such as movement, shielding and noise, and is more stable in the aspect of feature extraction compared with other methods.
SVD is a common numerical analysis tool. Singular values obtained through SVD satisfy the characteristics of nonnegativity, uniqueness, stability to disturbance, invariance to matrix transposition and the like.
The invention provides a GF-2 image integrity authentication method using SIFT and SVD perceptual hashing by combining the advantages of perceptual hashing technology, SIFT operators and SVD, and realizes the authentication of GF-2 image content integrity.
Disclosure of Invention
In view of the above, the present invention, aiming at the defects existing in the authentication process of the existing data authentication technology, maps the original GF-2 image into a unique digital digest by means of the perceptual hash function, and compresses and quantizes the digest; in the data authentication process, a perceptual hash sequence of an original image is compared with a perceptual hash sequence of an image to be authenticated, the Euclidean distance between the original perceptual hash sequence which should be perceived first and the perceptual hash sequence of the image to be authenticated is calculated by a method of calculating the Euclidean distance, the difference between the original perceptual hash sequence and the perceptual hash sequence of the image to be authenticated is measured according to the Euclidean distance, if the Euclidean distance is smaller than a given threshold value, the image content is judged to be complete, if not, the image content is detected to be tampered, and the difference of the perceptual hash sequences between corresponding sub-blocks is further compared to finish the accurate positioning of a tampered area in the image. The method can effectively realize the detection and the positioning of local ground object tampering in the GF-2 image, can resist the content retention operations such as DAT format conversion, BMP format conversion, LSB watermark embedding and the like, and is a practical GF-2 image content integrity authentication method with high robustness.
In order to achieve the purpose, the invention adopts the following technical scheme:
a GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing comprises the following steps: generating an image hash sequence and authenticating the integrity of the image;
the image hash sequence generation steps are as follows:
s1: dividing an original image into N sub-blocks with fixed sizes by adopting a fixed partitioning mode;
s2: SIFT feature extraction is carried out on each sub-block respectively, and feature description of corresponding feature points is generated;
s3: compressing the extracted feature description by adopting SVD, extracting the first 10% of large singular values after SVD decomposition, and further generating a perceptual hash sequence of each subblock;
s4: connecting the perception hash sequences of all the sub-blocks in series to construct a final perception hash sequence of the whole image;
the image integrity authentication step is as follows:
s5: extracting each subblock perception hash sequence of the image to be authenticated and a final perception hash sequence of the image;
s6: calculating Euclidean distances between the image to be authenticated and each Hash sequence of the original image;
s7: judging the relation between the obtained Euclidean distance and a given threshold value to obtain a conclusion;
s8: end up
The method is advanced and scientific, ensures effective extraction of fingerprint information, has good robustness, and can ensure that data content can be well identified and authenticated after multiple data are subjected to various transformation operations. Experiments show that the method has better robustness and use value for content maintenance operations such as DAT format conversion, BMP format conversion, LSB watermark embedding and the like.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart of a GF-2 image integrity authentication method using SIFT and SVD perceptual hashing according to the present invention;
FIG. 2 shows image data after GF-2 image data and blocking processing for experiments according to the present invention;
fig. 3 is image final authentication result data provided by the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1, an embodiment of the present invention discloses a GF-2 image integrity authentication method using SIFT and SVD perceptual hashing, including:
1. generation of image hash sequence
And Step 1, dividing the original image into N sub-blocks with fixed sizes by adopting a fixed partitioning mode.
The size of N is determined by the specific image size, the feature extraction efficiency, the robustness, and other factors. In order to ensure reasonable blocking, the invention carries out histogram statistics on the blocked result and calculates the Shannon entropy of the statistical result. The result shows that in order to ensure effective feature extraction and accurate tampering positioning, the shannon entropy of the histogram statistical result of each sub-block should be between 6.50 and 7.00 after the image is partitioned;
and Step 2, extracting and extracting SIFT feature points of each block and feature description information corresponding to the SIFT feature points respectively. For each block, an n × 128 dimensional SIFT feature description is generated, where n is the number of extracted key points in the image. Integrating the n × 128 dimensional characterizer descriptors into an n × 128 matrix, where n is the number of key points extracted from the sub-blocks, and using the matrix as the characteristic matrix M of each sub-blocki(i＝1，2，3...，N)；
And Step 3, compressing the image key point feature matrix extracted by SIFT by adopting SVD (singular value decomposition) to ensure that the image feature information has good robustness on content retention operation while keeping the sensitivity on local tampering. In general, the sum of the singular values of the first 10% of the matrix after SVD occupies more than 99% of the sum of all the singular values. Thus, the feature matrix M of each sub-blockiAfter SVD decomposition, the first 10% of singular value is selected as the characteristic data of the subblock, and a perceptual hash sequence h of the subblock is further formedi(i＝1，2，3...，N)；
Step 4, concatenating all the sub-blocks hi to generate the final perceptual hash sequence H of the whole image, i.e. H ═ H1，h2，h3，...，hN]；
2. Authentication of video content integrity
Step 5, generating a block perceptual hash sequence H 'of the image to be authenticated and a final perceptual hash sequence H' of the image by using an image hash sequence generation method;
step 6, calculating the Euclidean distance between the original image and the perception hash sequence of the image to be authenticated, wherein the calculation method comprises the following steps:
And Step 7, determining an authentication result, and finishing the authentication of the data by judging the size relationship between the calculation result of the Euclidean distance and a given threshold value. Firstly, judging the relation between the Euclidean distance between an original image and a final perception Hash sequence of an image to be authenticated and a threshold value, if the calculation result is less than or equal to the given threshold value, the content of the image to be authenticated is complete and is not tampered, and if not, the image is detected to be tampered. And further respectively judging the relationship between the Euclidean distance of the perceptual hash sequence of each sub-block in the original image and the to-be-authenticated image and a threshold value, and when the calculation result is greater than the threshold value, positioning the current tampered area.
In conclusion, the method effectively solves the problem of content integrity authentication of the GF-2 image, has better robustness on content maintaining operations such as DAT format conversion, BMP format conversion, LSB watermark embedding and the like of the GF-2 image, extracts the perceptual hash sequence of the image by applying the perceptual hash methods of SIFT and SVD, has high robustness on the same data of the content, can be used for local ground object tampering identification and positioning of the GF-2 image, and has better practical value.
Claims (3)
1. A GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing comprises the following steps: generating an image hash sequence and authenticating the integrity of the image;
the image hash sequence generation steps are as follows:
s1: dividing an original image into N sub-blocks with fixed sizes by adopting a fixed partitioning mode;
s2: SIFT feature extraction is carried out on each sub-block respectively, and feature description of corresponding feature points is generated;
s3: compressing the extracted feature description by adopting SVD, extracting the first 10% of large singular values after SVD decomposition, and further generating a perceptual hash sequence of each subblock;
s4: connecting the perception hash sequences of all the sub-blocks in series to construct a final perception hash sequence of the whole image;
the image content integrity authentication step is as follows:
s5: extracting each subblock perception hash sequence of the image to be authenticated and a final perception hash sequence of the image;
s6: calculating Euclidean distances between the image to be authenticated and each Hash sequence of the original image;
s7: judging the relation between the obtained Euclidean distance and a given threshold value to obtain a conclusion;
s8: and (6) ending.
2. The method of claim 1, wherein in step S5, the video hash generation process from steps S1 to S4 is applied.
3. The GF-2 image integrity certification method using SIFT and SVD perceptual hashing as claimed in claim 1 or claim 2, wherein in step S7, when the image content integrity certification is completed, if the image is subjected to local malicious tampering, the tampered region can be precisely located.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911210657.XA CN111104872A (en) | 2019-11-29 | 2019-11-29 | GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911210657.XA CN111104872A (en) | 2019-11-29 | 2019-11-29 | GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111104872A true CN111104872A (en) | 2020-05-05 |
Family
ID=70421042
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911210657.XA Pending CN111104872A (en) | 2019-11-29 | 2019-11-29 | GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111104872A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111861847A (en) * | 2020-07-12 | 2020-10-30 | 兰州交通大学 | GF-2 image double watermarking method applying DWT and SIFT |
| CN112232428A (en) * | 2020-10-23 | 2021-01-15 | 上海电力大学 | Image hash acquisition method based on three-dimensional characteristics and energy change characteristics |
| CN112561769A (en) * | 2020-12-03 | 2021-03-26 | 兰州交通大学 | GF-2 image security protection method using exchange cipher watermark |
| CN113592744A (en) * | 2021-08-12 | 2021-11-02 | 长光卫星技术有限公司 | Geometric precise correction method suitable for high-resolution remote sensing image |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103747255A (en) * | 2014-01-27 | 2014-04-23 | 深圳大学 | Video tamper detection method and device based on airspace perceptual hashing |
| CN104091303A (en) * | 2014-07-11 | 2014-10-08 | 湖南大学 | Robust image hashing method and device based on Radon transformation and invariant features |
| CN107423768A (en) * | 2017-08-02 | 2017-12-01 | 上海应用技术大学 | The image Hash sequence generating method combined based on SURF and PCA |
| CN110175642A (en) * | 2019-05-22 | 2019-08-27 | 南京农业大学 | A kind of chrysanthemum similarity calculation method based on PCA dimensionality reduction and feature binary |
-
2019
- 2019-11-29 CN CN201911210657.XA patent/CN111104872A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103747255A (en) * | 2014-01-27 | 2014-04-23 | 深圳大学 | Video tamper detection method and device based on airspace perceptual hashing |
| CN104091303A (en) * | 2014-07-11 | 2014-10-08 | 湖南大学 | Robust image hashing method and device based on Radon transformation and invariant features |
| CN107423768A (en) * | 2017-08-02 | 2017-12-01 | 上海应用技术大学 | The image Hash sequence generating method combined based on SURF and PCA |
| CN110175642A (en) * | 2019-05-22 | 2019-08-27 | 南京农业大学 | A kind of chrysanthemum similarity calculation method based on PCA dimensionality reduction and feature binary |
Non-Patent Citations (1)
| Title |
|---|
| 刘霞: "《基于尺度不变与视觉显著特征的图像感知哈希技术研究》", 《中国优秀硕士学位论文全文数据库 (信息科技辑)》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111861847A (en) * | 2020-07-12 | 2020-10-30 | 兰州交通大学 | GF-2 image double watermarking method applying DWT and SIFT |
| CN112232428A (en) * | 2020-10-23 | 2021-01-15 | 上海电力大学 | Image hash acquisition method based on three-dimensional characteristics and energy change characteristics |
| CN112232428B (en) * | 2020-10-23 | 2021-11-16 | 上海电力大学 | Image hash acquisition method based on three-dimensional characteristics and energy change characteristics |
| CN112561769A (en) * | 2020-12-03 | 2021-03-26 | 兰州交通大学 | GF-2 image security protection method using exchange cipher watermark |
| CN113592744A (en) * | 2021-08-12 | 2021-11-02 | 长光卫星技术有限公司 | Geometric precise correction method suitable for high-resolution remote sensing image |
| CN113592744B (en) * | 2021-08-12 | 2024-03-19 | 长光卫星技术股份有限公司 | Geometric fine correction method suitable for high-resolution remote sensing image |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111104872A (en) | GF-2 image integrity authentication method applying SIFT and SVD perceptual hashing | |
| US8483427B2 (en) | System and method for image authentication | |
| Lu et al. | Multimedia forensic hash based on visual words | |
| CN106503721B (en) | Hash algorithm and authentication method based on cmos image sensor PUF | |
| CN102096894B (en) | Image fragile watermarking algorithm capable of realizing accurate positioning of tampered region | |
| CN101489133B (en) | Geometric attack resisting real-time video watermarking method | |
| Wang et al. | Reversible fragile watermarking for locating tampered blocks in 2D vector maps | |
| CN111968027B (en) | Robust color image zero watermarking method based on SURF and DCT features | |
| CN116385250B (en) | Track data double watermarking method based on robust watermarking and fragile watermarking | |
| CN103294676A (en) | Content duplicate detection method of network image based on GIST (generalized search tree) global feature and SIFT (scale-invariant feature transform) local feature | |
| Xijin et al. | The application research of MD5 encryption algorithm in DCT digital watermarking | |
| Qi et al. | Robust authentication for paper-based text documents based on text watermarking technology | |
| CN116757909B (en) | BIM data robust watermarking method, device and medium | |
| CN102881008B (en) | Based on the anti-rotation image Hash method of annulus statistical nature | |
| Wang et al. | RST invariant fragile watermarking for 2D vector map authentication | |
| CN104637484A (en) | MP3 audio steganography detection method based on co-occurrence matrix analysis | |
| CN110349072B (en) | Watermark synchronization method in vector geographic data watermark embedding and detecting process | |
| Shang et al. | Double JPEG detection using high order statistic features | |
| CN106952211B (en) | Compact image hashing method based on feature point projection | |
| Singh et al. | Analysis of contrast enhancement forensics in compressed and uncompressed images | |
| CN202929680U (en) | Digital watermarking device based on image feature and Huffman coding theory | |
| Nasiri et al. | Exposing forgeries in soccer images using geometric clues | |
| CN111339238A (en) | Vector geographic data digital watermarking method capable of resisting projective transformation | |
| US11983789B1 (en) | Generation method, detection method, generation device, and detection device of zero watermarking for trajectory data, and storage medium | |
| Zhao et al. | Passive detection of paint-doctored JPEG images |
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 |