CN100346322C - Digital image recovering method based on digital water mark technology - Google Patents
Digital image recovering method based on digital water mark technology Download PDFInfo
- Publication number
- CN100346322C CN100346322C CNB2003101001183A CN200310100118A CN100346322C CN 100346322 C CN100346322 C CN 100346322C CN B2003101001183 A CNB2003101001183 A CN B2003101001183A CN 200310100118 A CN200310100118 A CN 200310100118A CN 100346322 C CN100346322 C CN 100346322C
- Authority
- CN
- China
- Prior art keywords
- image
- digital
- watermark
- water mark
- data
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a digital image recovering method based on a digital water mark technique, which helps a damaged image to be recovered by the digital water mark technique. The digital image recovering method based on a digital water mark technique comprises the following steps: the characteristic of the image extracted by small wave multi-resolution disintegration operation is converted into a water mark signal by a quantization encoding process, and a water mark is embedded into the image by a water mark embedding algorithm; the digital image embedded with the information of the self characteristic is obtained by small wave multi-resolution synthesis operation, when the processed digital image loses data in transmission because of packet loss or hacker attack, the water mark can be extracted from the image by a water mark extracting algorithm, the information of the characteristic of the image is recovered by a decoding process, and prior knowledge is provided for the lost data; the information of the characteristic and the damaged image are fused by an image fusion concept so as to recover the lost contents of the image. The present invention can enhance the visual quality of the digital image under the adverse condition of the transmission, and the present invention can be used for protecting the integrity of the contents of the digital image.
Description
Technical field:
The present invention relates to the multimedia messages process field.Be to utilize digital watermark technology to help damaged image to carry out content to restore specifically, can improve the visual quality of images of digital picture under abominable transmission conditions, also can be used for protecting the integrality of digital image content.Be applicable to the diverse network of units such as government, enterprise, army, bank, library.
Background technology:
Multimedia messages the phenomenon of loss of data and distortion all might occur in transmission course when if network congestion taking place or being subjected to hacker's malicious attack.In order to improve signal transmission quality, minimizing is also repaired the information of losing, and the various countries scholar has proposed the multiple technologies strategy.From the angle of information coding, can be divided into initiatively and passive two kinds of solutions.Active solution mainly comprises fault-tolerant encoding technology and channel-information source combined coding technology.Though this class scheme can be accomplished the intact reconstruction of data, but exist the problem compatible mutually with the signal encoding standard, and compare under the serious situation in loss of data, active solution might lose efficacy, and at this moment needed to adopt the passive-type scheme to remedy.The passive-type solution mainly refers to the signal processing technology damaged data recovered at receiving end, as error concealment and image restoration technology, promptly utilizes the data that receive, and estimates the data with reconstructing lost.The digital image restoration algorithm based on digital watermark technology that the present invention proposes belongs to the passive-type solution.
At present, digital image restoration technology both domestic and external mainly is to utilize the correlativity of data block on the spatial domain, promptly utilizes the character of continuous gray scale conversion between spatial domain epigraph piece and the piece, estimates the image block of losing by linear interpolation techniques.Lose the value of each pixel in the image block, obtain through linear interpolation by the pixel on the same locus in the adjacent block.If the video data block of losing is more concentrated, then need to restore through recursive operation, promptly at first rebuild the obliterated data piece that those have more adjacent block, the data block that these reconstructions are obtained is handled the image block of losing that those have less adjacent block again as known conditions then.Obviously, owing to when data are rebuild, lack the priori of reconstructed object, therefore the data volume when missing image be difficult in this way recover the picture material that the obliterated data piece is expressed, so quality of reconstructed images was relatively poor when very big (as through the bigger shearing manipulation of area).
Summary of the invention:
The present invention utilizes digital watermark technology to help damaged image to carry out the content recovery, and the feature that is about to image is hidden in the image self as watermark, utilizes the characteristic information that extracts from image when signal restoring, helps damaged image to recover the data message of losing.The specific implementation method is as follows:
1. digital picture is before entering transmission channel, at first utilize wavelet multiresolution rate operation splitting to extract characteristics of image, then with the characteristic information of real number character through the quantization encoding process convert to be suitable as that watermark embeds { 0, the 1} signal repeats watermark to be embedded in the details subgraph of image different resolution and different directions by watermarking algorithm.Neither can influence the visual quality and the use of host image after watermark embeds, also can not increase the volume of host image.After this, obtain being embedded with the digital picture of unique characteristics information, send into transmission channel through after the common digital coding processing procedure through wavelet multiresolution rate synthetic operation.
2. at the image receiving end,, then image is carried out the wavelet multiresolution rate and decompose, utilize watermark extracting algorithm and weighting watermark recovery mechanism to obtain watermark signal then, revert to the characteristic information of image then by decode procedure if find that picture material is impaired.Utilize the thought of image co-registration at last, characteristic information and damaged image are merged, recover the picture material of losing.
3. the approximate subgraph that the image wavelet multiresolution is decomposed is as the characteristic information of image, and this is because the data volume of approximate subgraph is little, only is 1/4 of former figure
L, (L is the number of plies of wavelet decomposition), but comprised the Global Information of image, so the importance height.
4. the coding-decoding operation that characteristic information and watermark signal are changed mutually is as follows:
(1) quantization encoding operation: at first find the maximum of T of characteristic information, and determine 8 intervals: (0, T/8], (T/8,2T/8], (2T/8,3T/8], (3T/8,4T/8], (4T/8,5T/8], (5T/8,6T/8], (6T/8,7T/8], (7T/8, T], the position of each element on number axis in the comparative feature information then.If certain numerical digit is then used the final coding form of the binary representation of m as this number in m interval.Because 8 states have been represented in 8 intervals, so only need the just value of a characteristic information of energy perfect representation of 3 bits.
(2) decode operation: after receiving end utilizes the watermark extracting algorithm to obtain watermark information, at first recover the interval numbering at 1 characteristic information place by per 3 watermark signals, recover the value of characteristic information then, that is: if this characteristic information in m interval, is then used the value of middle this characteristic information of value representation in this interval.
(3) the important parameter T that uses in the encoding-decoding process is embedded in the image after also needing to convert binary mode to.It can select the embedded location different with further feature information.The details subgraph that among the present invention all characteristic informations is embedded into the host, and T is embedded in the approximate subgraph.
5. the embedding of watermark and extraction algorithm are as follows:
(1) the embedding algorithm of watermark: the embedding of watermark is to finish by the wavelet coefficient in the image detail subgraph is carried out replacement operation.When watermark signal was 0, wavelet coefficient was replaced by nearest with it category-A point; When watermark signal was 1, wavelet coefficient was replaced by nearest with it category-B point.Wherein, the category-A point is defined as (2k+1/2) Δ, and the category-B point is defined as (2k-1/2) Δ, k=0, and ± 1, ± 2K, Δ are the human eye threshold of visibility of wavelet coefficient correspondence.Carry out inverse wavelet transform then, obtain being embedded with the image of unique characteristics information.Because the transformation range of wavelet coefficient is less than human eye threshold of visibility Δ, so this embedding strategy invisibly has been embedded among the image watermark.
(2) extraction algorithm of watermark: the image that at first will contain unique characteristics carries out the wavelet multiresolution rate identical with watermark embed process and decomposes, then with the wavelet coefficient in the details subgraph divided by Δ, carry out Modulo-two operation after the merchant rounded, just obtained watermark signal.
6. in order to guarantee that as far as possible image still can correctly recover characteristic information after the lost part data, the recovery for damaged image provides condition then, has adopted following measure:
(1) before embedding operation, watermark signal is carried out pseudo-random permutation.
(2) watermark signal is repeated to be embedded in the image.If wherein a part of watermark is damaged because of loss of image data, so still can in remaining watermark, find correct watermark signal.
(3) adopt weighting watermark recovery mechanism.The ability that different wavelet coefficient opposings is attacked is inequality, the less coefficient of the energy force rate absolute value that the opposing of the bigger coefficient of absolute value is attacked is strong, wavelet coefficient under the low resolution is stronger than the coefficient resistivity under the high resolving power, and the coefficient resistivity of the wavelet coefficient comparison angular direction of level, vertical direction is strong.Therefore can be according to the ability of wavelet coefficient opposing attack, for the watermark signal that extracts from these wavelet coefficients is given different credibilities.The final watermark value of determining is the weighted sum of the watermark signal that is repeated to embed.
7. if find that at receiving end picture material loses, can from image, extract characteristic information, utilize the method for image co-registration to recover the picture material of losing.That is: the priori of characteristic information as obliterated data directly interpolated in the corresponding data block, combine with other Image Restoration Algorithm then, make the grey scale change between reconstructed data block and the adjacent data blocks more level and smooth.This method can improve the reconstruction quality of image largely, even when the missing image data block is relatively concentrated, when area is big, utilize the still content implication of recover data losses part largely of this method, its quality of reconstructed images has had significant raising than other algorithm.
Description of drawings:
Fig. 1. utilize digital watermark technology embedded images unique characteristics information in digital picture;
Fig. 2. utilize digital watermark technology to extract characteristic information in the image and the process flow diagram that picture material is restored.
Embodiment:
Below in conjunction with the digital image restoration method performing step of description of drawings based on digital watermark technology:
Fig. 1: illustrate that digital picture is before entering transmission channel, at first digital picture being carried out the wavelet multiresolution rate decomposes, then the approximate subgraph under the image lowest resolution is encoded as the characteristic information of image, the encoded process of characteristic information that is about to real number character converts to and is suitable as { 0 of watermark embedding, the 1} signal, and repeat to be embedded in watermark in the details subgraph of image different resolution and different directions by watermarking algorithm, obtain being embedded with the digital picture of unique characteristics information then through wavelet multiresolution rate synthetic operation, send into transmission channel through after the common digital coding processing procedure.
Fig. 2: at the image receiving end, if find that picture material is impaired, then image being carried out the wavelet multiresolution rate decomposes, utilize watermark extracting algorithm and weighting watermark recovery mechanism to obtain watermark information then in the details subgraph under different resolution and different directions, then recover the characteristic information of image by decode procedure, with the characteristic information of image priori, merge at last, thereby recover the picture material of losing through Image Fusion and damaged image as obliterated data.
Claims (4)
1, based on the digital image restoration method of digital watermark technology, be to utilize digital watermark technology to help damaged image to carry out content to restore, be to be hidden in the middle of the image self as watermark behind the feature coding of digital picture specifically, if image generation loss of data, then can be from the image that contains watermark characteristic information extraction, for image restoration provides priori, help damaged image to restore the data of losing, improve the reconstruction quality of image.
2, the digital image restoration method based on digital watermark technology according to claim 1 is characterized in that:
Method is after the image wavelet multiresolution decomposes, directly will comprise the feature of the approximate subgraph of integral image information as image, convenient and swift, then by the quantization encoding process with the characteristic information of real number character convert to be suitable as that watermark embeds 0, the 1} signal.
3, the digital image restoration method based on digital watermark technology according to claim 1 is characterized in that:
For under the prerequisite of not damaging the digital picture visual quality, the big as far as possible watermark signal of energy is embedded in the image, make watermark can not be subjected to the influence of data coding-decoding operation and correctly extraction, in the watermarking algorithm with the human eye threshold of visibility as quantized interval, after embedding, watermark neither can influence the visual quality and the use of host image, also can not increase the volume of host image, host image still can be sent into transmission channel after common processing procedure.
4, the digital image restoration method based on digital watermark technology according to claim 1 is characterized in that:
At the image receiving end, the watermark information that utilizes watermark extracting algorithm and weighting watermark recovery mechanism from image, to obtain, can revert to the characteristic information of image by decode procedure, and provide priori for the data of missing image, this method is the content implication of recover data losses part largely, thereby improves the reconstruction quality of image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2003101001183A CN100346322C (en) | 2003-10-10 | 2003-10-10 | Digital image recovering method based on digital water mark technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2003101001183A CN100346322C (en) | 2003-10-10 | 2003-10-10 | Digital image recovering method based on digital water mark technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1492338A CN1492338A (en) | 2004-04-28 |
| CN100346322C true CN100346322C (en) | 2007-10-31 |
Family
ID=34256191
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2003101001183A Expired - Fee Related CN100346322C (en) | 2003-10-10 | 2003-10-10 | Digital image recovering method based on digital water mark technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100346322C (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100385459C (en) * | 2006-07-11 | 2008-04-30 | 电子科技大学 | Image watermark method based on finite ridgelet transform |
| JP5036643B2 (en) * | 2007-07-18 | 2012-09-26 | 株式会社リコー | Information processing apparatus, information processing method, and program |
| US8639062B2 (en) | 2007-10-09 | 2014-01-28 | Bank Of America Corporation | Ensuring image integrity using document characteristics |
| CN101727572A (en) * | 2008-10-20 | 2010-06-09 | 美国银行公司 | Method for ensuring image integrity by using file characteristics |
| CN101930601B (en) * | 2010-09-01 | 2012-02-08 | 浙江大学 | Edge information-based multi-scale blurred image blind restoration method |
| CN102129655B (en) * | 2011-02-20 | 2012-10-24 | 西安电子科技大学 | Wavelet domain-based method for weighting fractional differential image digital watermark |
| CN102695059B (en) * | 2012-05-31 | 2014-10-08 | 西安空间无线电技术研究所 | Method for hiding, compressing and transmitting images |
| CN103796024B (en) * | 2014-01-17 | 2017-05-03 | 西安空间无线电技术研究所 | Method for transmitting multiple paths of images at the same time based on information hiding |
| CN103955877B (en) * | 2014-01-21 | 2018-01-26 | 广州大学 | A kind of method and device of multimedia sensing network watermark verification and image repair |
| US10437778B2 (en) | 2016-02-08 | 2019-10-08 | Bank Of America Corporation | Archive validation system with data purge triggering |
| US10460296B2 (en) | 2016-02-08 | 2019-10-29 | Bank Of America Corporation | System for processing data using parameters associated with the data for auto-processing |
| US9823958B2 (en) | 2016-02-08 | 2017-11-21 | Bank Of America Corporation | System for processing data using different processing channels based on source error probability |
| US10437880B2 (en) | 2016-02-08 | 2019-10-08 | Bank Of America Corporation | Archive validation system with data purge triggering |
| US10067869B2 (en) | 2016-02-12 | 2018-09-04 | Bank Of America Corporation | System for distributed data processing with automatic caching at various system levels |
| US9952942B2 (en) | 2016-02-12 | 2018-04-24 | Bank Of America Corporation | System for distributed data processing with auto-recovery |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1199304A (en) * | 1997-05-08 | 1998-11-18 | 株式会社东芝 | Additional information insertion and regeneration apparatus, method and record medium |
| CN1254230A (en) * | 1998-07-01 | 2000-05-24 | 三星电子株式会社 | Digital image coding/decoding device using watermark and its method |
| CN1276565A (en) * | 1999-06-08 | 2000-12-13 | 中国科学院自动化研究所 | Method for protecting multimedia file and data |
| CN1325577A (en) * | 1998-09-10 | 2001-12-05 | 李相夔 | Watermarking of digital images using wavelet and discrete cosine tranforms |
| JP2002319924A (en) * | 2001-01-18 | 2002-10-31 | Canon Inc | Method and device for processing data signal |
| CN1379324A (en) * | 2002-05-09 | 2002-11-13 | 宣国荣 | Digital watermark method based on integer wavelet without damage to image |
-
2003
- 2003-10-10 CN CNB2003101001183A patent/CN100346322C/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1199304A (en) * | 1997-05-08 | 1998-11-18 | 株式会社东芝 | Additional information insertion and regeneration apparatus, method and record medium |
| CN1254230A (en) * | 1998-07-01 | 2000-05-24 | 三星电子株式会社 | Digital image coding/decoding device using watermark and its method |
| CN1325577A (en) * | 1998-09-10 | 2001-12-05 | 李相夔 | Watermarking of digital images using wavelet and discrete cosine tranforms |
| CN1276565A (en) * | 1999-06-08 | 2000-12-13 | 中国科学院自动化研究所 | Method for protecting multimedia file and data |
| JP2002319924A (en) * | 2001-01-18 | 2002-10-31 | Canon Inc | Method and device for processing data signal |
| CN1379324A (en) * | 2002-05-09 | 2002-11-13 | 宣国荣 | Digital watermark method based on integer wavelet without damage to image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1492338A (en) | 2004-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100346322C (en) | Digital image recovering method based on digital water mark technology | |
| Miller et al. | Applying informed coding and embedding to design a robust high-capacity watermark | |
| US7389420B2 (en) | Content authentication and recovery using digital watermarks | |
| Lancini et al. | A robust video watermarking technique in the spatial domain | |
| CN107688731B (en) | Digital watermarking algorithm based on text document protection | |
| Chang et al. | An effective image self-recovery based fragile watermarking using self-adaptive weight-based compressed AMBTC | |
| Lin et al. | Digital images watermarking by vector quantization | |
| Sarkar et al. | Large scale image tamper detection and restoration | |
| CN110097490A (en) | The insertion of area-of-interest reversible water mark and extracting method based on IWT-Schur | |
| Kishor et al. | A review on steganography through multimedia | |
| CN106875324B (en) | Lossless image information concealing method based on SBDE | |
| CN115880125A (en) | Soft fusion robust image watermarking method based on Transformer | |
| ShuangKui et al. | A modification-free steganography method based on image information entropy | |
| Yang et al. | Reversible data hiding in the spatial and frequency domains | |
| CN1377180A (en) | Method for protecting multimedia video data | |
| Ohyama et al. | Lossless data hiding using bit-depth embedding for JPEG2000 compressed bit-stream | |
| CN111583086A (en) | Adaptive digital image watermarking and repairing method based on AMBTC | |
| CN116523758B (en) | End cloud combined super-resolution video reconstruction method and system based on key frames | |
| Lee et al. | Temporal feature modulation for video watermarking | |
| Huang | A reversible visible watermarking technique for BTC-compressed images | |
| Seo et al. | QIM Watermarking for Image with Two Adaptive Quantization Ste-sizes | |
| Liu et al. | Image Self-Recovery Based on Authentication Feature Extraction | |
| Ghasal et al. | Payload minion stenographic technique for color images | |
| Kumar et al. | An optimally robust digital watermarking algorithm for stereo image coding | |
| Yongqiang et al. | Gray image watermark method based on fractal compression |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |