CN1226860C - Image processing apparatus and image processing method and program and storage media - Google Patents
Image processing apparatus and image processing method and program and storage media Download PDFInfo
- Publication number
- CN1226860C CN1226860C CNB021419965A CN02141996A CN1226860C CN 1226860 C CN1226860 C CN 1226860C CN B021419965 A CNB021419965 A CN B021419965A CN 02141996 A CN02141996 A CN 02141996A CN 1226860 C CN1226860 C CN 1226860C
- Authority
- CN
- China
- Prior art keywords
- character
- mentioned
- image processing
- characters
- radicals
- 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
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Editing Of Facsimile Originals (AREA)
- Image Processing (AREA)
Abstract
To ensure the accuracy and amount of information embedding above a certain level while suppressing the degradation of the font at a minimum. SOLUTION: The character which appears most frequently is decomposed into radicals and their reference values are obtained. In this case, a reference value is represented by a relative distance among coordinates of four edges of a character image. The relative position of each radical of the second or later character which is selected in the step S412 and corresponds to each specified bit of document access control information to be embedded is changed according to the embedding information in consideration of the reference value.
Description
Technical field
The present invention relates to the document image that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is carried out the image processing apparatus of the embedding of eletric watermark, the document image that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries extracted the image processing apparatus and the image processing method of the eletric watermark that embeds.
Background technology
Form in the device in digital pictures such as printer, photocopiers in recent years, its picture quality significantly improves, and can easily obtain high-quality printed article.Be anyone image processing, can both obtain desired printed article by being undertaken by high performance scanner, printer, photocopier and computer.Therefore, file takes place in regular meeting wrongfully problem such as duplicates, distorts, and in order to prevent or suppress the generation of this type of phenomenon, in recent years access control information is embedded the work very active (eletric watermark) of printed article itself.
As the method that realizes such requirement, generally be to make naked eyes invisibly access control information be embedded in the printed article now, perhaps will embed the blank space of file corresponding to the bitmap graphics of access control information, perhaps the scramble password is added in the document image.Wherein, making naked eyes embed the method for access control information invisibly, generally is to adopt following forms to realize: by the space amount between the control English character string, embed the form of information; By the rotation amount of control character, embed the form of information; By the amplification reduction volume of control character, embed the form of information etc.
Fig. 9 is the space amount between the explanation control English character string, carries out the figure of method of the embedding of information.Here, be called the space with 801～804.In addition, the interval in space 801 is set at p, the interval in space 802 is set at s.Under this state, if the position of the information that embeds is 0, then interval p, the s in space 801,802 changed to (p+s)/2 of (p+s)/2 of p ← (1+p), s ← (1-p),, then change to (p+s)/2 of (p+s)/2 of p ← (1-p), s ← (1+p) if the position of the information that embeds is 1.This can be applicable to space 803,804 equally.
Figure 10 is the rotation amount of explanation control character, carries out the figure of method of the embedding of information.Here, the state before this figure left side expression rotation, postrotational state is represented on this figure right side.The anglec of rotation of 901 expression characters.Identical with method shown in Figure 9, make of the position variation of the angle of its rotation corresponding to the information that embeds.
Figure 11 is the amount that explanation is dwindled by the amplification of control character, embeds the figure of the method for information.The original size of 1001 expressions.Size after 1002 expressions are amplified.Identical with method shown in Figure 9, make of the position variation of the amount of its amplification corresponding to the information that embeds.The situation of dwindling too.
; though embedding the method that access control information loses with seeing, above-mentioned naked eyes help maintaining secrecy; but in the document image that the redundancy of information embedded images is few (being generally the diadic image); and inharmonious sense to character, space can take place, it is very showy that the deterioration of original copy grade becomes.In addition, a little less than the anti-printing of so in general image (to the confining force of paper output back information) also.
The present invention finishes in view of above problem, and purpose is that degradation inhibiting with font in Min., guarantees that simultaneously to a certain degree above information embeds precision and embedded quantity.
Summary of the invention
In order to reach purpose of the present invention, image processing apparatus of the present invention comprises following structure.
That is, a kind of image processing apparatus to the document image embed watermark information that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is characterized in that comprising: the draw-out device that extracts character from above-mentioned document image; In the character that selection is extracted by above-mentioned draw-out device, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character gets the choice device of the character of predetermined structure; And the position of the predetermined character radicals by which characters are arranged in traditional Chinese dictionaries by making the character of being selected by above-mentioned choice device is changed to position according to the watermark information decision, the flush mounting of embed watermark information in the character of above-mentioned selection with respect to unchanged other the character radicals by which characters are arranged in traditional Chinese dictionaries in position in this character.
In addition, in order to reach purpose of the present invention, image processing apparatus of the present invention comprises following structure.
That is, a kind of the document image that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is extracted the image processing apparatus of the watermark information that is embedded into, it is characterized in that comprising: the character extraction device that from above-mentioned document image, extracts character; In the character that selection is extracted by above-mentioned character extraction device, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character gets the choice device of the character of predetermined structure; And extraction is according to the watermark information draw-out device of the watermark information of the determining positions of the character radicals by which characters are arranged in traditional Chinese dictionaries of the character of above-mentioned choice device selection.
In addition, in order to reach purpose of the present invention, image processing method of the present invention may further comprise the steps.
That is, a kind of image processing method to the document image embed watermark information that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is characterized in that comprising: the extraction step that extracts character from above-mentioned document image; The selection step of be chosen in the character that extracts in the above-mentioned extraction step, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character being got the character of predetermined structure; And the position of the predetermined character radicals by which characters are arranged in traditional Chinese dictionaries by making the character in above-mentioned selection step, selected with respect to this character in unchanged other the character radicals by which characters are arranged in traditional Chinese dictionaries in position be changed to position according to the watermark information decision, the embedding step of embed watermark information in the character of above-mentioned selection.
In addition, in order to reach purpose of the present invention, image processing method of the present invention may further comprise the steps.
That is, a kind of the document image that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is extracted the image processing method of the watermark information that is embedded into, it is characterized in that comprising: the character extraction step that from above-mentioned document image, extracts character; The selection step of be chosen in the character that extracts in the above-mentioned character extraction step, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character being got the character of predetermined structure; And extraction is according to the watermark information extraction step of the watermark information of the determining positions of the character radicals by which characters are arranged in traditional Chinese dictionaries of the character of selecting in above-mentioned selection step.
Description of drawings
Fig. 1 is the block diagram of basic structure of electronic watermark embedded device that document image is carried out the embedding of eletric watermark of expression first form of implementation of the present invention.
Fig. 2 is that expression has embedded the block diagram of the basic structure of the eletric watermark draw-out device that extracts eletric watermark the document image of eletric watermark from utilizing electronic watermark embedded device shown in Figure 1.
Fig. 3 is the flow chart that processor 4 embeds eletric watermark the processing in the document image.
Fig. 4 is processor 24 extracts eletric watermark from 1 document image that embeds eletric watermark a process chart.
Fig. 5 is the detail flowchart of the processing among step S206, the S306.
Fig. 6 is the schematic diagram that the embedding processing of the file access control information of carrying out in step S412 is described.
Fig. 7 is the figure of explanation with the processing in 9 the actual embedding of the information file.
Fig. 8 is the figure that expression has a plurality of Chinese modes of the above radicals by which characters are arranged in traditional Chinese dictionaries of some.
Fig. 9 be explanation by the space amount between the control English character string, carry out the figure of method of the embedding of information.
Figure 10 is the rotation amount of explanation by control character, carries out the figure of method of the embedding of information.
Figure 11 is the amplification reduction volume of explanation by control character, carries out the figure of method of the embedding of information.
Figure 12 is the figure of expression embedding information example.
Figure 13 is that the schematic diagram of handling according to the embedding of the file access control information of the fiducial value of trying to achieve with the first form of implementation diverse ways is used in explanation.
Embodiment
Below, with reference to accompanying drawing, describe the present invention in detail according to preferred implementing form.
[first form of implementation]
Fig. 1 is the block diagram of basic structure of electronic watermark embedded device that document image is carried out the embedding of eletric watermark of this form of implementation of expression.
Among this figure, the 2nd, input embed eletric watermark obj ect file by scanner, camera, or the input part that constitutes such as file reader unit, the 4th, the processor that carries out various processing, the 6th, to the keyboard of processor 4 input commands, the 8th, preserve embedding information or the dish of the document image that reads in, the 10th, temporarily store data etc. in order in processor 4, to carry out various processing, or the memory of the document image that reads in by input part 2 of storage, the 12nd, show the order input that processor 4 is carried out and the display of treatment state, the 14th, output embedded access control information document image by printer, or internet, the efferent that network interfaces such as LAN constitute.
On the other hand, Fig. 2 is that expression has embedded the basic block diagram that extracts the eletric watermark draw-out device of eletric watermark the document image of eletric watermark from utilizing electronic watermark embedded device shown in Figure 1.
In the figure, the 22nd, the input embedded eletric watermark file by scanner, camera, or file reader unit, the input part that network interface etc. constitute, the 24th, the processor that carries out various processing, the 26th, to the keyboard of processor 24 input commands, the 28th, preserve the document image that reads in, or the dish of the original document of the file that reads in retrieval usefulness, the 30th, temporarily store data etc. in order in processor 24, to carry out various processing, or the memory of the document image that reads in by input part 22 of storage, the 32nd, show the order input that processor 24 is carried out and the display of treatment state, 34,36 is respectively the network interface that the utilization file access control information of reading in is used, printer.
In addition, in this form of implementation, though electronic watermark embedded device and eletric watermark draw-out device are used as independent device separately, but be not limited thereto, also these can be installed (electronic watermark embedded device, eletric watermark draw-out device) and use as the eletric watermark Embedded Division in the device, eletric watermark extracting part.
Below, illustrate that eletric watermark embeds the rough flow process of handling.At first,, obtain the electronic document image that is embedded into, in memory 10, launch from input part 2 according to order from keyboard 6 inputs.From keyboard 6 or coil 8 input embedding information (file access control information), this information is embedded in the memory 10 in the document image that launches again by processor 4.The document image that has embedded predetermined file access control information is exported as the file that embeds eletric watermark from efferent 14.
Below, the rough flow process that extracts the processing of eletric watermark from the file that embeds eletric watermark of efferent 14 outputs is described.At first,, import the file that has embedded eletric watermark, in memory 30, launch by input part 22 according to order from keyboard 26 inputs.Secondly by the document image of processor 24, read the file access control information of embedding, the processing of being scheduled to according to its indication from memory 30, launching.So-called predetermined processing, be for example to find under the wrongful situation about reading, to outside circular, to inner disk 28 or the outside is carried out the retrieval of original document, perhaps printout attribute information etc. uses network I/F34, printer 36 in order to carry out these processing.
Below, describe processor 4 in detail with the processing method in document image of eletric watermark embedding.The flow process of this processing has been shown among Fig. 3.
In step S200, read in file from input part 2, give memory 10 as the electronic image transfer of data.In addition in this step, pre-treatments such as the direction of the file that reads in, tilt correction.In step S202, the document image that launches in memory 10 in step S200 is carried out zone identification, characters in images piece (text) is all extracted.This work can application examples such as the Japanese Patent Application Publication spy open the piece selection technology of putting down in writing in the flat 6-068301 communique and wait and realize.In step S204, the character to comprising in the alphabet piece that extracts in step S202 carries out character recognition, generates the character code as character identification result.
In step S206, in the character that from the character block that among step S202, extracts, comprises, extract the object character that embeds the file access control information.The object character of supposing extraction is the character of the word size that is predetermined.The back will explain the processing method in this step.In step S208, the file access control information that embeds in the character that input is extracted in step S206.Here, the called file access control information for example is a copy limit information, distort the information of preventing, original document management information etc.
In step S210, the file access control information that will import in step S208 is embedded in the character that extracts among the step S206.With the processing that describes in detail in the back in this step.In step S212, output has embedded the document image of file access control information in step S210.
Below, describe processor 24 extracts eletric watermark from a document image that has embedded eletric watermark processing method in detail.This handling process is shown among Fig. 4.
In step S300, be taken into the file that embeds eletric watermark from input part 22, give memory 30 as the electronic image data delivery.Processing in this step is identical with step S200, also comprises the pre-treatment such as direction, tilt correction of the file that reads in.
In step S302, the document image that embeds eletric watermark that launches in memory 30 in step S300 is carried out zone identification, the character block in the document image is all extracted.Processing in similarly carry out this step with the processing among the step S202.In step S304, the alphabet piece to extracting in step S302 carries out character recognition.Processing in similarly carry out this step with the processing among the step S204.
In step S306, in the character that from the character block that among step S302, extracts, comprises, only extract the character that embeds the file access control information.The back will explain the processing method in this step.In step S308, from the character that among step S306, extracts, read the file access control information.The back will explain the processing method in this step.
In step S310, according to the file access control information of in step S308, reading, carry out expectant control, for example duplicate and forbid processing, document retrieval processing etc.
Fig. 5 is the detail flowchart of the processing among step S206～S210 and the S306.In step S400, will flow to the character extraction working storage in the memory 10 based on the character code of character identification result.In step S402, judge whether that the alphabet sign indicating number that will comprise in the file has flowed to the character extraction working storage.Carrying under the situation about being all over, transfer among the step S404 and handle, under situation about not being all over, transfer among the step S400 and handle.
In step S404, use to flow to the character code of character extraction, to the predefined character of each character count with working storage.Here, what is called preestablishes, and is to preestablish the employing Chinese character of complicated radicals by which characters are arranged in traditional Chinese dictionaries structure in a way, for example has the word size " the formation radicals by which characters are arranged in traditional Chinese dictionaries are the Chinese character more than 3 " of 10 points such.In other words, in step S404, counting is fed to the number of character extraction with character code identical with the character code of predefined character in the character code of working storage.By carrying out such setting, can not embed the above information of some boldly and reliably.The back will describe this point in detail.
In step S406,, classify by the character that counting is counted in step S404.In step S408, judge whether counts reaches to a certain degree more than, promptly whether the character that occurrence frequency is high in the file reaches more than certain number of times.This is the embedding precision in order to ensure eletric watermark, gets the above character as object of certain number of times, and same information is embedded in the same character and the measure of taking repeatedly.This also is the extraction precision in order to ensure eletric watermark in addition.Here more than said certain number of times, though the many more precision of number of times are high more, even for example also can for twice.The back will describe this point in detail.
Here, be judged as under the situation that does not have certain above object number of characters, be judged as and embed/extract predetermined amount of information, handling and transfer to step S414, otherwise step S410 is transferred in processing.
In step S410, the character of occurrence frequency maximum in the select File in embedding/extraction object character is calculated and is used to embed/fiducial value of extraction operation.The back will describe this fiducial value in detail.
In step S412, except the character of having obtained fiducial value, from sorting result among step S406, obtaining above-mentioned occurrence frequency is more than second later characters, the embedding/extraction operation of carrying out the file access control information.The back will illustrate concrete method.
Step S414, be judged as that in step S408 embedding/extraction object character is few, under the situation that can not embed/extract, the step of the processing of being scheduled to.So-called predetermined processing is the processing that shows the warning that for example can not embed/extract etc. on display 12 or 32.
Fig. 6 is the schematic diagram that the embedding processing of the file access control information of carrying out in step S412 is described, Fig. 7 illustrates the figure that in fact 9 information is embedded the processing in the file.Fig. 6 is the figure of method that explanation is asked the method for fiducial value and embedded 3 information (8 kinds of information) respectively.In the following description, though the file access control information is decided to be 9 information, be not limited to this.In addition, in Fig. 6,7, illustrate that the number of radicals by which characters are arranged in traditional Chinese dictionaries (character radicals by which characters are arranged in traditional Chinese dictionaries) is for example 3, and for example " type " is such, little radicals by which characters are arranged in traditional Chinese dictionaries have two on top, and big radicals by which characters are arranged in traditional Chinese dictionaries have the pattern of one character in the bottom.In addition, describe figure to such an extent that more or less exaggerate for explanation.
At first, the character picture (being the image of Chinese character " type " among Fig. 6) that will extract from character block in step S206 resolves into each radicals by which characters are arranged in traditional Chinese dictionaries, asks its fiducial value.Method as character being resolved into each radicals by which characters are arranged in traditional Chinese dictionaries is not particularly limited, so adopt general disclosed method to get final product.So-called fiducial value is a most important value when embedding in the file with the invisible form of naked eyes the file access control information.Here said fiducial value as defining among Fig. 6, is sat up straight target relative distance K, P, M, N with four of character picture and is represented.
Concrete information embedding method considers to use four fiducial value K, P, M, the N that defined just now here, embeds 3 information at each character.In step S410, obtain fiducial value K, P the reliability maximum, that be the highest character of occurrence frequency, M, N (being equivalent to the 3rd step among Fig. 7).Corresponding, the file access control information that preparation should embed (9), by per 3,, carry out shown in Figure 6 certain and handle (processing of the relative position of each radicals by which characters are arranged in traditional Chinese dictionaries of change character) (being equivalent to the 4th step among Fig. 7) the more than second later characters of in step S412, selecting.Specifically, when embedding per 3 information in more than the second later characters, for example in the character with initial 3 embeddings (classification results) more than second.In the character with next 3 embeddings more than the 3rd, in the character with 3 last embeddings more than the 4th., be not limited to this order, also can carry out on the contrary, that is, for example in the character with 3 initial embeddings more than the 4th.In the character with next 3 embeddings more than the 3rd, in the character with 3 last embeddings more than second.
Generally speaking, the information (embedding information) that per 3 information has been embedded in which many character deposits in the memory 10.The example of the information of this embedding has been shown among Figure 12.In the figure, embedding information 1201 is deposited in the memory 10, and embedding information 1201 is made of the classified order that has embedded 3 initial character, the classified order that has embedded second 3 character, the information of classified order that has embedded the 3rd 3 character.
In the extraction of eletric watermark is handled,, during from per 3 information of each character extraction, can specificly which kind of rearrange these per 3 information in proper order according to original 9 file access control information is restored by information with reference to this embedding.The back will explain the extraction of eletric watermark and handle.
When embedding per 3 information in each character, as mentioned above, making the change in location of the radicals by which characters are arranged in traditional Chinese dictionaries of character, the pattern of this variation according to the information that embeds is a certain corresponding in the changing pattern of each information shown in Figure 6 (000,001,010,011,100,101,110,111).
In addition, from above explanation as can be known,, can keep the embedding/extraction precision of eletric watermark and the balance that embeds figure place by adjusting the number (minimum occurrence frequency) of object character.
K ' among Fig. 6, P ', M ', N ' are the relative distance of four ends after the change in location.
In order to prevent the deterioration of character, make maximum radicals by which characters are arranged in traditional Chinese dictionaries, be that the radicals by which characters are arranged in traditional Chinese dictionaries of lower end do not change in the case.Order according to above explanation can embed information arbitrarily.
On the other hand, same as the above-mentioned method when extracting eletric watermark, obtain fiducial value, be that the relative position of each radicals by which characters are arranged in traditional Chinese dictionaries of the character more than second below is compared with fiducial value with occurrence frequency, extract an arrangement that embeds in each character.In addition at this moment as mentioned above owing to embed information stores in memory 10, so with reference to this embedding information, the embedding of specific extraction the character arranged of position be which character, recover original file access control information.
By above explanation as can be known, adopt the image processing apparatus and the image processing method of this form of implementation, on the basis of carrying out zone identification, character recognition, use the relative position of each complicated character radicals by which characters are arranged in traditional Chinese dictionaries in the radicals by which characters are arranged in traditional Chinese dictionaries structure to change dexterously, can guarantee that simultaneously to a certain degree above information embeds precision, quantity (can use according to the classification of occurrence frequency and be controlled) with the degradation inhibiting of font in Min..In addition, when extracting eletric watermark, also can realize the eletric watermark that the noise resistance performance is high.In addition, on principle, there is not the dependence of word size fully, so, obviously be a kind of effective method even for the few original copy of character quantity yet.
[second form of implementation]
In first form of implementation, the radicals by which characters are arranged in traditional Chinese dictionaries structure that adopts as embedded object is a single pattern shown in Figure 6, but is not limited thereto, and as shown in Figure 8, also can set a plurality of Chinese modes with the above radicals by which characters are arranged in traditional Chinese dictionaries of some simultaneously.In the case, can adopt the method for using in first form of implementation, further increase the embedding amount of information character with various radicals by which characters are arranged in traditional Chinese dictionaries structures.
[the 3rd form of implementation]
In first form of implementation, each character has embedded 3, but is not limited thereto, if the figure place in the possible combination of the Move Mode of radicals by which characters are arranged in traditional Chinese dictionaries then can freely be set.But, embed figure place if increase, then the deformation extent of character increases.
[the 4th form of implementation]
In first form of implementation, watermark information has been embedded in the Chinese character, but be not to be defined in this, if the character that constitutes by a plurality of structural elements (character radicals by which characters are arranged in traditional Chinese dictionaries), for example then can fully similarly embed in Korea S's literal, the Thailand's literal etc.
[the 5th form of implementation]
Use fiducial value for example shown in Figure 6, under the situation in " 000 " bit string embedding Chinese character " type ", make N '=N, but under the medium situation that has an interference effect of the superimposed document image that has embedded bit string of noise, the processing of from the Chinese character " type " that has embedded this bit string, extracting bit string " 000 " difficulty that will become.This be because, even in extract handling, ask N ', be not N '=N strictly sometimes, its result often can not extract bit string " 000 ".
Therefore, in extracting processing, can have certain width ground and change by comparing N ' and N.If promptly satisfy | N '-N|＜ε then is judged as N '=N.This processing also can be applicable to other fiducial values, if other is for example satisfied | and M '-M|＜ε then is judged as M '=M.
[the 6th form of implementation]
In first form of implementation, the object character as embedding the file access control information has adopted the character that identical word size is arranged.This is because by the file access control information is embedded in each object character, makes the amount of movement of mobile radicals by which characters are arranged in traditional Chinese dictionaries certain, and radicals by which characters are arranged in traditional Chinese dictionaries move the back makes character between each object character the roughly certain cause of balance.
, even the word size of each object character is inequality, if but each word size is pre-determined the amount of movement of mobile radicals by which characters are arranged in traditional Chinese dictionaries, also can move the back and between each object character, make the balance of character roughly certain at radicals by which characters are arranged in traditional Chinese dictionaries.In addition, also can when embedding, ask the amount of movement of the radicals by which characters are arranged in traditional Chinese dictionaries of each word size.In the case, for example suppose in the character of 10 points that the amount of movement of radicals by which characters are arranged in traditional Chinese dictionaries is c, then under the situation of 12 points, can obtain amount of movement by calculating (c * (12 character sizes)/(10 character sizes)).
[the 7th form of implementation]
In first form of implementation, as the character of asking fiducial value to use, adopted the character of occurrence frequency maximum, but be not limited to this, also can be in each radicals by which characters are arranged in traditional Chinese dictionaries pattern for example, in advance according to making of stroke number, radicals by which characters are arranged in traditional Chinese dictionaries etc., even be divided into visually also inconspicuous group of mobile radicals by which characters are arranged in traditional Chinese dictionaries and showy group, set as embedded object character, benchmark character respectively, with the character in a plurality of benchmark character group that in fact occur as the benchmark character.The fiducial value of this situation can be passed through, if for example by mobile radicals by which characters are arranged in traditional Chinese dictionaries then each character that comprises in the visually showy group asks the mean value of fiducial value to obtain.
[the 8th form of implementation]
In first form of implementation, the character as asking fiducial value to use has adopted the character of occurrence frequency maximum, but has been not to be defined in this, even for example adopt the initial character of file or character block also passable.In this case, when initial character and predetermined radicals by which characters are arranged in traditional Chinese dictionaries pattern are inconsistent, control the feasible for example character late that adopts.
[the 9th form of implementation]
In first form of implementation, the character as asking fiducial value to use has adopted the character of occurrence frequency maximum, but has been not to be defined in this, for example also can be with occurrence frequency more than second with down to the inferior character of pre-determined bit as the benchmark character.The fiducial value of this situation, the mean value of fiducial value that can be by asking each benchmark character obtains.
[the tenth form of implementation]
In first form of implementation, shown in the definition among Fig. 6, sit up straight target relative distance K, P, M, N with four of character picture and represent fiducial value K, P, M, N.But be not limited to this, figure 13 illustrates other examples of fiducial value.
Figure 13 is that the figure according to the embedding handling principle of the file access control information of the fiducial value of trying to achieve with the first form of implementation diverse ways is used in explanation.Fiducial value in this form of implementation with the decomposition of character picture each radicals by which characters are arranged in traditional Chinese dictionaries width, highly the ratio of absolute altitude, width is represented.If further specify, that is exactly said here each radicals by which characters are arranged in traditional Chinese dictionaries, and top has only two radicals by which characters are arranged in traditional Chinese dictionaries, does not define its magnitude proportion about the radicals by which characters are arranged in traditional Chinese dictionaries of bottom.This is that then relative fiducial value can change because if define with whole radicals by which characters are arranged in traditional Chinese dictionaries, and in addition, the distortion of word is also big, and it is very showy that deterioration becomes, so surplus so next topmost radicals by which characters are arranged in traditional Chinese dictionaries, (in the case) only uses other two radicals by which characters are arranged in traditional Chinese dictionaries.
Then, similarly use fiducial value K, P, M, the N that tries to achieve like this, carry out the embedding of file access control information and handle with first form of implementation.
As mentioned above, adopt the present invention, can guarantee that to a certain degree above information embeds precision, embeds quantity with the degradation inhibiting of font in Min..
Claims (26)
1. image processing apparatus is to the image processing apparatus of the document image embed watermark information that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries, it is characterized in that comprising:
From above-mentioned document image, extract the draw-out device of character;
In the character that selection is extracted by above-mentioned draw-out device, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character gets the choice device of the character of predetermined structure; And
The position of the predetermined character radicals by which characters are arranged in traditional Chinese dictionaries by making the character of being selected by above-mentioned choice device is changed to position according to the watermark information decision, the flush mounting of embed watermark information in the character of above-mentioned selection with respect to unchanged other the character radicals by which characters are arranged in traditional Chinese dictionaries in position in this character.
2. image processing apparatus according to claim 1 is characterized in that:
Above-mentioned draw-out device also comprises
From above-mentioned document image, extract the character block draw-out device of character block; And
The character that contains in the character block that is extracted by above-mentioned character block draw-out device is carried out character recognition, generate character code, from above-mentioned character block, extract the character recognition device of the image of above-mentioned character as recognition result.
3. image processing apparatus according to claim 1 and 2 is characterized in that:
Above-mentioned choice device comprise in the character that extracts by above-mentioned draw-out device, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character gets the counting device that the number of the character of predetermined structure is counted by every kind of character,
More than the counts of being undertaken by above-mentioned counting device reached predetermined counts, promptly the number of character was that above-mentioned flush mounting embeds above-mentioned watermark information in the character of being selected by above-mentioned choice device under the above situation of some.
4. image processing apparatus according to claim 3 is characterized in that:
Above-mentioned counting device is got the number of the character of predetermined structure and is used the character code of this character to count by every kind of character to the character radicals by which characters are arranged in traditional Chinese dictionaries of character.
5. image processing apparatus according to claim 3 is characterized in that:
Under the little situation of the predetermined counts of the counts ratio that is undertaken by above-mentioned counting device, on predetermined display unit, show the warning of the impossible embedding of eletric watermark.
6. image processing apparatus according to claim 4 is characterized in that:
Under the little situation of the predetermined counts of the counts ratio that is undertaken by above-mentioned counting device, on predetermined display unit, show the warning of the impossible embedding of eletric watermark.
7. image processing apparatus according to claim 1 is characterized in that:
The calculation element that also comprises the fiducial value that calculating is used when embedding above-mentioned watermark information.
8. image processing apparatus according to claim 7 is characterized in that:
The aforementioned calculation device calculates the said reference value with predetermined character.
9. according to claim 7 or 8 described image processing apparatus, it is characterized in that:
The external relative distance of target sat up straight in external four of aforementioned calculation device calculating selecteed character in order to calculate fiducial value, as the said reference value.
10. according to claim 7 or 8 described image processing apparatus, it is characterized in that:
The aforementioned calculation device calculates the width of each character radicals by which characters are arranged in traditional Chinese dictionaries of selecteed character in order to calculate fiducial value, highly to the width of this character, the ratio of height, as the said reference value.
11. image processing apparatus according to claim 1 is characterized in that:
Above-mentioned flush mounting also generates expression above-mentioned watermark information is embedded information in which character.
12. an image processing apparatus is the image processing apparatus that the document image that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is extracted the watermark information that is embedded into, and it is characterized in that comprising:
From above-mentioned document image, extract the character extraction device of character;
In the character that selection is extracted by above-mentioned character extraction device, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character gets the choice device of the character of predetermined structure; And
Extraction is according to the watermark information draw-out device of the watermark information of the determining positions of the character radicals by which characters are arranged in traditional Chinese dictionaries of the character of above-mentioned choice device selection.
13. image processing apparatus according to claim 12 is characterized in that:
Above-mentioned character extraction device also comprises
From above-mentioned document image, extract the character block draw-out device of character block; And
The character that contains in the character block that is extracted by above-mentioned character block draw-out device is carried out character recognition, generate character code, from above-mentioned character block, extract the character recognition device of the image of above-mentioned character as recognition result.
14. image processing apparatus according to claim 12 is characterized in that:
Above-mentioned choice device comprise in the character that extracts by above-mentioned character extraction device, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character gets the counting device that the character number of predetermined structure is counted by every kind of character,
More than the counts of being undertaken by above-mentioned counting device reached predetermined counts, promptly the number of character was that above-mentioned character extraction device extracts above-mentioned watermark information under the above situation of some from the character of being selected by above-mentioned choice device.
15. image processing apparatus according to claim 14 is characterized in that:
Above-mentioned counting device is got the number of the character of predetermined structure and is used the character code of this character to count by every kind of character to the character radicals by which characters are arranged in traditional Chinese dictionaries of character.
16., it is characterized in that according to claim 14 or 15 described image processing apparatus:
Under the little situation of the predetermined counts of the counts ratio that is undertaken by above-mentioned counting device, on predetermined display unit, show the warning of the impossible extraction of eletric watermark.
17. image processing apparatus according to claim 12 is characterized in that:
Above-mentioned watermark information draw-out device extracts above-mentioned watermark information by with reference to expression above-mentioned watermark information having been embedded information in which character.
18. image processing apparatus according to claim 12 is characterized in that:
The calculation element that also comprises the fiducial value of using when calculate extracting above-mentioned watermark information.
19. image processing apparatus according to claim 18 is characterized in that:
The aforementioned calculation device calculates the said reference value with predetermined character.
20., it is characterized in that according to claim 18 or 19 described image processing apparatus:
The external relative distance of target sat up straight in external four of aforementioned calculation device calculating selecteed character in order to calculate fiducial value, as the said reference value.
21., it is characterized in that according to claim 18 or 19 described image processing apparatus:
The aforementioned calculation device calculates the width of each character radicals by which characters are arranged in traditional Chinese dictionaries of selecteed character in order to calculate fiducial value, highly to the width of this character, the ratio of height, as the said reference value.
22. image processing apparatus according to claim 1 is characterized in that:
Above-mentioned watermark information is the file access control information, comprises copy limit information, distorts the information of preventing, original document management information.
23. image processing apparatus according to claim 1 is characterized in that:
The character that comprises in the above-mentioned document image comprises Chinese character, Korea S's literal, Thailand's literal.
24. image processing apparatus according to claim 1 is characterized in that:
Above-mentioned character radicals by which characters are arranged in traditional Chinese dictionaries comprise the radicals by which characters are arranged in traditional Chinese dictionaries of Chinese character.
25. an image processing method is to the image processing method of the document image embed watermark information that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries, it is characterized in that comprising:
From above-mentioned document image, extract the extraction step of character;
The selection step of be chosen in the character that extracts in the above-mentioned extraction step, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character being got the character of predetermined structure; And
The position of the predetermined character radicals by which characters are arranged in traditional Chinese dictionaries by making the character selected in above-mentioned selection step is changed to position according to the watermark information decision, the embedding step of embed watermark information in the character of above-mentioned selection with respect to unchanged other the character radicals by which characters are arranged in traditional Chinese dictionaries of this character position.
26. an image processing method is the image processing method that the document image that comprises the character that a character is made of a plurality of character radicals by which characters are arranged in traditional Chinese dictionaries is extracted the watermark information that embeds, and it is characterized in that comprising:
From above-mentioned document image, extract the character extraction step of character;
The selection step of be chosen in the character that extracts in the above-mentioned character extraction step, the character radicals by which characters are arranged in traditional Chinese dictionaries structure of character being got the character of predetermined structure; And
Extraction is according to the watermark information extraction step of the watermark information of the determining positions of the character radicals by which characters are arranged in traditional Chinese dictionaries of the character of selecting in above-mentioned selection step.
Applications Claiming Priority (6)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2001266436 | 2001-09-03 | ||
| JP266436/2001 | 2001-09-03 | ||
| JP386137/2001 | 2001-12-19 | ||
| JP2001386137A JP3848150B2 (en) | 2001-12-19 | 2001-12-19 | Image processing apparatus and method |
| JP226588/2002 | 2002-08-02 | ||
| JP2002226588A JP3833154B2 (en) | 2001-09-03 | 2002-08-02 | Image processing apparatus, image processing method, program, and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1404298A CN1404298A (en) | 2003-03-19 |
| CN1226860C true CN1226860C (en) | 2005-11-09 |
Family
ID=27347433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB021419965A Expired - Fee Related CN1226860C (en) | 2001-09-03 | 2002-09-02 | Image processing apparatus and image processing method and program and storage media |
Country Status (2)
| Country | Link |
|---|---|
| KR (1) | KR100485554B1 (en) |
| CN (1) | CN1226860C (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1326383C (en) * | 2004-06-30 | 2007-07-11 | 佳能株式会社 | Image processing apparatus, image processing method, computer program and computer readable storage medium |
| JP2006050551A (en) | 2004-06-30 | 2006-02-16 | Canon Inc | Image processing apparatus, image processing method, program and storage medium |
| CN1684115B (en) * | 2004-10-18 | 2011-03-23 | 刘�东 | Text digital water printing technology based on character topoloical structure |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3573009B2 (en) * | 1999-08-11 | 2004-10-06 | 日本電気株式会社 | Digital watermark insertion system, digital watermark characteristic table generation system, and digital watermark characteristic parameter table generation system |
| JP3643509B2 (en) * | 1999-09-30 | 2005-04-27 | 株式会社東芝 | Digital watermark embedding method and apparatus, and digital watermark detection method and apparatus |
| KR20010070865A (en) * | 2001-06-14 | 2001-07-27 | 최종욱 | Apparatus for preventing duplication and forgery/alternation of document and authenticating it |
-
2002
- 2002-09-02 CN CNB021419965A patent/CN1226860C/en not_active Expired - Fee Related
- 2002-09-03 KR KR10-2002-0052878A patent/KR100485554B1/en not_active IP Right Cessation
Also Published As
| Publication number | Publication date |
|---|---|
| KR100485554B1 (en) | 2005-04-27 |
| CN1404298A (en) | 2003-03-19 |
| KR20030020250A (en) | 2003-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101609283B (en) | Image processing apparatus and image processing method | |
| CN1195280C (en) | Method and system for inserting information into piles | |
| JP5253352B2 (en) | Method for embedding a message in a document and method for embedding a message in a document using a distance field | |
| JP4310288B2 (en) | Image processing apparatus and method, program, and storage medium | |
| CN1269069C (en) | Symbol identifying device and method | |
| JP4758461B2 (en) | Text direction determination method and system in digital image, control program, and recording medium | |
| US7411702B2 (en) | Method, apparatus, and computer program product for embedding digital watermark, and method, apparatus, and computer program product for extracting digital watermark | |
| JP2009003937A (en) | Method and system for identifying text orientation in digital image, control program and recording medium | |
| CN1719865A (en) | Image processing system and image processing method | |
| CN1933536A (en) | Adaptive, image content dependent mark placement | |
| JP2007221794A (en) | Method and apparatus for creating high-fidelity glyph prototype from low-resolution glyph images | |
| CN1955981A (en) | Character recognition device, character recognition method and character data | |
| CN1704990A (en) | Information embedding device, information detecting device, information embedding and detecting system, information embedding method, information detecting method, information embedding program, infor | |
| CN1226860C (en) | Image processing apparatus and image processing method and program and storage media | |
| CN1945622A (en) | Digital water mark embedding and extracting method and device | |
| JP3980983B2 (en) | Watermark information embedding method, watermark information detecting method, watermark information embedding device, and watermark information detecting device | |
| CN1771513A (en) | Method of detecting watermarks | |
| US8126193B2 (en) | Image forming apparatus and method of image forming | |
| EP2529331B1 (en) | Parallel test payload | |
| CN1275191C (en) | Method and appts. for expanding character zone in image | |
| CN1497525A (en) | Technique for setting printing width of outline character | |
| Davarzani et al. | Farsi text watermarking based on character coding | |
| CN1084010C (en) | Word generating device | |
| JP2005157928A (en) | Stamp processor, electronic approval system, program and recording medium | |
| US8155376B2 (en) | Image processor, image forming apparatus, image processing method and computer readable medium |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20051109 Termination date: 20140902 |
|
| EXPY | Termination of patent right or utility model |