CN102446211A - Method and system for filing and verifying image - Google Patents
Method and system for filing and verifying image Download PDFInfo
- Publication number
- CN102446211A CN102446211A CN2011102625278A CN201110262527A CN102446211A CN 102446211 A CN102446211 A CN 102446211A CN 2011102625278 A CN2011102625278 A CN 2011102625278A CN 201110262527 A CN201110262527 A CN 201110262527A CN 102446211 A CN102446211 A CN 102446211A
- Authority
- CN
- China
- Prior art keywords
- image
- record
- putting
- verified
- retrieval
- 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
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000001514 detection method Methods 0.000 claims abstract description 23
- 238000012795 verification Methods 0.000 claims description 14
- 239000004744 fabric Substances 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 abstract description 10
- 239000003814 drug Substances 0.000 description 49
- 230000008569 process Effects 0.000 description 18
- 238000010586 diagram Methods 0.000 description 6
- 239000000463 material Substances 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 229940079593 drug Drugs 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 230000006835 compression Effects 0.000 description 3
- 238000007906 compression Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 230000001105 regulatory effect Effects 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 230000001902 propagating effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 101100217298 Mus musculus Aspm gene Proteins 0.000 description 1
- 241001441724 Tetraodontidae Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 230000035876 healing Effects 0.000 description 1
- 235000013402 health food Nutrition 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000010008 shearing Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Landscapes
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for filing and verifying an image, comprising the following steps of: when filing an image, generating unforgeable watermark data according to the filing information, and embedding the watermark data into the filed image, and realizing the filing of the filed image by storing the filed image embedded with a digital watermark and extracting image characteristics from the same, and simultaneously, applying the filed image embedded with the digital watermark in stead of the original filed image; and then, when verifying the image, verifying whether the image is already filed by joining image retrieval and digital watermark detection, and detecting whether the filed image is falsified by comprising the image to be verified which has the watermark confirmed through detection with the corresponding filed image. The invention also discloses a system for filing and verifying an image. The method and the system provided in the invention are capable of realizing the filing of the image by joining the image retrieval and the digital watermark technology in combination with image falsification detection; and the method and the system also are capable of verifying whether an unknown image is already filed and whether the image is a falsified filed image based on the filing data.
Description
Technical field
The present invention relates to the multimedia information management technical field, relate in particular to a kind of image and put on record and verification method and system.
Background technology
Along with the development of computer and network technologies, and the popularizing of ecommerce, the internet becomes the ad distribution channel and the marketing channel of extensive stock just gradually, and scale is increasing.But for some special commodity, such as medicine, there is relevant laws and regulations regulations in country, arbitrarily releasing advertisements and selling.
Special in advertisement; Relevant regulations according to the Advertisement Law of the People's Republic of China, " Pharmaceutical Administration Law of the People's Republic of China ", " Pharmaceutical Administration Law of the People's Republic of China implementing regulations ", " advertisement for pharmaceuticals examination method " and laws and regulations such as " advertisement for pharmaceuticals examination issue standards "; The advertisement of commodity such as medicine, medicine equipment, health food (hereinafter to be referred as: be need transfer to food and medicine at different levels Surveillance Authority to examine advertisement for pharmaceuticals); Pass and put on record after externally issue, and the advertisement for pharmaceuticals that issue is used must be in full accord with the content that examination is put on record.
At present; Comprise mainly that to the flow process of putting on record of advertisement for pharmaceuticals pharmaceutical production merchant or agency submit the content file of various associated materials and advertisement for pharmaceuticals (like advertising image) to; After examining by food and medicine at different levels Surveillance Authority; Relevant information and ad content are preserved, notify enterprise whether to pass through then.Handling of reports against wrongdoings is then mainly adopted in supervision to advertisement for pharmaceuticals on the internet; To suspect that then illegal network advertisement for pharmaceuticals and relevant information thereof inquire about, verify in the storehouse of putting on record; Whether illegal, and handle accordingly according to discrimination result by manual work if mainly differentiating this advertisement.
But, ad distribution channel enormous amount on the internet, and internet drug market coverage is wide, disguised strong, need the advertisement for pharmaceuticals of magnanimity be verified and supervise, in case illegal activities take place, control is difficult, evidence obtaining is difficult, the supervision difficulty is very big.In the face of the magnanimity information on the internet, above-mentioned traditional regulatory measure is difficult to realize the effective supervision to advertisement for pharmaceuticals on the internet, can not reach and hit illegal advertisement for pharmaceuticals, protection consumer's purpose.So, be necessary very much to adopt the infotech means, improve supervisory efficiency to the internet advertisement for pharmaceuticals, ad content is put on record and the method verified then is the improved emphasis of needs.
Image is one of main issue form of the Commdity advertisement on the internet, and advertisement for pharmaceuticals is also issued on the internet and propagated with the form of image mostly.Therefore, be badly in need of putting on record and verification technique of a kind of image, can realize effective checking, specifically can solve following technical matters large nuber of images:
1) can utilize the data of putting on record of advertisement for pharmaceuticals image, not have under the situation of other supplementary, realize quick checking, judge whether it is the advertising image of putting on record unknown advertising image.
Because according to the rules, can only issue and use and the on all four advertisement for pharmaceuticals of ad content of putting on record, whether be the content of putting on record so the unknown advertising image of a width of cloth is judged in checking, be the groundwork of advertisement for pharmaceuticals supervision.And the characteristics decision of internet advertisement for pharmaceuticals supervision except a unknown advertising image to be verified, does not often have other extra side information to utilize.
2) utilizing the data of putting on record of advertisement for pharmaceuticals image, in the proof procedure to new advertising image, can judge whether it is distorting the advertising image of putting on record.
Possibly comprise description in the advertisement for pharmaceuticals image, distort, just possibly make the character of advertisement and content have greatly changed the advertising image of putting on record is very a little bit smaller to action of drug, curative effect and usage.Such as on advertising image, adding terms such as " definitely ", " healing ", perhaps " onset in 30 days " changed into " onset in 3 days ".These are revised concerning image itself, and change changes very little, and still the content to advertisement for pharmaceuticals then is very large change.So need in proof procedure, find this type distorts.
3) utilizing the data of putting on record of advertisement for pharmaceuticals image, in the proof procedure to new advertising image, can tolerate some reasonable editors, like format conversion, size scaling etc. to the advertising image of putting on record.
Often size is bigger than normal for the advertisement for pharmaceuticals image of putting on record, and its image file is also bigger.And in the actual on the internet issue use of these advertisement for pharmaceuticals images,, often need carry out some scalings to its size in order to improve communication effect, perhaps change into the bigger picture format of compressibility.These editors to putting the advertisement for pharmaceuticals image on record are reasonably, in the proof procedure to the advertisement for pharmaceuticals image, if think that these are rationally edited is to distort, and with the rate of false alarm that increases checking, reduces the efficient of supervision.
Summary of the invention
The method and system that the embodiment of the invention provides a kind of image to put on record and verify, these method and system can effectively be put on record to picture material, and realize the efficient verification to large nuber of images.These method and system combine CBIR and digital watermarking, and are aided with the distorted image detection, in order to needing to this type of quasi-drugs advertisement to realize the image of Content Advisor; The file of putting on record; And on this basis, whether the checking unknown images put on record, and at tolerance compression of images, size scaling to a certain degree; And under the condition of format conversion, whether the checking unknown images is distorting the image of putting on record.
Image of the present invention is put on record and is comprised with verification method: step 11, generate digital watermark data based on the information calculations of putting on record of the image of putting on record, and it is embedded in the image of putting on record after preserving this watermark data, obtain the image of putting on record of embed digital watermark; Step 12 is extracted put on record the characteristics of image of image of embed digital watermark, preserves embed digital watermark put on record image and characteristic thereof, and is associated with corresponding information and the watermark data put on record; Step 13 is with the original image of putting on record of image replacement of putting on record of embed digital watermark; Step 14; To image to be verified, extract its characteristics of image and in all put image on record, retrieve, result for retrieval comprises put on record image and related put on record information and the watermark data thereof that satisfies certain condition; If do not retrieve, explain that then image to be verified puts on record; Step 15; Based on result for retrieval; Similarity order from high to low according to put on record image and image to be verified; Detect the corresponding digital watermarking of image of putting on record of each width of cloth in the result for retrieval one by one and whether be present in the image to be verified,, explain that then image to be verified puts on record if in image to be verified, do not detect all corresponding digital watermarkings of image of putting on record in the result for retrieval.
Image of the present invention is put on record and is comprised with verification system: the digital watermarking module, be used for generating digital watermark data based on the information calculations of putting on record of the image of putting on record, and it is embedded in the image of putting on record after preserving this watermark data, obtain the image of putting on record of embed digital watermark; The storage center of putting on record module is used to preserve the image of putting on record of embed digital watermark, and is associated with corresponding information and the watermark data put on record, extract simultaneously embed digital watermark put on record image characteristics of image and set up index; Image retrieval module; Be used for to image to be verified, extract its characteristics of image and in all put image on record, retrieve, result for retrieval comprises put on record image and related put on record information and the watermark data thereof that satisfies certain condition; If do not retrieve, explain that then image to be verified puts on record; Said digital watermarking module is based on result for retrieval; Similarity order from high to low according to put on record image and image to be verified; Detect image to be verified whether contain with result for retrieval in the corresponding digital watermarking of image of putting on record; If do not detect corresponding digital watermarking, explain that then image to be verified puts on record.
Adopt the present invention to realize the large nuber of images that environment such as internet use, propagate is down carried out efficient verification, judge whether it put on record, and whether on the image of putting on record, distort through file that image is put on record.Adopt the present invention can make that to image, particularly the intellectual monitoring of advertisement for pharmaceuticals image becomes possibility, let computer generation replace manual work to carry out the checking of picture material, for the propagation of the magnanimity advertising image under the internet environment provides a kind of effective regulatory measure.
Description of drawings
Fig. 1 carries out the process flow diagram that image is put on record and verified in the embodiment of the invention;
Fig. 2 is for generating the process flow diagram of watermark data in the embodiment of the invention according to the information of putting on record of the image of putting on record;
Fig. 3 is for carrying out the process flow diagram that digital watermarking detects based on result for retrieval in the embodiment of the invention;
Fig. 4 is the synoptic diagram through moving window traversing graph picture;
The process flow diagram that Fig. 5 detects for the distorted image of putting on record in the embodiment of the invention;
Fig. 6 is the structural representation with verification system of putting on record of the image in the embodiment of the invention.
Embodiment
For making the object of the invention, technical scheme and advantage clearer, below in conjunction with specific embodiment, and with reference to accompanying drawing, to further explain of the present invention.
Image of the present invention is put on record and the principle of work of verification method comprises following two aspects: first aspect; Carrying out image when putting on record; After at first generating the digital watermark data that to forge and preserving, this watermark data is embedded among the image of putting on record according to the information of putting on record of the image of putting on record; Through preserving the image of putting on record of embed digital watermark, extract characteristics of image and set up index then, and put on record information and watermark data of this image and its is associated and accomplishes putting on record of image; The image of putting on record of embed digital watermark then returns to the user that puts on record and supplies its subsequent applications, promptly in external environment conditions such as internet, propagates the image of putting on record after this image version of putting on record of using must be embed digital watermark.Second aspect when carrying out image authentication, is at first passed through to extract the characteristic of image to be verified, and in all put image on record, is retrieved, and obtains some put on record image and related put on record information and watermark datas thereof that satisfy certain condition; Detect one by one after result for retrieval sorted according to similarity from big to small and whether have the result for retrieval corresponding digital watermarking of image of putting on record in the image to be verified; After detecting digital watermarking, put on record image and image to be verified that this digital watermarking is corresponding are compared, and detect whether there be distorting the former figure that puts on record in the image to be verified; Finally gather and return the checking result.
Can see that from above-mentioned principle of work the present invention is with the application image digital watermarking algorithm.This algorithm is divided into watermark embed process and watermark detection process two parts usually.Realize in the watermark embed process group identity information (being digital watermark data) is directly embedded among the digital picture; When needed, can from the image of embed watermark, extract the identification information that is embedded into, perhaps judge whether embedded certain known identification information in this image through watermark detection process.
The digital watermarking of embedded images does not influence the use value of original image, also is difficult for being perceived by people's vision, and be the basic demand of image digital watermark algorithm, i.e. the invisibility of watermark.
After the image of embed watermark is had a mind to or is by mistake carried out modifications such as filtering, interpolation noise, amplification/dwindle, lossy compression method, shearing, displacement; Whether watermark can also normally be detected; Be the important indicator of evaluation map as digital watermarking algorithm, i.e. robustness.
The image digital watermark algorithm is the technique known means; There is multiple disclosed algorithm to adopt; Its main technical principle is in watermark embed process, to utilize image transformation algorithms such as DFT or discrete cosine transform that digital service unit is arrived frequency domain; In frequency domain, select some position (like the intermediate frequency zone) carry out watermark add calculate after again inverse transformation go back, obtain the image of embed watermark; In watermark detection process then equally with the image transformation of embed watermark to frequency domain, utilize method such as correlation calculations to extract watermark then or detection has or not known watermark.
In practical implementation, the present invention can adopt the public image digital watermarking algorithm that possesses following performance: do not need former figure in the watermark detection process, be blind watermarking algorithm; Need from image, not extract digital watermark data in the watermark detection process, and only need judge whether embedded certain known digital watermark data in the image; Consider that the image of putting on record does not allow to make amendment in concrete the application, so less demanding to the robustness of digital watermarking algorithm, the digital watermarking that only needs assurance to embed can be resisted attack patterns such as scale, lossy compression method to a certain extent and got final product; Under the prerequisite that guarantees watermark robustness, improve the invisibility of watermark, make the embedding better effects if, reduce influence to the image practical application as far as possible.
Therefore, the present invention can select suitable digital watermarking algorithm in more wide in range selection face, as based on spread-spectrum watermark algorithm of DFT or discrete cosine transform etc.Simultaneously, along with the digital watermark technology progress, the innovation digital watermarking algorithm that possesses These characteristics also can directly apply among the present invention.
Fig. 1 is the image of the present invention process flow diagram with verification method of putting on record.
With reference to Fig. 1, wherein first three step is to carry out the operation of image when putting on record, and back three steps are the operation when carrying out image authentication.
Step 11 generates for digital watermarking and embeds step; Number the digital watermark data that waits the information calculations generation of putting on record to forge based on putting on record of the image of putting on record in this step; After preserving this watermark data it is embedded in the image of putting on record, obtains the image of putting on record of embed digital watermark.
Corresponding in order to obtain with the image of putting on record, and the digital watermark data that can not forge, will wait the information calculations of putting on record to generate the digital watermark data that can not forge based on the numbering of putting on record of the image of putting on record; As shown in Figure 2; The concrete computing method of digital watermark data are following: step 21, to the information of putting on record of the image of putting on record, utilize covert key; Adopt DEA to encrypt, obtain encrypting the information of putting on record.Step 22 to the encryption information of putting on record, adopts the data summarization algorithm to carry out abstract extraction, obtains a string summary yard.Step 23 according to digital watermarking desired data figure place, is taken out the plurality of fixed numerical digit in the summary sign indicating number, forms digital watermark data.
Wherein, the information of putting on record is carried out the ciphered data AES, can adopt commonly used, disclosed symmetric encipherment algorithm such as DES, TripleDES, RC4 and Blowfish, then need to be keep secret of encrypted secret key, underground.And the encryption information of putting on record is carried out the data summarization algorithm of abstract extraction, can adopt then that MD5, SHA1 etc. are commonly used, disclosed method.The summary sign indicating number that is obtained through symmetric cryptography and abstract extraction like this is corresponding with the information of putting on record of the image of putting on record, and can't forge because encryption key is underground.The figure place D of summary sign indicating number so can from the summary sign indicating number, extract the fixedly numerical digit of figure place, fixed position according to the scheme of setting, forms digital watermark data generally greater than the required figure place L of digital watermarking.For example, can extract the preceding L position of summary sign indicating number, form the digital watermark data of L bit 0-1 sequence.
Replace the original image of putting on record to be used for subsequent applications such as externally issue the image of putting on record of embed digital watermark in step 13.
When carrying out image authentication, need utilize CBIR technology, promptly according to image to be verified,, all retrieve in putting image on record, through extracting characteristics of image, find out and the most similar part of the image to be verified image of putting on record.Can adopt the various characteristics of image in the CBIR technology, comprise color, texture, edge etc., and the color histogram feature application be comparatively general, and calculate simple, distinguish functional, so can be adopted by the present invention.The similarity of calculating two width of cloth color of image histogram features then can adopt histogram to ask the method for friendship to calculate.In other words, two width of cloth image p
1And p
2Similarity S (p
1, p
2) calculate by formula:
Wherein h (p) is the normalization color histogram characteristic of image p, and B is the dimension of histogram feature, and the number of histogram bin (the bar post in the histogram).
Recall precision when making image authentication improves; After the characteristic of the image of putting on record that carries out having extracted when image is put on record embed watermark; Can utilize data clusters, local sensitivity Hash (Locality-Sensitive Hashing; LSH) etc. known method is set up index to the characteristic of the image of putting on record of all embed watermarks, reduces the number of times that characteristic similarity calculates, and improves recall precision.
When carrying out image authentication; The present invention retrieves in all put image on record through utilizing image to be verified, and acquisition is put on record by the N bar and write down the result for retrieval of forming; Wherein every put on record record by the image of putting on record, with the similarity of image to be verified; And corresponding put on record information and watermark data composition, and the image of putting on record of entering result for retrieval should possess following condition: 1, the similarity of this put on record image and image to be verified is greater than threshold value T
h, promptly fully similar with image to be verified.2, the area of the big image of area and the ratio of the area of the less image of area are less than threshold value T in this put on record image and image to be verified
s, promptly the area of this put on record image and image to be verified can not have big difference.
In step 15; Watermark detection based on result for retrieval; Based on result for retrieval,, detect the corresponding digital watermarking of image of putting on record of each width of cloth in the result for retrieval one by one and whether be present in the image to be verified according to the similarity order from high to low of put on record image and image to be verified; If in image to be verified, do not detect in the result for retrieval all corresponding digital watermarkings of image of putting on record, explain that then image to be verified puts on record.
When carrying out image authentication, the result who in all put image on record, retrieves based on image to be verified carries out the digital watermarking detection, obtains efficiently and accurately image authentication.As shown in Figure 3; Concrete testing process comprises: step 31: carry out initialization; Process object comprises image to be verified; And by the N bar result for retrieval formed of record of putting on record, wherein every record of putting on record comprise the image of putting on record that satisfies above-mentioned condition 1 and condition 2, with the similarity of image to be verified, and put on record the information and the watermark data of correspondence.Simultaneously, initializing variable i=1, i represent the i bar record of putting on record.Step 32: with the N bar in the result for retrieval put on record the record sort from big to small according to similarity with image to be verified.In step 33, utilize digital watermarking algorithm, in image to be verified, detect it and whether contain the digital watermarking in the record of putting on record of i bar in the result for retrieval.
If in step 33, detect digital watermarking, then jump to step 35, otherwise then jump to step 36.Step 35 detects successfully, and image to be verified is the image of putting on record, and corresponding to the record of putting on record of the i bar in the result for retrieval.Step 36 judges whether i equals N, if equate, the digital watermarking that does not exist all result for retrieval corresponding in the image to be verified is described then, jumps to step 38, otherwise then need detect the corresponding digital watermarking of next bar result for retrieval, jumps to step 37.Step 37: the variable i value increases by 1.Step 38 detects failure, and image to be verified is the image of not putting on record.
The image to be verified that detects the corresponding digital watermark data of image of putting on record must be the image through putting on record, still, also possibly carried out tiny distorting.Because the robustness of digital watermarking algorithm, image is tiny distorts the detection that possibly not influence wherein digital watermarking, but might cause the change of picture material.In order to detect and to prevent this effect that influence is put on record of distorting, the embodiment of the invention detects the place of distorting that possibly exist through the image to be verified and the image of putting on record are compared.Wherein as shown in Figure 4; Mainly utilize a length of side for the square moving window of l at every turn to the right or move down l/3 come to image travel through, piecemeal; Through detecting image to be verified and putting the pixel extent of image in same piecemeal position on record, find whether distort in this window.
It is concrete that to distort testing process as shown in Figure 5; Comprise: step 41: establish image to be verified and correspondence and put on record in the image; The picture traverse that area is less and highly be respectively w and h; The image that area is bigger is through convergent-divergent, is converted into width and highly is respectively the image of w and h, promptly makes the width of two width of cloth images and highly all be w and h.Step 42: use a length of side to come image is carried out piecemeal, begin from the upper left corner as the square moving window of l, window at every turn to the right or move down l/3, window travels through image to be verified and correspondence to be put on record behind the image, obtains M to image block, wherein M equals:
Step 43: poor according to the formula calculating pixel to every pair of image block:
Dp wherein
iBe the pixel difference of i to image block, (R '
Ij, G '
Ij, B '
Ij) and (R "
Ij, G "
Ij, B "
Ij) be respectively in image to be verified and the corresponding image of putting on record the RGB color component of j pixel in i the image block.
Step 44: it is poor to establish the right mean pixel of all images piecemeal
To every pair of image block, if DP
i＞max (λ DP
m, T
d), explain that then i distorts the place to the position existence of image block.Wherein λ is a pre-determined factor, and T
dIt then is a predetermined threshold value.
Step 45: if the right position of all images piecemeal does not exist do not distort, then jump to step 46, otherwise then jump to step 47.Step 46: verify successfully the put image of image to be verified for not distorting.Step 47: authentication failed, image to be verified is for existing the image of distorting of putting on record, and tampered position detects the position of the image block of distorting for all.
After the embodiment of the invention was treated authentication image and carried out image authentication operation, whether the checking result who obtains comprises: the result 1: whether image to be verified was the image of putting on record, be the image through putting on record promptly.As a result 2: if image to be verified is the image of putting on record, whether it distorts the picture material of putting on record.As a result 3:, point out the position of distorting if image to be verified is the image of putting on record after distorting.
In the embodiment of the invention, realize the putting on record and efficient verification of image, and be aided with the distorted image detection technique and obtain more complete checking result through combining CBIR technology and image digital watermark technology.
Based on same inventive concept; The system that the embodiment of the invention also provides a kind of image to put on record and verify; As shown in Figure 6, in this system, be provided with and comprise: at least one client modules 51 of putting on record is used for when image is put operation on record; Put on record image file input of reception, and return to put on record and be used for the image of subsequent applications after handling; At least one checking client module 52 is used for when image authentication is operated, and receives image file input to be verified, and returns the checking result; Scheduler module 53 is used to dispatch correlation module and handles; The storage center of putting on record module 54 is used for preserving put on record image and characteristics of image thereof, and is associated with put on record information and watermark data; Digital watermarking module 55 is used for generating digital watermark data in image put on record when operation according to the information of putting on record of the image of putting on record, and embeds in the image of putting on record; Simultaneously also be used for when image authentication is operated, detecting the digital watermarking whether image to be verified exists appointment.An image retrieval module 56 is used to realize through extracting the characteristic of image, and sets up aspect indexing, in all put image on record, retrieve fast eligible 1 with the image of putting on record of condition 2.Whether a distorted image detection module 57 is used for the image to be verified and the image of putting on record are compared, detect it and exist and distort.
As shown in Figure 6, the client modules of putting on record links to each other with scheduler module respectively with the checking client module, and digital watermarking module, image retrieval module and distorted image detection module then link to each other with the storage center module of putting on record with scheduler module simultaneously respectively.
Carrying out image when putting on record, the embodiment of the invention at first through the client modules of putting on record receive put on record image with put information on record, and send it to the digital watermarking module through scheduler module; After the digital watermarking module generates the digital watermark data that can not forge according to the information of putting on record of image, its embedding is put on record among the image, and the image of putting on record of embed digital watermark is sent to scheduler module respectively and puts the storage center module on record; Scheduler module returns to the user that puts on record through the client modules of putting on record with the image of putting on record of embed digital watermark and carries out subsequent applications; The storage center of putting on record module is preserved the image of putting on record of all embed digital watermarks, and related with put on record accordingly information and digital watermarking; Image retrieval module is extracted the characteristic of all images of putting on record of preserving in the storage center module of putting on record, and these characteristics of image are set up index.
When carrying out image authentication, the embodiment of the invention at first receives image to be verified through the checking client module, and sends it to image retrieval module through scheduler module; Image retrieval module is extracted the characteristic of image to be verified; In all put image on record retrieval obtain some satisfy condition 1 with put on record image and put on record the accordingly information and the digital watermarking of condition 2, and send to scheduler module after result for retrieval sorted according to similarity from big to small; Scheduler module is utilized the digital watermarking module according to result for retrieval, verifies the corresponding digital watermarking of information of putting on record that whether exists result for retrieval to put image on record in the image to be verified one by one; After detecting digital watermarking, whether scheduler module utilizes the distorted image detection module that the image to be verified and the image of putting on record are compared, detect to exist in the image to be verified and distort; Finally gather and return the checking result to the user through the checking client module by scheduler module.
After the embodiment of the invention is carried out image authentication; Draw the checking result who comprises result 1, result 2 and result 3; This checking result can be exported in the XML file outside the system by the checking client module; These XML files are loaded in the other system, can make other system obtain the result of image authentication.
Adopt the present invention to realize using the large nuber of images of propagating to carry out efficient verification down, judge whether it put on record, and whether on the image of putting on record, distort through file that image is put on record to environment such as internets.Because in proof procedure; Utilize the image retrieval technologies of rapid speed to filter out a large amount of and the dissimilar image of putting on record of image to be verified earlier; And then utilize slow digital watermarking detection technique according to image similarity to be verified sequence detection from big to small image to be verified in digital watermarking, last again to the image that detects digital watermarking carry out speed the slowest distort detection.The flow scheme design of this similar " waterfall " makes the present invention carry out the rapid speed of image authentication, satisfies the checking demand to large nuber of images.
Especially, put on record system and internet advertisement for pharmaceuticals acquisition system of the present invention and advertisement for pharmaceuticals integrated and implemented, can realize intellectual monitoring medicine image ad on the internet.Use through putting on record and the advertisement for pharmaceuticals image of embed watermark through requiring the advertisement for pharmaceuticals publisher; Let computer generation replace manual work to carry out the checking of advertisement for pharmaceuticals image; What whether affirmation advertisement for pharmaceuticals publisher was issued is the advertisement for pharmaceuticals image through putting on record; Whether the advertisement for pharmaceuticals image of putting on record is distorted, for the propagation of the magnanimity advertisement for pharmaceuticals image under the internet environment provides a kind of effective regulatory measure.
The present invention is applicable to advertisement for pharmaceuticals management and monitoring system.The relevant unit that supervises utilizes this system that advertisement for pharmaceuticals is registered, examines, puts on record, managed, and the advertisement for pharmaceuticals of propagating is on the internet monitored, and finds out and wherein puts on record or the illegal advertisement for pharmaceuticals through examining.The present invention is used for the image with quasi-drugs advertisement character is supervised, and this type image need be examined before application, put on record, and in application process, does not allow picture material is distorted.
Obviously, the engineering technical personnel of this area can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention, particularly can adopt different digital watermarking algorithms and characteristics of image.If these changes of the present invention and modification are belonged within claim of the present invention and the equivalent technologies scope thereof, then the present invention also is intended to comprise these changes and modification.
Claims (14)
1. an image is put on record and verification method, and the method comprising the steps of:
Step 11 generates digital watermark data based on the information calculations of putting on record of the image of putting on record, it is embedded in the image of putting on record after preserving this watermark data, obtains the image of putting on record of embed digital watermark;
Step 12 is extracted put on record the characteristics of image of image of embed digital watermark, preserves embed digital watermark put on record image and characteristic thereof, and is associated with corresponding information and the watermark data put on record;
Step 13 is with the original image of putting on record of image replacement of putting on record of embed digital watermark;
Step 14; To image to be verified, extract its characteristics of image and in all put image on record, retrieve, result for retrieval comprises put on record image and related put on record information and the watermark data thereof that satisfies certain condition; If do not retrieve, explain that then image to be verified puts on record;
Step 15; Based on result for retrieval; Similarity order from high to low according to put on record image and image to be verified; Detect the corresponding digital watermarking of image of putting on record of each width of cloth in the result for retrieval one by one and whether be present in the image to be verified,, explain that then image to be verified puts on record if in image to be verified, do not detect all corresponding digital watermarkings of image of putting on record in the result for retrieval.
2. method according to claim 1 is characterized in that, after step 15, further comprises:
Step 16, whether the corresponding former figure comparison of putting on record with it of the image to be verified that will contain digital watermarking detects to exist and distorts, and distorts if detect, and explains that then image to be verified is the image of putting on record after distorting, otherwise explains that then image to be verified is the image of putting on record normally.
3. the method for claim 1 is characterized in that, step 11 further comprises:
Step 21: to the information of putting on record of putting image on record, utilize covert key, adopt DEA to encrypt, obtain encrypting the information of putting on record;
Step 22: to encrypting the information of putting on record, adopt the data summarization algorithm to carry out abstract extraction, obtain a string summary sign indicating number;
Step 23: according to digital watermarking desired data figure place, take out the plurality of fixed numerical digit in the summary sign indicating number, form digital watermark data.
4. the method for claim 1 is characterized in that, the image of putting on record in the said result for retrieval need meet the following conditions: the similarity of this put on record image and image to be verified is greater than threshold value T
h, promptly fully similar with image to be verified; And the area of the big image of area and the ratio of the area of the less image of area are less than threshold value T in this put on record image and image to be verified
s, promptly the area of this put on record image and image to be verified can not have big difference.
5. the method for claim 1 is characterized in that, step 15 further comprises:
Step 31: make variable i=1, i representes the i bar that the N bar is put on record and write down in the result for retrieval.
Step 32: with the N bar in the result for retrieval put on record the record sort from big to small according to similarity with image to be verified;
Step 33: utilize digital watermarking algorithm, in image to be verified, detect and whether contain the digital watermarking in the record of putting on record of i bar in the result for retrieval;
Step 34: if in step 33, detect digital watermarking, then jump to step 35, otherwise then jump to step 36;
Step 35: detect successfully, image to be verified is the image of putting on record, and corresponding to the record of putting on record of the i bar in the result for retrieval;
Step 36: judge whether i equals N,, the digital watermarking that does not exist all result for retrieval corresponding in the image to be verified is described then, jump to step 38, otherwise then need detect the corresponding digital watermarking of next bar result for retrieval, jump to step 37 if equate;
Step 37: the variable i value increases by 1;
Step 38: detect failure, image to be verified is the image of not putting on record.
6. the method for claim 1 is characterized in that, step 16 further comprises:
Step 41: establish image to be verified and correspondence and put on record in the image; The picture traverse that area is less and highly be respectively w and h; The image that area is bigger is through convergent-divergent, is converted into width and highly is respectively the image of w and h, promptly makes the width of two width of cloth images and highly all be w and h;
Step 42: using a length of side is that 1 square moving window comes image is carried out piecemeal, begin from the upper left corner, window at every turn to the right or move down 1/3, window travels through image to be verified and correspondence to be put on record behind the image, obtains M to image block, wherein M equals:
Step 43: poor according to the formula calculating pixel to every pair of image block:
DP wherein
iBe the pixel difference of i to image block, (R '
Ij, G '
Ij, B '
Ij) and (R "
Ij, G "
Ij, B "
Ij) be respectively in image to be verified and the corresponding image of putting on record the RGB color component of j pixel in i the image block;
Step 44: it is poor to establish the right mean pixel of all images piecemeal
To every pair of image block, if DP
i＞max (λ DP
m, T
d), explain that then i distorts the place to the position existence of image block, wherein, λ is a pre-determined factor, and T
dIt then is a predetermined threshold value;
Step 45: if the right position of all images piecemeal does not exist do not distort, then jump to step 46, otherwise then jump to step 47;
Step 46: verify successfully the put image of image to be verified for not distorting;
Step 47: authentication failed, image to be verified is for existing the image of distorting of putting on record, and tampered position detects the position of the image block of distorting for all.
7. an image is put on record and verification system, and this system comprises:
The digital watermarking module is used for generating digital watermark data based on the information calculations of putting on record of the image of putting on record, it is embedded in the image of putting on record after preserving this watermark data, obtains the image of putting on record of embed digital watermark;
The storage center of putting on record module is used to preserve the image of putting on record of embed digital watermark, and is associated with corresponding information and the watermark data put on record, extract simultaneously embed digital watermark put on record image characteristics of image and set up index;
Image retrieval module; Be used for to image to be verified, extract its characteristics of image and in all put image on record, retrieve, result for retrieval comprises put on record image and related put on record information and the watermark data thereof that satisfies certain condition; If do not retrieve, explain that then image to be verified puts on record;
Said digital watermarking module is based on result for retrieval; Similarity order from high to low according to put on record image and image to be verified; Detect image to be verified whether contain with result for retrieval in the corresponding digital watermarking of image of putting on record; If do not detect corresponding digital watermarking, explain that then image to be verified puts on record.
8. system according to claim 7 is characterized in that this system also comprises
The distorted image detection module; Be used for to contain the corresponding former figure comparison of putting on record of image to be verified of digital watermarking, detect whether to exist and distort, distort if detect with it; Explain that then image to be verified is the image of putting on record after distorting, otherwise explain that then image to be verified is the image of putting on record normally.
9. system as claimed in claim 7 is characterized in that, the digital watermarking module generates digital watermarking in the following manner:
Step 21: to the information of putting on record of putting image on record, utilize covert key, adopt DEA to encrypt, obtain encrypting the information of putting on record;
Step 22: to encrypting the information of putting on record, adopt the data summarization algorithm to carry out abstract extraction, obtain a string summary sign indicating number;
Step 23: according to digital watermarking desired data figure place, take out the plurality of fixed numerical digit in the summary sign indicating number, form digital watermark data.
10. system as claimed in claim 7 is characterized in that, the image of putting on record in the said result for retrieval need meet the following conditions: the similarity of this put on record image and image to be verified is greater than threshold value T
h, promptly fully similar with image to be verified; And the area of the big image of area and the ratio of the area of the less image of area are less than threshold value T in this put on record image and image to be verified
s, promptly the area with image to be verified can not have big difference.
11. system as claimed in claim 7 is characterized in that, the digital watermarking module detects digital watermarking in the following manner:
Step 31: make variable i=1, i representes the i bar that the N bar is put on record and write down in the result for retrieval.
Step 32: with the N bar in the result for retrieval put on record the record sort from big to small according to similarity with image to be verified;
Step 33: utilize digital watermarking algorithm, in image to be verified, detect and whether contain the digital watermarking in the record of putting on record of i bar in the result for retrieval;
Step 34: if in step 33, detect digital watermarking, then jump to step 35, otherwise then jump to step 36;
Step 35: detect successfully, image to be verified is the image of putting on record, and corresponding to the record of putting on record of the i bar in the result for retrieval;
Step 36: judge whether i equals N,, the digital watermarking that does not exist all result for retrieval corresponding in the image to be verified is described then, jump to step 38, otherwise then need detect the corresponding digital watermarking of next bar result for retrieval, jump to step 37 if equate;
Step 37: the variable i value increases by 1;
Step 38: detect failure, image to be verified is the image of not putting on record.
12. system as claimed in claim 7 is characterized in that, the distorted image detection module is further carried out:
Step 41: establish image to be verified and correspondence and put on record in the image; The picture traverse that area is less and highly be respectively w and h; The image that area is bigger is through convergent-divergent, is converted into width and highly is respectively the image of w and h, promptly makes the width of two width of cloth images and highly all be w and h;
Step 42: using a length of side is that 1 square moving window comes image is carried out piecemeal, begin from the upper left corner, window at every turn to the right or move down 1/3, window travels through image to be verified and correspondence to be put on record behind the image, obtains M to image block, wherein M equals:
Step 43: poor according to the formula calculating pixel to every pair of image block:
DP wherein
iBe the pixel difference of i to image block, (R '
Ij, G '
Ij, B '
Ij) and (R "
Ij, G "
Ij, B "
Ij) be respectively in image to be verified and the corresponding image of putting on record the RGB color component of j pixel in i the image block;
Step 44: it is poor to establish the right mean pixel of all images piecemeal
To every pair of image block, if DP
i＞max (λ DP
m, T
d), explain that then i distorts the place to the position existence of image block, wherein, λ is a pre-determined factor, and T
dIt then is a predetermined threshold value;
Step 45: if the right position of all images piecemeal does not exist do not distort, then jump to step 46, otherwise then jump to step 47;
Step 46: verify successfully the put image of image to be verified for not distorting;
Step 47: authentication failed, image to be verified is for existing the image of distorting of putting on record, and tampered position detects the position of the image block of distorting for all.
13., further comprise according to each described system of claim 1-12:
At least one client modules of putting on record is used for putting on record when operation at image, is used for receiving the image file input of putting on record, and returns to put on record and be used for the image of subsequent applications after handling;
At least one checking client module is used for when image authentication is operated, and is used to receive image file input to be verified, and returns the checking result.
Scheduler module is used to dispatch correlation module and handles.
14. image as claimed in claim 13 is put on record and verification system; It is characterized in that; In this system; The client modules of putting on record links to each other with scheduler module respectively with the checking client module, and digital watermarking module, image retrieval module and distorted image detection module then link to each other with the storage center module of putting on record with scheduler module simultaneously respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102625278A CN102446211A (en) | 2011-09-06 | 2011-09-06 | Method and system for filing and verifying image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102625278A CN102446211A (en) | 2011-09-06 | 2011-09-06 | Method and system for filing and verifying image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102446211A true CN102446211A (en) | 2012-05-09 |
Family
ID=46008711
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011102625278A Pending CN102446211A (en) | 2011-09-06 | 2011-09-06 | Method and system for filing and verifying image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102446211A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104980799A (en) * | 2014-04-04 | 2015-10-14 | 北京秒针信息咨询有限公司 | Advertisement detecting method and system |
| CN109313659A (en) * | 2016-06-21 | 2019-02-05 | 电子湾有限公司 | The abnormality detection of web document revision |
| CN109325933A (en) * | 2017-07-28 | 2019-02-12 | 阿里巴巴集团控股有限公司 | A kind of reproduction image-recognizing method and device |
| CN110276709A (en) * | 2019-06-18 | 2019-09-24 | 周晓明 | It is a kind of for the generation of protectiveness document, discrimination method and information management system |
| CN111461254A (en) * | 2020-04-17 | 2020-07-28 | 支付宝(杭州)信息技术有限公司 | Picture detection method and device and electronic equipment |
| CN111640051A (en) * | 2019-03-01 | 2020-09-08 | 浙江大学 | Image processing method and device |
| CN111814793A (en) * | 2020-09-09 | 2020-10-23 | 耀方信息技术(上海)有限公司 | Medicine image-text information violation detection method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1710577A (en) * | 2005-07-08 | 2005-12-21 | 王树勋 | Digit water-mark certificate anti-fake method based on mobile communication network |
| CN1858799A (en) * | 2005-05-08 | 2006-11-08 | 中国科学院计算技术研究所 | Digital image hash signature method |
| CN101059863A (en) * | 2006-04-20 | 2007-10-24 | 刘瑞祯 | Embed and detection method for identifying water mark, its system and uses |
| CN101082981A (en) * | 2007-05-22 | 2007-12-05 | 中山大学 | Watermark embeding and extracting method of binary image |
| CN102004874A (en) * | 2010-11-05 | 2011-04-06 | 王义海 | Digital resource monitoring and early warning method based on synthetic characteristics |
| CN102013088A (en) * | 2010-11-19 | 2011-04-13 | 王义海 | Digital watermark generation and embedding method based on comprehensive characteristics of digital archival resources |
-
2011
- 2011-09-06 CN CN2011102625278A patent/CN102446211A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1858799A (en) * | 2005-05-08 | 2006-11-08 | 中国科学院计算技术研究所 | Digital image hash signature method |
| CN1710577A (en) * | 2005-07-08 | 2005-12-21 | 王树勋 | Digit water-mark certificate anti-fake method based on mobile communication network |
| CN101059863A (en) * | 2006-04-20 | 2007-10-24 | 刘瑞祯 | Embed and detection method for identifying water mark, its system and uses |
| CN101082981A (en) * | 2007-05-22 | 2007-12-05 | 中山大学 | Watermark embeding and extracting method of binary image |
| CN102004874A (en) * | 2010-11-05 | 2011-04-06 | 王义海 | Digital resource monitoring and early warning method based on synthetic characteristics |
| CN102013088A (en) * | 2010-11-19 | 2011-04-13 | 王义海 | Digital watermark generation and embedding method based on comprehensive characteristics of digital archival resources |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104980799A (en) * | 2014-04-04 | 2015-10-14 | 北京秒针信息咨询有限公司 | Advertisement detecting method and system |
| CN109313659A (en) * | 2016-06-21 | 2019-02-05 | 电子湾有限公司 | The abnormality detection of web document revision |
| CN109313659B (en) * | 2016-06-21 | 2022-07-12 | 电子湾有限公司 | Anomaly detection for revisions to web documents |
| CN109325933A (en) * | 2017-07-28 | 2019-02-12 | 阿里巴巴集团控股有限公司 | A kind of reproduction image-recognizing method and device |
| CN109325933B (en) * | 2017-07-28 | 2022-06-21 | 阿里巴巴集团控股有限公司 | Method and device for recognizing copied image |
| CN111640051A (en) * | 2019-03-01 | 2020-09-08 | 浙江大学 | Image processing method and device |
| CN111640051B (en) * | 2019-03-01 | 2023-07-25 | 浙江大学 | Image processing method and device |
| CN110276709A (en) * | 2019-06-18 | 2019-09-24 | 周晓明 | It is a kind of for the generation of protectiveness document, discrimination method and information management system |
| CN111461254A (en) * | 2020-04-17 | 2020-07-28 | 支付宝(杭州)信息技术有限公司 | Picture detection method and device and electronic equipment |
| CN111814793A (en) * | 2020-09-09 | 2020-10-23 | 耀方信息技术(上海)有限公司 | Medicine image-text information violation detection method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102446211A (en) | Method and system for filing and verifying image | |
| Davarzani et al. | Perceptual image hashing using center-symmetric local binary patterns | |
| Karsh et al. | Robust image hashing through DWT-SVD and spectral residual method | |
| Farfoura et al. | A novel blind reversible method for watermarking relational databases | |
| CN102903075B (en) | Robust watermarking method based on image feature point global correction | |
| Shojae Chaeikar et al. | PSW statistical LSB image steganalysis | |
| CN102073980B (en) | Compression sensing theory-based interactive supported dual watermark generating and detecting method | |
| Hadmi et al. | Perceptual image hashing | |
| CN102622545A (en) | Picture file tracking method | |
| US11216536B2 (en) | Data watermarking and fingerprinting system and method | |
| CN104166955A (en) | Conformal transformation-based image Hash generation method and image tampering detection and positioning method | |
| Lou et al. | Construction information authentication and integrity using blockchain-oriented watermarking techniques | |
| US20210026598A1 (en) | Salting Text and Fingerprinting in Database Tables, Text Files, and Data Feeds | |
| CN105069158B (en) | Data digging method and system | |
| Alkhowaiter et al. | Evaluating perceptual hashing algorithms in detecting image manipulation over social media platforms | |
| Abdulqader et al. | Detection of tamper forgery image in security digital mage | |
| CN103106656B (en) | Image signatures based on profile wave convert generates and tampering detection and localization method | |
| Chai et al. | A robust and reversible watermarking technique for relational dataset based on clustering | |
| CN103177219A (en) | Medical biochemical detection report critical data tamper-proofing method | |
| Toraskar et al. | Efficient computer forensic analysis using machine learning approaches | |
| Nazli et al. | Comparison between image forgery detection algorithms | |
| Behura et al. | Counterfeit product detection analysis and prevention as well as prepackage coverage assessment using machine learning | |
| Kishore et al. | Faster file imaging framework for digital forensics | |
| Yan et al. | Adaptive local feature based multi-scale image hashing for robust tampering detection | |
| Xiang et al. | Histogram-based image hashing for searching content-preserving copies |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120509 |