CN101393414B - Image data output processing apparatus and image data output processing method - Google Patents

Abstract

In an image data output processing apparatus of the present invention, an image matching section is capable of determining whether a similarity exists between each image of an allocation document and a reference document when input image data is indicative of the allocation document. An output process control section is capable of regulating an output process of each image in accordance with a result of determining whether the similarity exists between each image of the allocation document and the reference document. Thus, it is able to detect with high accuracy a document image under regulation on the output process and regulating the output process, when the input image data is indicative of an allocation document and includes the document image under regulation on the output process.

Description

Image data output processing apparatus and method thereof

Technical field

The present invention relates to the image data output processing apparatus that the outputs such as replication processes, transmission processing, editing and processing or filing processing to input image data are handled, more specifically, relate to have to the image of input image data and in advance the image of the registration original copy of registration compare and judge similarity, and based on this result of determination, the image data output processing apparatus of the function that the output of control input image data is handled.

Background technology

Following image contrast technology has just been arranged in the past, and promptly the view data of the registration original copy of view data that obtains reading original copy with scanner and prior registration compares, and judges its homophylic image contrast technology.

In homophylic judgement, known: as for example to use OCR (Optical Character Reader) etc. to extract key word, and carry out method based on the coupling of key word from image; Object images is defined as the image of bill with lines, carries out method (with reference to patent documentation 1) based on the coupling of the feature of lines; Perhaps based on the tone method (with reference to patent documentation 2) that distributes and judge of input picture and registration image.

In addition, in patent documentation 3, disclose following technology, promptly the feature from the input document generates descriptor, use this descriptor and record descriptor and point out to comprise the descriptor database of tabulation of the document of the feature that descriptor generates, import the technology of the coupling between the document in document and the document database.Descriptor is selected as, and becomes constant with respect to the difference between the document of the coupling in the distortion that is produced by the digitizing of document or input document and the document database.

In this technology, scanning is during descriptor database, and accumulative total is to the ballot of each document in the document database, and 1 document or the number of votes obtained of the highest number of votes obtained surpassed the document of certain threshold value as the coupling document.

In addition, in patent documentation 4, disclose following technology, promptly extract a plurality of unique points from digital picture, to each unique point of extracting, the set of the unique point that decision is local, select the part set of unique point from each set of decision, with the each several part set selected amount as supplementary features, and based on a plurality of combinations of the unique point in the part set, obtain invariant respectively, each invariant that combination is obtained, and calculated characteristics amount to geometric transformation, or/and image is voted, retrieval looks like corresponding document or/and the technology of image with above-mentioned digitized map 1 thus to the document in the database of characteristic quantity with calculating.

And, in the past, in the image data output processing apparatus that outputs such as duplicating machine, facsimile unit, scanister or compounding machine etc. carry out that to input image data replication processes, transmission processing, editing and processing or filing are handled are handled, utilize such original copy contrast technology, when input image data is similar to the view data of registration original copy, limits its output and handle.

For example, known in coloured image formation device, in order to tackle the forgery of bank note and securities, based on the pattern that detects from input image data, judge whether input image data is the view data of bank note and securities (registration original copy), when being the view data of registration image, at the additional specific pattern of the image of being exported, perhaps be coated with full duplicating image, perhaps forbid replication actions, so that can determine the technology of the image processing system that duplicates from the image of being exported.

In addition, in patent documentation 5, disclose as technology, the quantity of the page or leaf image that promptly comprises in the source images based on the comparison, the character features amount that extraction comprises image feature amount such as the color characteristic amount, brightness characteristic quantity, architectural feature amount, shape facility amount of the page or leaf treatment of picture subject area that comprises in the reference source image or is made of character code, use the characteristic quantity that is extracted, retrieval comparison object image extracts original electronic data thus from the paper spare document that Nup prints.

Patent documentation 1: Japanese publication communique spy opens flat 8-255236 communique (on October 1st, 1996 is open)

Patent documentation 2: Japanese publication communique spy opens flat 5-110815 communique (on April 30th, 1993 is open)

Patent documentation 3: Japanese publication communique spy opens flat 7-282088 communique (October 27 nineteen ninety-five is open)

Patent documentation 4: international No. 2006/092957 pamphlet (on September 8th, 2006 is open) that disclose of Japanese publication communique

Patent documentation 5: Japanese publication communique spy opens 2005-4724 communique (on January 6th, 2005 is open)

Yet in image data output processing apparatus in the past, input image data is when distributing original copy, even existence includes the image that the original copy of handling is exported in restriction in distributing original copy, also can't differentiate it, and the problem that the output that permission should limit is handled.

Distributing original copy is the original copy that distributes a plurality of original images on an original copy, the Nin1 that for example concentrates N to open original image on an original copy to form (N=2,4,6,8,9 etc.) original copy etc.

In image data output processing apparatus in the past, distribute the image of original copy also processed, and judge and register the similarity of original copy as an original image.For example, input image data is that 2 of original copy A and the original copy B that be assigned shown in Figure 26 (a) closes 1 original copy, also as an original image, comes the characteristic quantity of computed image, and judges and register the similarity of original copy.

Therefore, even original copy A and original copy B both are the registration original copys, also shown in the example of Figure 26 (b), the characteristic quantity that has with the original copy A that is registered is consistent a lot, and when number of votes obtained surpassed threshold value, being judged to be this, 2 to close 1 original copy be original copy A, about the characteristic quantity that original copy B has, do not judge similarity.

Its result, the restriction of original copy A when reducing that the exploring degree is exported and wait the restriction of pine, even original copy B is an original copy of forbidding exporting processing, is also handled together, can't avoid the information outflow of original copy B.

And then in above-mentioned example, expression 2 is closed the number of votes obtained of original copy A of a side in 1 original copy greater than the situation of threshold value.Yet, the result of ballot, 2 original copy A, the original copy B that close 1 original copy are below the threshold value, also can be judged as the situation that is not the registration original copy.At this moment, be not subjected to any restriction, intactly original copy A, original copy B are handled in output.

Otherwise in addition, as shown in figure 27, the number of votes obtained that also has original copy A, an original copy B is all greater than the situation of threshold value, at this moment, can be judged to be and be similar to original copy A, original copy B, 2 not close 1 original copy but can not be judged as to be similar to, so be similar to the highest original copy A of number of votes obtained if be judged as by what original copy A, original copy B constituted, then, just can't limit any its output for original copy B.Promptly with above-mentioned same, input image data is when distributing 2 of original copy A and original copy B to close 1 original copy, to be similar to the original copy A that allows output if be judged as, even then original copy B is an original copy of forbidding exporting processing, also handled together, can't avoid the information of original copy B to flow out.

When distributing original copy, the original image that is comprised is than original size reduction.In the image that dwindles, the characteristic quantity and the different situation of registration original copy of not dwindling of calculating appears sometimes, judge that precision reduces.

If illustrate in greater detail, then in the gimmick based on unique point calculated characteristics amount, the connecting area of calculated characteristics point is different sometimes in the distribution original copy that dwindles and the registration original copy that does not dwindle.For example, in the registration original copy, be separated from each other, the parts of images that is identified as different connecting areas closes to be close in 1 original copy at the N that dwindles and is identified as 1 connecting area together.

As described above, if the connecting area difference of calculated characteristics point then calculates the characteristic quantity different with registering original copy, thus, as mentioned above, may take place even register the also non-existent judgement of original copy.

And record the image-region that will distribute in the patent documentation 5 as the process object zone, extract image feature amount or character features amount, and the scheme of retrieval comparison object image, but need to urge the user to import the allotment that distributes original copy, whether be the formation of distributing the judgement of original copy perhaps.The former the time because allotment is not general, so need carry out explanation to allotment, and the formation of input allotment, be not efficient on handling, during the latter, have necessary whether judgement is set in addition be the problem of distributing the formation of original copy.

In addition,, use the character code of handling based on OCR, therefore be necessary in dictionary, to register in advance term, also have the problem of determination processing spended time as the character features amount.

Summary of the invention

The object of the present invention is to provide following image data output processing apparatus, it judges whether the image of input image data is similar to the image of registration original copy, and according to result of determination, the output of control input image data is handled, in this device, need not before carrying out homophylic judgement, whether make the user import input image data is the information of distributing original copy, perhaps be provided with in addition by constituting that input image data is judged, and be to distribute original copy at input image data, and when including the original image of restriction output processing, can go out it, and restriction output is handled with high Precision Detection.

To achieve these goals, image data output processing apparatus of the present invention is a device of input image data being exported processing, and it comprises: the image comparing part, it extracts the characteristic quantity of image from input image data, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, judge the image of above-mentioned input image data and register similarity between the image of original copy; Output processing controls portion, it is based on the result of determination of above-mentioned image comparing part, control is handled the output of above-mentioned input image data, wherein, above-mentioned image comparing part is when being assigned the distribution original copy of each image of a plurality of original copys, to be unit to distribute the image that comprises in the original copy at above-mentioned input image data, judges the similarity with the registration original copy, above-mentioned output processing controls portion is that unit control output is handled based on the result of determination of distributing each image that comprises in the original copy with the image.

For example, at input image data is 2 to close 1 original copy, and when including the registration original copy respectively in 2 zones, before carrying out the similarity judgement, whether not allowing the user import input image data is the information of distributing original copy, perhaps be provided with under the situation of judging by input image data about constituting in addition, in the judgement in the past, can only proceed to the judgement of the registration original copy which side is similar to.Therefore, the problem that exist to allow the output of the input image data that should limit to handle.

According to above-mentioned formation, be to distribute original copy at input image data, and include when limiting the image of exporting the registration original copy of handling, image with distribution is a unit, judge similar with the registration original copy, output processing controls portion is that unit control output is handled based on result of determination with the image.

Therefore, before carrying out the similarity judgement, whether be the information of distributing original copy, perhaps be provided with in addition by constituting that input image data is judged if need not to allow the user import input image data, just can handle with the output that limits the input image data that limit accurately.

Thus, can be provided at input image data and be and distribute original copy, and when including the original image that restriction output handles, can be detecting it accurately, and the image data output processing apparatus handled of restriction output.

In the present invention, comprise:, can add control device with the image data output processing apparatus of the equal function of image data output processing apparatus of the present invention by in the image data output processing apparatus that had, carry in the past; By being applicable in the past in the image data output processing apparatus that exists, can add image data output processing method with the equal function of image-data output device of the present invention.

In addition, above-mentioned image data output processing apparatus also can be realized by computing machine, at this moment, record by computing machine is moved as above-mentioned each parts, utilize computing machine to realize the recording medium of embodied on computer readable of the program of above-mentioned image data output processing apparatus, be also contained in the scope of the present invention.

By following record, can fully understand other purposes of the present invention, feature and advantage.In addition, in the following description of reference accompanying drawing, it is clearer that advantage of the present invention will become.

Description of drawings

Fig. 1 (a) be the formation digital color copier that relates to of expression an embodiment of the invention image data output processing apparatus in the process flow diagram of the step handled of the ballot of the document control treatment portion that had, Fig. 1 (b) is the process flow diagram of step of the similar degree determination processing of the similar degree determination processing portion in the identical document control treatment portion of expression.

Fig. 2 is the block diagram of concise and to the point formation of the image data output processing apparatus of the formation digital color copier that relates to of expression an embodiment of the invention.

Fig. 3 is the block diagram of the concise and to the point formation of the document control treatment portion that had in the expression image data output processing apparatus shown in Figure 2.

Fig. 4 is the block diagram of the concise and to the point formation of the unique point calculating part that had in the expression image data output processing apparatus shown in Figure 2.

Fig. 5 is the block diagram of the concise and to the point formation of the feature value calculation unit that had in the expression image data output processing apparatus shown in Figure 2.

Fig. 6 is the key diagram of an example of the filter factor of the compound filter that had in the MTF handling part of expression image data output processing apparatus shown in Figure 2.

The key diagram of an example of the center of gravity of Fig. 7 connecting area that to be the unique point calculating part of expression by image data output processing apparatus shown in Figure 2 extract from input image data and this connecting area.

Fig. 8 is the key diagram of an example of each center of gravity (unique point) of a plurality of connecting areas of extracting of text line that the unique point calculating part of expression by image data output processing apparatus shown in Figure 2 comprises from input image data.

The key diagram of a concern unique point that Fig. 9 extracts when being the feature value calculation unit calculated characteristics amount of expression by image data output processing apparatus shown in Figure 2 and an example of peripheral unique point.

The key diagram of an example of the concern unique point that Figure 10 (a)～Figure 10 (d) extracts when being the feature value calculation unit calculated characteristics amount of expression by image data output processing apparatus shown in Figure 2 and the combination of peripheral unique point.

The key diagram of an example of the concern unique point that Figure 11 (a)～Figure 11 (d) extracts when being the feature value calculation unit calculated characteristics amount of expression by image data output processing apparatus shown in Figure 2 and the combination of peripheral unique point.

Figure 12 (a) and Figure 12 (b) are in the expression image data output processing apparatus shown in Figure 2, the key diagram about an example of the index of the hashed value of each unique point and expression input image data of registering in the hash table.

Figure 13 be in the ballot handling part of expression image data output processing apparatus shown in Figure 2, to the curve map of an example of the votes of each registration image.

Figure 14 is the key diagram of the content of the coordinate figure admin table that is provided with in the document control treatment portion of expression image data output processing apparatus shown in Figure 2, manage each characteristic point coordinates value by each registration original copy.

Figure 15 is the key diagram of the content of the index of the unique point of the index of registration original copy of the hashed value unanimity that is provided with in the document control treatment portion of expression image data output processing apparatus shown in Figure 2, will obtain for each unique point of contrast object images and the registration original copy unique point corresponding tables of storing explicitly.

Figure 16 is the key diagram of 4 characteristic point coordinates values of 4 characteristic point coordinates values of the registration original copy that uses in the calculating of coordinate transform coefficient A of the similar degree determination processing portion in the document control treatment portion of expression image data output processing apparatus shown in Figure 2 and the contrast object images corresponding with them.

Figure 17 is the key diagram of the corresponding figure between the image of 4 unique points of 4 unique points of expression registration original copy shown in Figure 16 and contrast object images.

Figure 18 is which which unique point that expression is obtained by the similar degree determination processing portion in the document control treatment portion of image data output processing apparatus shown in Figure 2, the contrast object images register the consistent key diagram of characteristic quantity of image with.

Figure 19 (a) and Figure 19 (b) are that expression contrast object original copy is to distribute original copy, and when determining with the similar original image of registration original copy, determine the key diagram of method of the scope of the output processing in the image of restriction contrast object original copy.

Figure 20 is the process flow diagram of the flow process of the similarity determination processing in the expression image data output processing apparatus shown in Figure 2.

Figure 21 is whether the judgement in the original copy classification judegment part that is had in the image data output processing apparatus of expression variation is the key diagram that distributes the processing of original copy and allotment.

Figure 22 is whether the judgement in the original copy classification judegment part that is had in the image data output processing apparatus of expression variation is the key diagram that distributes the processing of original copy and allotment.

Figure 23 is the block diagram of concise and to the point formation of the image data output processing apparatus of the formation color digital compounding machine that relates to of expression other embodiments of the present invention.

Figure 24 is the key diagram of flow process in the expression image data output processing apparatus shown in Figure 23, the view data when fax is handled.

Figure 25 is the block diagram of the concise and to the point formation of the image data output processing apparatus that relates to of expression other embodiments of the present invention.

Figure 26 (a) is the figure that is used for illustrating the problem of image data output processing apparatus in the past, the image of expression contrast object original copy, and Figure 26 (b) is the curve map of the voting results of the contrast object images shown in expression Figure 26 (a).

Figure 27 be in the ballot handling part of the document control treatment portion that had in the image data output processing apparatus of the formation digital color copier that relates to of expression other embodiments of the present invention, to the curve map of an example of the votes of each registration image.

Figure 28 is that the similar degree determination processing of the document control treatment portion that had in the image data output processing apparatus of the formation digital color copier that relates to of expression other embodiments of the present invention is partly cut the key diagram of the processing of contrast object original copy.

Figure 29 be in the ballot handling part of the document control treatment portion that had in the image data output processing apparatus of the formation digital color copier that relates to of expression other embodiments of the present invention, to the curve map of an example of the votes of each registration image.

Figure 30 be the document control treatment portion that had in the image data output processing apparatus of the formation digital color copier that relates to of expression other embodiments of the present invention the ballot handling part, to the curve map of an example of the votes of each registration image.

Figure 31 be in the ballot handling part of the document control treatment portion that had in the image data output processing apparatus of the formation digital color copier that relates to of expression other embodiments of the present invention, to the curve map of an example of the votes of each registration image.

Figure 32 is the process flow diagram of the flow process of the similarity determination processing in the image data output processing apparatus of the formation digital color copier that relates to of expression other way of example of the present invention.

Embodiment

An embodiment of the invention are described.In addition, in the present embodiment, an example when being applicable to the present invention in the digital color copier is described, but the present invention is not limited thereto.

Fig. 2 is the block diagram of the concise and to the point formation of the digital color copier (image data output processing apparatus) 1 that relates to of expression present embodiment.

As shown in Figure 2, digital color copier 1 has coloured image input media 2, color image processing apparatus 3, coloured image output unit 4 and guidance panel 6.

Coloured image input media 2 is the devices that read the image of original copy and generate view data, for example constitutes by having the scanner section (not shown) that optical information with CCD (Charge Coupled Device) etc. is transformed to the device of electric signal.Here, coloured image input media 2 will (R: red, G: green, B: simulating signal indigo plant) outputs to color image processing apparatus 3 as RGB from the reflected light picture of original copy.

3 pairs of color image processing apparatus carry out various processing from the analog picture signals of coloured image input media 2 inputs, and are transformed to the form that coloured image output unit 4 can be handled, and to the device of coloured image output unit output.

Color image processing apparatus 3 has A/D (analog/digital) transformation component 11 that the simulating signal of RGB is transformed to digital signal in input stage.View data to the simulating signal of color image processing apparatus 3 input is transformed to digital signal by this A/ID transformation component 11.

Be transformed to the rgb signal of digital signal, after this, be sent to shade correction portion 12, original copy classification judegment part 13, document control treatment portion 14, input gray level grade correction portion 15, editing and processing portion 16, regional separating treatment portion 17, color correct portion 18 successively, the black background color that generates is removed portion 19, spatial filtering handling part 20, output gray level grade correction portion 21, grayscale reproduction handling part 22, and finally becomes the CMYK signal of digital signal.Then, temporarily store storer 23 into from the CMYK signal of the digital signal of grayscale reproduction handling part 22 output after, be output to coloured image output unit 4.

The processing of above-mentioned each one that constitutes color image processing apparatus 3 is described.The rgb signal of the numeral that 12 pairs of shade correction portions are sent here from A/D transformation component 11 is removed the processing of the various distortion that produce in the illuminator, imaging system, camera system of coloured image input media 2.In addition, shade correction portion 12 processing implementing the adjustment of colour balances and be transformed to the easy processing signals of the image processing system that concentration signal etc. adopts in color image processing apparatus 3.

Remove various distortion by shade correction portion 12, the rgb signal (concentration signal of RGB) that carried out the adjustment of colour balance is output to original copy classification judegment part 13.

Original copy classification judegment part 13 is differentiated the original copy classification of the view data of being imported from the rgb signal of being sent here by shade correction portion 12.Differentiating original copy is original copy for type composition, still prints the photo original copy, or the original copy classifications such as literal photograph print original copy of literal and photograph print mixing existence.The differentiation result of such original copy classification is used in the processing of back one-level.

In addition, original copy classification judegment part 13 also will remain untouched from the input signal of shade correction portion 12 outputs and output to the document control treatment portion 14 of back one-level.

Document control treatment portion 14 (similarity detection unit/output processing controls portion) has from rgb signal (input image data) extract minutiae of being sent here by original copy classification judegment part 13, and based on the function of the characteristic quantity of the image of each the feature point extraction input image data that extracts.

Document control treatment portion 14 from the characteristic quantity of the image data extraction image of the registering object imported, and with the characteristic quantity that extracts as described later, is registered in the memory storage with the index (original copy ID) of expression registration original copy in enrollment mode.

In addition, in normal control mode, carry out the similarity determination processing, promptly from the image data extraction characteristic quantity of the contrast object original copy imported, and compare with the characteristic quantity of the registration original copy of having registered, judge whether the image (below be also referred to as the contrast object images) of contrast object original copy is similar to the treatment of picture of registration original copy.In addition, below, with the comparison that relatively shows as contrast object original copy and registration original copy of the image of contrast object original copy and the image of registering original copy.

And then, document control treatment portion 14 also has the function as output processing controls portion 14, in the similarity determination processing, be similar to the registration original copy if be judged to be, then output forbids the output of input image data is handled the control signal of (handling for image forms) in color copy machine.

And, though details is described in the back, but in the present embodiment, document control treatment portion 14 is when distributing original copy at input image data, can be that the image that whether is similar to the registration original copy is judged by unit with a plurality of original images that comprised, based on each result of determination, limit the output that distributes in a plurality of original image units that comprise in the original copy and handle.

In addition, document control treatment portion 14 intactly input gray level grade correction portion 15 outputs of one-level backward of input signal that will come by original copy classification judegment part 13.

The rgb signals that 15 pairs of input gray level grade correction portions are sent here from document control treatment portion 14 implement background colors (the concentration composition of background color: background color concentration) remove or the picture quality adjustment of contrast etc. is handled.

Editing and processing portion 16 is to distribute original copy at input image data (contrast object original copy), when needing editing and processing, to the background color of enforcement sent here from input gray level grade correction portion 15 (the concentration composition of background color: removing or rgb signal that the picture quality adjustment of contrast etc. is handled background color concentration), carry out editing and processing.

As described above, in the present embodiment, in normal control mode, contrast object original copy is when distributing original copy, can be that the image that whether is similar to the registration original copy is judged by unit with a plurality of original images that it was comprised, only restriction distributes the output of the specific original image in a plurality of original images that comprise in the original copy to handle.

Such processing is undertaken by editing and processing portion 16, based on the control signal of sending here from document control treatment portion 14, only to being judged as the original image that is similar to the registration original copy, limiting its output and handles.

For example, similarly register the output of original copy and handle when being under an embargo being judged as certain original image, 16 in editing and processing portion is to being equivalent to the view data of this original image, data value is replaced into " 0 " or " 255 " (8 time), so that have only the corresponding original image that distributes in the original copy to become invisible.Thus, distribute original copy to become the part of having only forbidden original image and be coated with full state or blank sheet of paper state.

Like this, be sent to regional separating treatment portion 17 by the editor's of editing and processing portion 16 rgb signal.In addition, the rgb signal of sending here from input gray level grade correction portion 15 is not when distributing original copy, even perhaps distribute original copy, but when not needing editing and processing, the rgb signal of sending here from input gray level grade correction portion 15 is sent to regional separating treatment portion 17 by editing and processing portion 16.

Zone separating treatment portion 17 from rgb signal with any one in character area, dot area, the photo zone of each pixel separation the input picture.Zone separating treatment portion 17 is based on separating resulting, and the regional identification signal that the expression pixel is belonged to which zone is removed portion 19, spatial filtering handling part 20 and 22 outputs of grayscale reproduction handling part to color correct portion 18, the black background color that generates.In addition, regional separating treatment portion 17 also will be from intactly color correct portion 18 outputs of one-level backward of input signal of input gray level grade correction portion 15 outputs.

Color correct portion 18 is for the loyalty realization of color reproduction, and removes the CMY (C: dark green that does not need to absorb composition based on comprising; M: peony; Y: the yellow) processing of the look muddiness of the dichroism of look material.

Black generate background color remove 3 chrominance signals that portion 19 carries out the CMY after the color correct generate black (K) signal black generation, deduct the K signal of obtaining by black generation and generate new CMY Signal Processing from original C MYK signal.Thus, 3 chrominance signals of CMY are transformed to 4 chrominance signals of CMYK.

Spatial filtering handling part 20 to the view data of removing the CMYK signal of portion's 19 inputs from black generation background color, based on regional identification signal, carries out handling based on the spatial filtering of digital filter, revises spatial frequency characteristic.Thus, can alleviate the fuzzy of output image or graininess deterioration.Grayscale reproduction handling part 22 is also same with spatial filtering handling part 20, to the view data of CMYK signal, based on regional identification signal, implements predetermined process.

For example, separate the zone of giving literal,, and during clear in handling based on the spatial filtering of spatial filtering handling part 20 emphasize to handle, the amount of emphasizing of high frequency is increased particularly in order to improve the repeatability of black literal or color literal by regional separating treatment portion 17.Simultaneously, in grayscale reproduction handling part 22, select to be suitable for binaryzation or many-valuedization processing in the screen of high-resolution of reproduction of high-frequency region.

In addition, about be separated to the zone of dot area by regional separating treatment portion 17, in spatial filtering handling part 20, implement to be used to remove the low pass/Filtering Processing of input site composition.

Then, in output gray level grade correction portion 21, carrying out signal transformations such as concentration signals is after the characteristic value of coloured image output unit 4 is the output gray level grade correcting process of dot area percentage, in grayscale reproduction handling part 22, separation of images is a pixel the most at last, and the grayscale reproduction of carrying out reproducing the processing of gray shade scale is separately handled (generation medium tone).About separating the zone of giving photo, pay attention to binaryzation or many-valuedization processing on the screen of grayscale reproduction by regional separating treatment portion 17.

In the present embodiment, original copy classification discrimination result is appended to during after these regional separating treatment each handle, throughout in the reason, differentiating does not mix when existing for each zone, same with above-mentioned regional separating treatment, when differentiation exists for a plurality of regional mixing, use intermediate parameters of each zone processing, do not use the parameter of the zone processing of in original copy classification juggling, not differentiating.

For example, when differentiation is original copy for type composition for input picture, in regional separating treatment, the position of being separated by the zone as literal and lines is effectively, differentiate position, even original copy for type composition for example, the also situation that may exist mistake to differentiate according to the difference of the classification of original copy for Continuous Gray Scale grades such as site and photographic paper, therefore be considered as wrong the separation, can't reflect.

Then,, in input gray level grade correcting process and grayscale reproduction processing, use and remove more hi-lite, perhaps increase the fair curve of contrast based on this zone separating treatment result.

In addition, for the color literal, pay attention to the color correct of colored degree and handle, for black literal, remove in the processing at black generation/background color, it is more to set black growing amount.In addition,, in spatial filtering is handled, emphasize the edge, weaken smoothing and handle the isoparametric switching of ground setting filter factor etc. for literal.

In addition, when to differentiate for input picture be literal/photographic paper photo original copy, throughout in the reason, use original copy for type composition to handle and the processing of the intermediate parameters of photographic paper photo original copy processing.In regional separating treatment, to be made as effectively as the position that literal, lines or photographic paper carry out separating in the zone, the position that separate in the zone is carried out in sites etc., even literal/photographic paper original copy sometimes for example, the also situation that can exist mistake to differentiate because of difference according to the classification of original copy, therefore be considered as wrong the separation, can't reflect.

According to paying attention to which of original copy for type composition or photographic paper photo original copy, and in input gray level grade correcting process and grayscale reproduction processing, use the intermediate parameters between processing of photographic paper photo original copy and the original copy for type composition processing, and carry out the adjustment with contrast of removing of high light, in addition, carry out the power of colored degree and the balance of gray shade scale and do not become extreme color correct processing.On the other hand, remove in the processing,, deceive the adjustment of growing amount so that the photographic paper photograph image is not had effect at black generation/background color.

Carry out like this being read out at official hour, and being imported into coloured image output unit 4 after view data that above-mentioned each handle temporarily is stored in the not shown storer.

Then, in the similarity determination processing based on the normal control mode of above-mentioned document control treatment portion 15 time, be judged to be and be similar to the registration image, when for example output forbids contrasting the control signal that the output of the original copy of image handles, the data value of the view data of this page is replaced into " 0 " or " 255 " (8 time), so that publish picture picture during data from this memory read, it is invisible that it is become.Thus, the output from 4 outputs of coloured image output unit becomes whole full state or the blank sheet of paper state of being coated with of page or leaf.

Coloured image output unit 4 will output on the recording materials (for example paper etc.) from the view data of color image processing apparatus 3 input.The formation of coloured image output unit 4 does not limit especially, for example can use the coloured image output unit that adopts electrofax mode or ink-jetting style.

Guidance panel 6 is for example by display part such as LCD with set button etc. and constitute (all not shown), the pairing information of indication of the master control part (not shown) of demonstration and digital color copier 1 on above-mentioned display part, and will pass to above-mentioned master control part from the information that the user imports by above-mentioned setting button.The user can by guidance panel 6 input to the processing requirements of input image data, handle number etc.

Above-mentioned master control part is for example by CPU formations such as (Central Processing Unit), based on institute's program stored among the not shown ROM etc. or various data, from the information of guidance panel 6 inputs etc., and the action that comes each one of control figure color copy machine 1.

Below, the details of document control treatment portion (image comparing part, output processing controls portion) 14 is described.

The document control treatment portion 14 that present embodiment relates at first extracts a plurality of unique points from input image data, to each unique point of extracting, and the set of the unique point that decision is local.Then, select the part set of unique point,,, obtain invariant respectively geometric conversion based on a plurality of combinations about the unique point in the part set with the each several part set selected amount as supplementary features from each set of decision.Then, each invariant that combination is obtained, by each unique point, the calculated characteristics amount is a hashed value.

At input image data is that the data of registering object original copy are in the enrollment mode, and the index (original copy ID) that original copy is registered in the hashed value and the expression of calculating stores in the table relatedly.

On the other hand, at input image data is in the normal control mode of contrast object original copy, the registration original copy corresponding with the hashed value of calculating voted, and the situation that the unique point of the image of storage contrast object original copy (below, contrast object images) is voted which unique point of which registration original copy.Then, if poll closing is then considered the distribution of the unique point of the registration original copy that voting results and the hashed value of contrast on the object images are consistent, judge the similarity of contrast object images and registration original copy.

Fig. 3 is the block diagram of the concise and to the point formation of expression document control treatment portion 14.As shown in the figure, document control treatment portion 14 has unique point calculating part 31, feature value calculation unit 32, ballot handling part 33, similar degree determination processing portion 34, registration process portion 35, control part 7, storer 8.

The action of each one of control part 7 control document control treatment portions 14.In addition, control part 7 also can be arranged on the master control part of the action of each one that is used for control figure color copy machine 1, also can be to separate in addition with master control part to be provided with, and with the parts of the action of master control part Collaborative Control document control treatment portion 14.

Control part 7 is in the result based on the similarity determination processing of document control treatment portion 14, and when being judged to be no similarity, output allows control signal that the output of this view data is handled.And control part 7 is when judgement has similarity, the control signal that export-restriction is handled the output of this input image data.

When restriction output is handled, press the content of each registration original copy decision restriction, the restriction of the strictness that priority is the highest is forbidding of output processing.In addition, control part 7 is that the method as exporting the restriction of handling has 2 patterns when distributing original copy at contrast object original copy.

One is only to being judged to be the part of the original image that is similar to the registration original copy, the normal mode that restriction output is handled.Under normal mode, 16 in above-mentioned editing and processing portion limits its output and handles being judged to be the original image that is similar to the registration original copy.

Another is the compulsory mode stricter than normal mode.Under compulsory mode, distribute original copy to comprise when being similar to the original image of registering original copy, for distributing the whole original images that comprise in the original copy, use being judged to be the restriction that similar registration original copy is set.And, at this moment, having a plurality of original images that are similar to the registration original copy, limiting content is selected the highest restriction (the strictest restriction) of priority respectively not simultaneously.For example, include the original image that is similar to the registration original copy of forbidding exporting processing if be judged to be, then output will distribute the whole original images that comprise in the original copy to become sightless control signal.

Employed various data, result etc. in the processing of each one of storer 8 storage document control treatment portions 14.At this storer 8 hash table described later, coordinate admin table, unique point corresponding tables etc. are set.

Unique point calculating part 31 calculates the center of gravity of coupling part from the linking portion of input image data extraction text line or lines as unique point.Here, input image data is the view data of original copy of registration image in enrollment mode, is the view data (below, be also referred to as the contrast object image data sometimes) of the original copy of contrast object in similar determination processing.

Fig. 4 is the block diagram of the concise and to the point formation of representation feature point calculating part 31.In addition, the formation of unique point calculating part 31 shown in Figure 4 is examples, be not limited thereto, for example also can by from the past with regard to well-known method, come the calculated characteristics point.

As shown in Figure 4, unique point calculating part 31 has signal transformation handling part 41, resolution changing portion 42, MTF handling part 43, binary conversion treatment portion 44, center of gravity calculation portion 45.

Signal transformation handling part 41, does not have coloured silkization with this view data, and is transformed to lightness signal or luminance signal when being coloured image in the view data of importing from shade correction portion 12 (rgb signal).

For example, signal transformation handling part 41 is transformed to brightness signal Y based on following calculating formula (1) with rgb signal.

Yj＝0.30Rj+0.59Gj+0.11Bj …(1)

Here, Yj is the luminance signal of each pixel, and Rj, Gj, Bj are each colour contents in the rgb signal of each pixel, and the j of footnote is the value (j is the integer more than 1) that each pixel is given.

Perhaps, also rgb signal can be transformed to CIE1976L ^*a ^*b ^*Signal (CIE:Commission International de l ' Eclairage, L ^*: lightness, a ^*, b ^*: colourity).

42 pairs of input image datas of resolution changing portion carry out the multiplying power change to be handled.For example, input image data is changed multiplying power optically by coloured image input media 2 sometimes.At this moment, resolution changing portion 42 changes the multiplying power of input image data once more, thereby becomes the exploring degree (below, default exploring degree) of regulation.

MTF (modulation transfer function) handling part 43 is used for the spatial frequency characteristic of absorption (adjustment) coloured image input media 2 and presses each machine class and different difference.In the picture signal of CCD output, produce aperture openings degree, transmission efficiency or the afterimage of the sensitive surface of opticses such as lens or catoptron, CCD, based on the storage effect of physical scan with scan the deterioration of the MTF that causes such as inhomogeneous.Since the deterioration of this MTF, the image blur of being read in.

MTF handling part 43 carries out suitable Filtering Processing (emphasizing to handle), repairs the fuzzy processing that is produced by the deterioration of MTF thus.In addition, unwanted radio-frequency component during the feature point extraction that also is used for suppressing the center of gravity calculation portion 45 of back one-level is handled.That is, use compound filter (not shown) to emphasize and the smoothing processing.In addition, Fig. 6 represents an example of the filter factor in this compound filter.

Binary conversion treatment portion 44 is by relatively not having the view data (brightness value (luminance signal) or brightness value (lightness signal)) and the pre-set threshold of coloured silkization, with the view data binaryzation.

Center of gravity calculation portion 45 gives label (label is given processing) based on the view data (for example by " 1 ", " 0 " expression) by binary conversion treatment portion 44 binaryzations to each pixel.Then, determine to give the connecting area that the pixel of label links, with the center of gravity of the connecting area determined as feature point extraction.At this, as the unique point of center of gravity by the coordinate figure in the bianry image (X, Y) expression, and with the characteristic point coordinates value calculated to feature value calculation unit 32 outputs.

The key diagram of an example of the center of gravity of Fig. 7 connecting area that to be expression extract from input image data and this connecting area, expression is called the pairing connecting area of text line and the center of gravity of " A ".In addition, Fig. 8 is the key diagram of an example of each center of gravity (unique point) of a plurality of connecting areas of extracting of the text line that comprises from input image data of expression.

Return Fig. 3 of the block diagram of the concise and to the point formation of representing document control treatment portion 14, feature value calculation unit 32 is used the unique point that calculates in unique point calculating part 31, calculate rotation with respect to original image, parallel the constant amount of geometric distortion such as to move, amplify, dwindle be characteristic quantity (hashed value and/or invariant).

Fig. 5 is the block diagram of the concise and to the point formation of representation feature amount calculating part 32.As shown in the figure, feature value calculation unit 32 has the 32a of feature point extraction portion, invariant calculating part 32b, hashed value calculating part 32c.

The 32a of feature point extraction portion as paying close attention to unique point, pays close attention to the unique point of the periphery of unique point with this with a unique point as shown in Figure 9, from beginning only to extract in order specified quantity (being 4 points) here as peripheral unique point from paying close attention to the near unique point of unique point.In the example of Fig. 9, at unique point a when paying close attention to unique point, with unique point b, c, d, e 4 as peripheral feature point extraction, unique point b when paying close attention to unique point, with unique point a, c, d, e 4 as peripheral feature point extraction.

In addition, the 32a of feature point extraction portion extracts 3 the combination of selecting from the peripheral unique point extracted as mentioned above 4.For example, shown in Figure 10 (a)～Figure 10 (d), with unique point a shown in Figure 9 when paying close attention to unique point, extract 3 combination among peripheral unique point b, c, d, the e, each combination of promptly peripheral unique point b, c, d, peripheral unique point b, c, e, peripheral unique point b, d, e, peripheral unique point c, d, e.

Then, invariant calculating part 32b each combination about extracting calculates invariant (one of characteristic quantity) Hij to geometric distortion.

Here, i is a number (i is the integer more than 1) of paying close attention to unique point, and j is the number (j is the integer more than 1) of the combination of 3 of the peripheral unique points of expression.In the present embodiment, be invariant Hij with connecting in the length of peripheral unique point line segment to each other 2 liken to.

Based on peripheral characteristic point coordinates value, can calculate the length of each above-mentioned line segment.For example, in the example of Figure 10 (a), if the length of the line segment of connection features point b and unique point c is A11, the length of the line segment of connection features point b and unique point d is B11, and then invariant H11 is H11=A11/B11.

In addition, in the example of Figure 10 (b), if the length of the line segment of connection features point b and unique point c is A12, the length of the line segment of connection features point b and unique point e is B12, and then invariant H12 is H12=A12/B12.In addition, in the example of Figure 10 (c), if the length of the line segment of connection features point b and unique point d is A13, the length of the line segment of connection features point b and unique point e is B13, and invariant H13 is H13=A13/B13.In addition, in the example of Figure 10 (d), if the length of the line segment of connection features point c and unique point d is A14, the length of the line segment of connection features point c and unique point e is B14, and then invariant H14 is H14=A14/B14.Like this, in the example of Figure 10 (a)～Figure 10 (d), calculate invariant H11, H12, H13, H14.

In addition, in above-mentioned example, connecting the peripheral unique point of the most approaching concern unique point and the line segment of the second approaching peripheral unique point is Aij, connecting the peripheral unique point of the most approaching concern unique point and the line segment of the 3rd approaching peripheral unique point is Bij, but be not limited thereto, the line segment that uses in the calculating of invariant Hij can be selected with arbitrary method.

Then, hashed value calculating part 32c is with following calculating formula (2)

Hi＝(Hi1×10 ³+Hi2×10 ²+Hi3×10 ¹+Hi4×10 ⁰)/D …(2)

In the value of remainder calculate as hashed value (one of characteristic quantity) Hi, store in the storer 8.In addition, above-mentioned D is which degree is scope according to the desirable value of remainder be set at and come predefined constant.

In addition, the computing method of invariant Hij do not limit especially, for example, the configuration of 5 compound ratio will extracting based near 5 compound ratio of paying close attention to unique point, near n point (n is the integer of n 〉=5), the m point (m is the integer of m＜n and m 〉=5) that extracts near n point and 5 the compound ratio of extracting from m point, conduct such as the value that calculates is about the above-mentioned invariant Hij of concern unique point.In addition, so-called compound ratio is meant 5 values of obtaining on from the straight line 4 or the plane, is a kind of promptly to the invariant of distortion of projection of conversion geometrically.

In addition,, also be not limited to aforementioned calculation formula (2), also can use other hash function (for example any one in the hash function of record in the patent documentation 4) about being used to calculate the calculating formula of hashed value Hi.

In addition, each one of feature value calculation unit 32, if the extraction of a peripheral unique point of paying close attention to unique point and the calculating of hashed value Hi are finished, then will pay close attention to unique point and change to the further feature point, and carry out the extraction of peripheral unique point and the calculating of hashed value, calculate hashed value about whole unique points.

In the example of Fig. 9, if with unique point a when paying close attention to unique point peripheral unique point and the extraction of hashed value finish, then then carry out with unique point b when paying close attention to unique point peripheral unique point and the extraction of hashed value.In the example of Fig. 9, when paying close attention to unique point, unique point a, c, e, f 4 are extracted as peripheral unique point with unique point b.

Then, shown in Figure 11 (a)～Figure 11 (d), 3 the combination that extraction is selected from these peripheral unique point a, c, e, f (peripheral unique point a, e, f, peripheral unique point a, e, c, peripheral unique point a, f, c, peripheral unique point e, f, c), go out hashed value Hi about each combination calculation, store in the storer 8.Then, repeat this processing, obtain respectively the hashed value of each unique point when paying close attention to unique point, and store in the storer 8 about each unique point.

In addition, feature value calculation unit 32, in carrying out, the hashed value (characteristic quantity) about each unique point of input image data that calculates is sent to registration process portion 35 shown in Figure 3 as implied abovely with the enrollment mode of input image data as the processing of registration original copy registration.

The hashed value that registration process portion 35 calculates feature value calculation unit 32, represent that this input image data promptly registers index ID (ID1, the ID2 of original copy about each unique point ...) and the index f (f1, the f2 that represent this unique point ...) register to successively in the not shown hash table that is provided with in the storer 8 (with reference to Figure 12 (a)).

Here, when hashed value has been registered, register the original copy index explicitly with this hashed value.The original copy index repeatedly is not assigned with numbering successively.

And in the not shown coordinate admin table of the index f of representation feature point in being arranged at storer 8, as shown in figure 14, each the index ID by the expression original copy is associated with coordinate figure, and in identical original copy, index f repeatedly is not assigned with numbering successively.

For example, if the content of the hash table shown in Figure 12 (a), then hashed value H1 is the index f1 (coordinate (X in the registration original copy of index ID1 ₁₁, Y ₁₁)) the value that has of unique point, in addition, hashed value H3 is the index f4 (coordinate (X in the registration original copy of index ID1 ₁₄, Y ₁₄)) unique point and the index f1 (coordinate (X in the registration original copy of index ID2 ₂₁, Y ₂₁)) the value that has of unique point.

In addition, the quantity of the original copy of being registered in hash table more also can be retrieved old index ID and deletion successively than setting (for example 80% of the quantity of registrable original copy).In addition, the index ID that the index ID that is deleted also can be used as new input image data reuses.In addition, the hashed value of being calculated is for when the value, and shown in the example of Figure 12 (b), also they can be gathered is one, and registers in the hash table.Here, H1=H5 gathers and is H1.

And, feature value calculation unit 32, when whether the image that carries out input image data is the determination processing of image of the registration original copy of having registered (normal control mode), with the hashed value that calculates as described above about each unique point of input image data, unique point index p (p1, the p2 of the unique point that is associated with each characteristic point coordinates value with expression ...) send to together the ballot handling part 33.

In addition,,,, use " f ",, use " p ", and distinguish about the unique point of contrast object images about the unique point of registration original copy in order to understand explanation easily as the unique point index.

The hashed value of being registered the hashed value of each unique point that ballot handling part 33 will calculate from the input image data of contrast object original copy and the hash table compares, and to the registration original copy with identical hashed value vote (with reference to Figure 13).Figure 13 is the curve map of expression to an example of the votes of registration image I D1, ID2, ID3, ID4.

For example, in the content of the hash table shown in Figure 12 (a), if, then the registration original copy of ID1 is thrown 1 ticket, if, then the original copy of respectively registering of ID1 and ID2 is respectively thrown 1 ticket from contrast object picture extraction hashed value H3 from contrast object picture extraction hashed value H1.

In addition, in the example of Figure 12 (b), H1=H5 gathers them and is one of H1, and registers in the hash table, therefore, if calculate hashed value H1 from the contrast object images, then the registration original copy of ID1 is thrown 2 tickets.

And in the present embodiment, ballot handling part 33 is in when ballot, carries out being associated between the unique point of the unique point of contrast object images of hashed value unanimity and registration original copy, and stores in the not shown unique point corresponding tables.

The content of Figure 15 representation feature point corresponding tables.In the example of Figure 15, expression is judged to be the hashed value obtained at the unique point of index p1 of contrast object images, consistent with the hashed value of unique point of index f1 in the registration original copy of index ID1, in addition, the hashed value of obtaining at the unique point of index p2 of contrast object images, with the consistent example of hashed value of the unique point of the index f2 of the registration original copy of index ID3.

The step that the ballot of being undertaken by ballot handling part 33 in Fig. 1 (a) expression is handled.Shown in Fig. 1 (a), ballot handling part 33 is at first about the unique point (below, unique point p) of the index p of contrast object images, compares (S1) with the hashed value of registration original copy, and to the registration original copy of hashed value unanimity vote (S2).Particularly, judge that at first whether the hashed value identical with the hashed value of the unique point p that contrasts object images is stored in the hash table, when storing, votes to the registration original copy with this hashed value.

In addition, ballot handling part 33 is read the index ID of the registration original copy with hashed value identical with the hashed value of the unique point p that contrasts object images and the index f of unique point from hash table, and is stored in the above-mentioned unique point corresponding tables in ballot.

Whole unique point p about extracting in the contrast object images carry out the processing of such S1～S3 repeatedly, in S4, if be judged to be end about the whole unique point p that extract in the contrast object images, then finish ballot and handle.

Handle by ballot, obtain expression as the curve map of above-mentioned voting results shown in Figure 13 and unique point corresponding tables as shown in figure 15.

Similar degree determination processing portion 34 is from storer 8, based on i.e. being associated of the ballot result of ballot handling part 33 to the unique point of the unique point of the contrast object images of each the registration number of votes obtained of image and hashed value unanimity and registration original copy, judge in the contrast object images and whether include the registration original copy, and when comprising, judge which part of contrast object images is similar with the registration original copy.

The step of the similar degree determination processing that expression is undertaken by similar degree determination processing portion 34 among Fig. 1 (b).Shown in Fig. 1 (b), similar degree determination processing portion 34 at first based on the voting results to each registration original copy, selects the registration original copy (S11) of obtaining maximum votes.When voting results as shown in figure 13, select the registration original copy of index ID1.

Then, similar degree determination processing portion 34 uses the coordinate of the unique point f of the hashed value unanimity in the registration original copy of selecting, and calculates the coordinate transform coefficient A (S12) that is used for being represented by the coordinate figure on the contrast object images this registration original copy.

Here, use Figure 16, Figure 17, obtaining of coordinate transform coefficient A is described.Similar degree determination processing portion 34 reference feature point corresponding tables (with reference to Figure 15) about selected registration original copy, for 4 unique points of hashed value unanimity, obtain the characteristic point coordinates of its coordinate and contrast object images.

Figure 16 represents 4 characteristic point coordinates values of the registration original copy obtained by similar degree determination processing portion 34 and 4 characteristic point coordinates values of the contrast object images corresponding with them, and Figure 17 represents to register each unique point corresponding figure to each other between original copy and the contrast object images.In addition, as can be seen from Figure 17, the contrast object images is that original copy originally dwindles when distributing original copy, and rotation.

Similar degree determination processing portion 34 will be made as Pin about the matrix of characteristic point coordinates of registration original copy, to be made as Pout about the matrix of characteristic point coordinates of contrast object images, the coordinate transform coefficient is made as A, and, comes calculating coordinate change coefficient A by following mathematical expression (1).

(mathematical expression 1)

$\mathrm{Pin}=\left(\begin{array}{ccc}x1& \mathrm{<figure-callout\; id="1"\; label="y"\; filenames="H2008101493405E00021.png,H2008101493405E00052.png"\; state="\{\{state\}\}">y</figure-callout>}1& 1\\ x2& \mathrm{<figure-callout\; id="2"\; label="y"\; filenames="H2008101493405E00052.png,H2008101493405E00112.png"\; state="\{\{state\}\}">y</figure-callout>}2& 1\\ x3& \mathrm{<figure-callout\; id="3"\; label="y"\; filenames="H2008101493405E00052.png,H2008101493405E00112.png"\; state="\{\{state\}\}">y</figure-callout>}3& 1\\ x4& \mathrm{<figure-callout\; id="4"\; label="y"\; filenames="H2008101493405E00052.png,H2008101493405E00091.png"\; state="\{\{state\}\}">y</figure-callout>}4& 1\end{array}\right),$ $\mathrm{Pout}=\left(\begin{array}{ccc}x1\&prime;& \mathrm{<figure-callout\; id="1"\; label="y"\; filenames="H2008101493405E00021.png,H2008101493405E00052.png"\; state="\{\{state\}\}">y</figure-callout>}1\&prime;& 1\\ x2\&prime;& \mathrm{<figure-callout\; id="2"\; label="y"\; filenames="H2008101493405E00052.png,H2008101493405E00112.png"\; state="\{\{state\}\}">y</figure-callout>}2\&prime;& 1\\ x3\&prime;& \mathrm{<figure-callout\; id="3"\; label="y"\; filenames="H2008101493405E00052.png,H2008101493405E00112.png"\; state="\{\{state\}\}">y</figure-callout>}3\&prime;& 1\\ x4\&prime;& \mathrm{<figure-callout\; id="4"\; label="y"\; filenames="H2008101493405E00052.png,H2008101493405E00091.png"\; state="\{\{state\}\}">y</figure-callout>}4\&prime;& 1\end{array}\right),$ $A=\left(\begin{array}{ccc}a& b& c\\ d& e& f\\ g& h& i\end{array}\right)$

Pout＝Pin×A …(1)

Here, Pin is not positive square matrix, so multiply by the transposed matrix Pin of Pin on both sides ^T, multiply by Pin again ^TThe inverse matrix of Pin.

(mathematical expression 2)

Pin ^TPout＝Pin ^TP1in×A

(Pin ^TPin) ^-1Pin ^TPout＝A

Use the coordinate transform coefficient A obtain like this, calculate the coordinate position of contrast object images, thus can (x y) be transformed to coordinate on the contrast object images (x ', y ') with the coordinate arbitrarily on the registration original copy.

(mathematical expression 3)

(x′，y′，1)＝(x，y，1)×A

Return Fig. 1 (b), similar degree determination processing portion 34 uses the coordinate transform coefficient A that is calculated, and whole unique point f of the hashed value unanimity in the registration original copy that will select in S11 are transformed to the coordinate (S13) on the contrast object images.

Thus, for example, as shown in figure 18, contrast object original copy is to distribute 2 of 2 original images to close 1 original copy, and when the original image near the first half of true origin O wherein is similar to the registration original copy of the index ID1 that obtains maximum votes, obtain the first half that is distributed in the feature point set of the hashed value unanimity in the registration original copy of index ID1 of maximum votes on the contrast object images.

Distribution on the contrast object images of the unique point of the hashed value unanimity of similar degree determination processing portion 34 from selected like this registration original copy (coordinate position of extraordinary point), the position (S14) on the contrast object images of definite registration original copy.If distribution shown in Figure 180, then similar degree determination processing portion 34 determines that contrast object original copys are 2 to close 1 original copy, and is judged to be and is equivalent to the registration original copy that 2 images category of closing the regulation zone in the contrast object images in zone of the first half of 1 original copy are similar to index ID1.

Then, similar degree determination processing portion 34 judges whether finish (S15) about the whole registration original copys that should select based on voting results.For example, voting results are that 2 numbers of votes obtained of registration original copy of the registration original copy of index ID1 as shown in figure 13 and index ID3 are more outstanding when a lot of than other registration original copys, and the possibilities that are similar to these 2 registration original copys are higher.

Therefore, at this moment, in S15,, be not judged as and finish about the whole registration original copys that should select, select the registration original copy of obtaining secondly big number of votes obtained, here select the registration original copy (S16) of index ID3, and turn back to S12, for the registration original copy of ID3, also repeat the processing of S12～S15 (if 4 close the original copy of 1 grade, then also comprise S16).

Thus, as shown in figure 18, at contrast object original copy is to distribute 2 of 2 original images to close 1 original copy, and when the original image of remaining the latter half is similar to the registration original copy of the index ID3 that obtains second largest number of votes obtained, in S14, similar degree determination processing portion 34 is judged to be and is equivalent to the registration original copy that 2 images category of closing the regulation zone in the contrast object images in zone of the latter half of 1 original copy are similar to index ID3.

In addition, preferably, when in S11, S15, S16, selecting the registration original copy, for example will obtain the number of votes obtained of registration original copy of the number of votes obtained of upper level handles as the similar degree of degree like the representation class, compare with the threshold value TH that is predetermined, and judge the formation of similarity (whether input image data is the view data of registration original copy).

Promptly, maximum number of votes obtained is that the threshold value TH that is predetermined is when above, be judged to be " having similarity ", in S11, select the registration original copy and carry out the later processing of S12, in S15, S16, second number of votes obtained is the threshold value TH that is predetermined when above, is judged to be " the registration original copy that should select is arranged " or " having similarity ", and turns back to S12.

At this moment, descend, reduce setting threshold TH successively, can improve the judgement precision according to counting to second, third number of votes obtained such as grade from the maximum gained vote.

And the decision signal of the result of determination that expression is undertaken by similar degree determination processing portion 34 is sent to control part 7 as mentioned above like that, control part 7 if normal mode then only handle being judged as the original image restriction output that is similar to the registration original copy.And if compulsory mode, then to distributing the whole original images that comprise in the original copy, the output of the regulation of the restriction of the highest strictness is according to priority handled.

When 1 original copy is closed in Figure 19 (a) expression 2, limit the method that its output is handled for each the regional image that is positioned at the first half, the latter half.As mentioned above, by the similar degree determination processing of similar degree determination processing portion 34, be when distributing original copy, can detect 2 and close 1 original copy or 4 and close distribution states such as 1 original copy.Promptly as shown in figure 18, when the unique point of registration original copy concentrates on contrast the first half of object images or the latter half regional, can be speculated as 2 and close 1 original copy.

Control part 7, shown in Figure 19 (a), be 2 when closing 1 original copy, testing result and effective image-region from the direction of original size and original copy (direction that the longitudinal and transverse original copy of A4 is placed), for example by the information of the coordinate figure of 2 points at diagonal angle (white circle) etc., obtain the coordinate figure of 2 regional Q1, Q2 in advance.The direction of original size and original copy detects, and for example can use the components of photo-electric conversion such as phototransistor that dispose by coloured image input media 2 in, detects the method for the original size of the main scanning direction placed on document board, sub scanning direction; Utilize control part to detect the method etc. of the original size of selecting from guidance panel 6 by the user.In addition, effective image-region is to close in 1 original copy 2, is predetermined the zone that view data exists.Therefore, if 2 close 1 original copy, and among 2 regional Q1, the Q2 that can determine to have which be similar to the registration original copy, then can this regional view data be implemented restriction output easily and handle.Replace 22 regional Q1, the Q2 (effective image-region) that close 1 original copy, also can use 1/2 zone of original size.

Equally, Figure 19 (b) expression be 4 when closing 1 original copy, the unique point of registration original copy concentrates on the left of the first half left side and right side, the latter half and during any one of 4 regional Q1～Q4 on right side, can be speculated as 4 and close 1 original copy.

Control part 7 is shown in Figure 19 (b), with above-mentioned same, the information of the coordinate figure of 2 points by the diagonal angle (white circle) etc., and obtain 4 the coordinate figures of 4 regional Q1～Q4 when closing 1 original copy in advance, thus if 4 close 1 original copy, and can determine that among 4 regional Q1～Q4 which is similar to the registration original copy, just can implement easily that restriction output is handled to this regional view data.

In addition, similar degree determination processing portion 34, and then by will be to each number of votes obtained of registering original copy divided by ballot sum (sum of the unique point of extracting from input image data) and standardization, calculate similar degree, relatively this similar degree and the threshold value TH that is predetermined (for example 80% of the ballot sum) judge similarity thus.

In addition, similar degree determination processing portion 34, by will be to each number of votes obtained of registering original copy divided by registrating number (maximum registrating number) and standardization about the hashed value of the maximum registration original copy of the registrating number of hashed value, calculate similar degree, relatively this similar degree and the threshold value TH that is predetermined (for example 80% of the ballot sum) judge similarity thus.

That is, the similar degree of calculating is threshold value TH when above, is judged to be " similarity is arranged ", when being lower than threshold value TH, is judged to be " no similarity ".In addition, at this moment, the sum of the hashed value of extracting from input image data is than above-mentioned maximum registrating number bigger (particularly, when at least a portion of original copy and/or registration original copy has hand-written part etc.) sometimes, so the calculated value of similar degree might surpass 100%.

In addition, the threshold value TH when judging similarity can be set at constant about each registration original copy, perhaps also can set by each registration original copy according to each importance degree etc. of registering original copy.The importance degree of registration original copy for example about bank note, securities, classified papers, to the file of maintaining secrecy company outside etc., importance degree be maximum, about secret papers, importance degree than bank note etc. more the lowland according to registering image, setting stage by stage.

At this moment, in storer 8, the weighting coefficient that will be corresponding with the importance degree of registration image and the index of this registration original copy are stored explicitly, and similar degree determination processing portion 34 uses and the registration original copy corresponding threshold TH that obtains maximum number of votes obtained, judges similarity.

In addition, when judging similarity, it is constant making threshold value TH, also can multiply by the weighting coefficient of respectively registering original copy at the number of votes obtained (respectively registering the number of votes obtained of original copy) to each registration original copy in addition, judges similarity.

At this moment, in storer 8, to store explicitly with each index of registering the corresponding weighting coefficient of the importance degree of original copy and each registration original copy, the correction number of votes obtained that the number of votes obtained of each registration original copy be multiply by the weighting coefficient of this registration original copy calculates in similar degree determination processing portion 34, judges similarity based on this correction number of votes obtained.

For example, also can compare maximum modified number of votes obtained and threshold value TH, also can be with the value of ballot after the total standardization and threshold value TH relatively with the maximum modified number of votes obtained, also can with the maximum modified number of votes obtained after with maximum registrating number standardization value and threshold value TH relatively.In addition, at this moment, weighting coefficient for example can be set to the value greater than 1, and along with the importance degree of registration image uprises big value.

In addition, in the present embodiment, a unique point (concern unique point) is calculated a hashed value, but be not limited thereto, also can calculate a plurality of hashed values a unique point (concern unique point).For example, also can use, about 6 combinations of extracting at 5, extract at 3 from 5 and obtain invariant, and calculate the method for hashed value from these 6 as peripheral feature point extraction 6 points of paying close attention to unique point.At this moment, for a unique point, can calculate 6 hashed values.

Below, on one side with reference to process flow diagram shown in Figure 20, the similarity determination processing in the digital color copier 1 is described on one side.

Control part 7 if accept to import from user's indication by guidance panel 6, then judges whether select enrollment mode (S21).Here, when selecting enrollment mode, the registration process input image data.

At first, each one of control part 7 control document control treatment portions 14, and each characteristic point coordinates calculating (S22, S23) of carrying out the calculating of unique point and being calculated.With the index f of each characteristic point coordinates calculated and representation feature point explicitly, store in the coordinate admin table.Then, carry out characteristic quantity computing (S24), carry out the characteristic quantity that to be calculated and store registration process (S25) in the hash table into the index f of the index ID of expression original copy and representation feature point.

And in S21, be not enrollment mode if be judged to be, and promptly be normal control mode, then after S27, calculated characteristics point (S27).Then, based on the unique point that calculates in S27, each characteristic point coordinates of similarly carrying out being calculated is calculated (S28).With the index p of each characteristic point coordinates calculated and representation feature point explicitly, store the operating area of storer 8 into.Then, carry out characteristic quantity computing (S29), the processing (S30) of voting, similarity determination processing (S31).

In this S30, S31, as described above, contrast object original copy is when distributing original copy, judges and the similarity of registration original copy to distribute the image as unit that comprises in the original copy, when being similar to the registration original copy, determines the picture position on the contrast object original copy.

Then, in S31,, then judge whether be set with compulsory mode (S32) if be judged to be similarly.Here, when being set with compulsory mode, for distributing the whole original images that comprise in the original copy, application is to being judged as the restriction that similar registration original copy is set, and have a plurality of similar registration original copys, and restriction is not simultaneously, and output stipulates to export the control signal (S33) of processing with the highest restriction of priority (the strongest restriction).

In addition, when not setting compulsory mode, enter S34, only to distributing the part that is judged as the original image that is similar to the registration original copy in each original image that comprises in the original copy, the control signal that export-restriction output is handled.

Be not when distributing original copy, with the setting of compulsory mode irrespectively, the control signal that export-restriction output is handled.

In addition, among the S31,, then enter S35 if be judged to be not similarly, as decision signal " 0 ", the control signal that output allows intact output to handle.

Yet, in the above-described embodiment, distribution (characteristic point coordinates position) on the contrast object images of the unique point of the hashed value unanimity of similar degree determination processing portion 34 from the registration original copy, position on the contrast object images of definite registration original copy, take a decision as to whether the distribution original copy thus, but, also can become following mode as other embodiments.

In the present embodiment, similar degree determination processing portion is to be judged to be when being similar to a plurality of registration image in the result of primary similar determination processing, as shown in figure 28, according to the quantity that is judged as similar registration original copy, to contrast the object original copy and be divided into 1/2 or 1/4, and the characteristic quantity that uses the unique point from each zone, exist to calculate, carry out secondary ballot and handle.

The image of contrast object original copy all as object, is contrasted the result of the primary similar degree determination processing of judgement, is the quantity that surpasses the registration original copy of threshold value when being a plurality of, according to the quantity of similar registration original copy, cuts apart contrast object original copy.Distribute the number of cutting apart of original copy to be predetermined like that by 2,4,6,8,9, when therefore similarly the quantity of registration original copy is consistent with above-mentioned allotment, utilize this value to cut apart, when inconsistent, use in the above-mentioned allotment and cut apart than the bigger value of quantity of similarly registering original copy.That is, be 3 if surpass the quantity of the registration original copy of threshold value, then select 4, if 5, then select 6, if 7, then select 8, cut apart.Then, carry out the ballot of each cut zone and handle, determine the similar registration original copy of each cut zone.In addition, the result that the ballot of cut zone is handled is the quantity that surpasses the registration original copy of threshold value when being a plurality of, uses bigger value to cut apart.Therefore in addition, cut apart also nonsensically too meticulously, cut apart several maximums and be decided to be about 8 or 9.In addition, each corresponding with cutting apart number picture position is predetermined as described above.

Below, illustrate that the quantity of the registration original copy that surpasses threshold value is 2, contrast object original copy is divided into 2 examples that carry out similar judgement.

The result of similar determination processing, be the number of votes obtained that obtains Figure 27, when being judged as the registration original copy B of the registration original copy A that is similar to index ID1 and index ID2, single-point line L1 as shown in figure 27 is such, to contrast the object original copy and be divided into 1/2, become a zone and b zone.Then, use the characteristic quantity of the unique point in each zone, carry out secondary ballot and handle.

Then, according to the voting results in 2 zones, judge as follows.

1) voting results in a zone of secondary ballot processing are to have only registration original copy A number of votes obtained to surpass the voting results shown in Figure 29 of threshold value, and the voting results in b zone are when having only registration original copy B number of votes obtained to surpass the voting results shown in Figure 30 of threshold value, similar degree determination processing portion be judged as contrast object original copy be comprise registration original copy A and registration original copy B both sides image 2 close 1 original copy.

2) voting results in a zone (perhaps b zone) of secondary ballot processing are voting results as shown in figure 29, and the voting results in the opposing party's b zone (perhaps a zone) are the numbers of votes obtained of registration original copy A, registration original copy B when all be the following voting results shown in Figure 31 of threshold value, and it is that (perhaps b zone) comprises and register 2 of original copy A and close 1 original copy in a zone that similar degree determination processing portion is judged as contrast object original copy.

3) voting results in a zone (perhaps b zone) of secondary ballot processing are the voting results shown in Figure 29, and the voting results in the opposing party's b zone (perhaps a zone) are the numbers of votes obtained of registration original copy A, registration original copy B when all be the following voting results shown in Figure 31 of threshold value, and it is that (perhaps b zone) comprises and register 2 of original copy B and close 1 original copy in a zone that similar degree determination processing portion is judged as contrast object original copy.

4) voting results in the secondary ballot a zone of handling, b zone all are that the number of votes obtained of registration original copy A, registration original copy B is when all be voting results shown in Figure 31 below the threshold value, it is the side's that number of votes obtained was many during primary ballot was handled registration original copy that similar degree determination processing portion is judged as contrast object original copy, promptly if the result of Figure 27 then is judged as registration original copy A (not being to distribute original copy).

5) at least one side's in the secondary ballot a zone of handling, b zone voting results are that the number of votes obtained of registration original copy A, registration original copy B is when all surpassing the voting results shown in Figure 27 of threshold value, similar degree determination processing portion will contrast the object original copy to be cut apart with dotted line L21 shown in Figure 28, will contrast the object original copy and be divided into 4.Then, in each zone, the ballot of carrying out is for the third time handled, and the result based on the ballot in 4 zones is handled utilizes with above-mentioned same benchmark and judges.

In addition, about the operation of handling according to output such result of determination, input image data, identical with above-mentioned embodiment.

In addition, in the above description, illustration similar degree determination processing portion with the result of similar determination processing as being judged to be the benchmark that is similar to a plurality of registration images, a plurality of registration original copys surpass the situation of threshold value, for example, the maximum registration original copy of number of votes obtained and second, third ... between the number of votes obtained registration original copy how, obtain the poor of number of votes obtained, difference all is made as similar registration original copy for the registration original copy below the setting.

Below, with reference to process flow diagram shown in Figure 32, the similarity determination processing in the digital color copier of other embodiments is described.

In addition, the process flow diagram of the Figure 20 that uses in the explanation of the processing of S21～S31, S32～S35 and just now embodiment is identical.Therefore, give common number of steps, and omit explanation.

Similarity determination processing among the S31 is the above-mentioned similarity determination processing first time.In S31, when similar registration original copy is arranged, enter into S41, judge similarly whether the registration original copy is a plurality of, not a plurality of promptly one the time, the control signal (S44) of output expression in S31, being judged as the predefined processing of similar registration original copy.

If in S41, be judged as a plurality of, then, cut apart contrast object original copy, to the processing (S42) of voting of each zone of cutting apart according to the similar quantity of registration original copy, based on the ballot result in each zone, carry out the similarity determination processing (S43) of this contrast object original copy.

In addition, though in Figure 32, omit, but the result that the ballot in each zone of cutting apart of S42 is handled, implement and the same step of S41, if be judged as and in 1 zone, comprise a plurality of similar registration original copys, then further improve and cut apart number, carry out the ballot processing in each zone, the similarity determination processing of following repeatedly.

In the similarity determination processing of S43, not to distribute original copy if be judged as, then enter into S44, the output expression is to the control signal of the processing that is judged as similar registration original copy and is predetermined.And if be judged as is to distribute original copy, then transfer to S32.

The variation of above-mentioned digital color copier 1 (image data output processing apparatus) is described here.

Here, whether this original copy that original copy classification judegment part 13 is also differentiated the source of reading input image data is to distribute original copy, and is when distributing original copy, obtains the state of this distribution.

This method is described.Original copy classification judegment part 13 is from rgb signal (view data), to the main scanning direction of this image and each line of sub scanning direction, obtain pixel value and become 1 from 0, become 0 the number of reversals (perhaps from 1, number of edges) distribution, whether based on it, detecting is to distribute original copy, be when distributing original copy, to detect its allotment.

Figure 21 be have paper laterally (original copy is horizontal) be the text line that sub scanning direction extends, have line space paper vertically (original copy is vertical) be the histogram of each number of reversals non-distribution original copy, on main scanning direction and the sub scanning direction of the image arranged of main scanning direction.

During such non-distribution original copy, corresponding in the distribution of the number of reversals on main scanning direction as shown in figure 21 with line space in the article, the number of reversals appears every the interval of regulation.And the distribution of the number of reversals on the sub scanning direction becomes except the surplus white part of original copy periphery and to be continuous distribution.

Utilize 2 to close the 1 distribution original copy of handling such original copy shown in Figure 21, as shown in figure 22, at the paper that distributes original copy vertically is to arrange on the main scanning direction, and distribute 2 images, the paper that these each images have at this original original copy laterally is the text line that extends on the main scanning direction, and line space vertically is to arrange on the sub scanning direction at paper.

During such distribution original copy, the histogram of each number of reversals on main scanning direction and sub scanning direction, as shown in figure 22, the number of reversals on main scanning direction forms 2 continuous distributions, and has the white space corresponding with blank parts in the distribution of the number of reversals with between distributing.And the number of reversals on the sub scanning direction is corresponding with line space, the number of reversals occurs every the interval of regulation.

Based on such histogram, the number of reversals on main scanning direction has 2 continuous distributions, in the number of reversals that forms between 2 distributions is white space below the setting (for example 20), for example exist about (though also based on the layout of distributing original copy) 20mm (during exploring degree 300dpi, about 170 lines) time, be judged to be this contrast object images and be 2 and close the image of 1 original copy.At this moment, allotment becomes 2.

But, be judged to be white space when existing, be line below the setting when comprising initial line or final line in the number of reversals, being judged as is the surplus white region of original copy periphery, is necessary except it.

Whether like this, obtain the distribution of the number of reversals on main scanning direction and the sub scanning direction, judge having or not of white space, be 2 to close 1 original copy thereby can differentiate.

In addition, here, illustration allotment be 22 discrimination standard when closing 1 original copy, but allotment is 44 when closing 1 original copy, on main scanning direction, sub scanning direction, all occur and the corresponding white space of blank parts that separates between the original image, can differentiate equally.

In addition,,, obtain the distribution of the number of reversals (the perhaps quantity at edge) here, but also can use the mean value or the dispersion value of the number of picture elements of each line each line of main scanning direction, sub scanning direction.

Whether whether in addition, distribute the method for original copy except using such view data to differentiate to be, can also differentiate from imposing a condition be to distribute original copy.

For example, by using the input operation of guidance panel 6, when selecting the split image output mode in the image model, the master control part (CPU) of action that is used for each one of control figure color copy machine 1 is discerned it, is the data of distributing original copy thereby can differentiate the view data of being imported.

Perhaps, when if coloured image input media 2 is the scanner that connects on computers, then in the setting picture (the setting picture of scanner driver) of reading conditions, select the original copy classification, therefore above-mentioned master control part (CPU) is discerned this selection result, can differentiate.

In addition, editing and processing portion 16 is according to the distribution state of being obtained by above-mentioned original copy classification judegment part 13, decide and cut apart the position of distributing original copy, and come split image based on it, image to the each several part cut apart is rotated processing, processing and amplifying, and the line translation of going forward side by side is the editing and processing of 1 page 1 page view data.Thus, 1 page 1 page ground output processing distributes the page or leaf processed in units of the original image of original copy to become possibility.

Document control treatment portion 14 if judging to distribute includes the original image that is similar to the registration original copy in the original copy, then in compulsory mode, limits the output of whole page or leaf and handles in the page or leaf processed in units, and under normal mode, only the output of the corresponding page or leaf of restriction is handled.

In addition, for example, precision is judged and when handling well, with reference to the distribution on the contrast object images of the unique point of registering the hashed value unanimity in the original copy, perhaps cut apart contrast object original copy, and use the unique point that exists in each zone, and vote when handling, also can be with reference to above-mentioned result of determination.For example, also can be made as testing result, carry out the formation that the ballot of distribution on the contrast object images of use characteristic point or unique point is handled with reference to allotment.

Like this, by the reference allotment, the formation that detects allotment necessitates, and judges the judgement precision of distributing original copy but can improve.In addition, about whether handling, for example also can set normal mode and high precision mode, and can select from guidance panel with reference to allotment.

Yet, the advantage of the present application is, if do not detect such distribution state, or allow the user import, and use image contrast technology, judge whether contrast object original copy is to distribute original copy, if distribute original copy, then can control output with image as unit and handle based on the result of determination of distributing each image that comprises in the original copy.

In addition, in the present embodiment, the situation that the present invention is applicable to digital color copier 1 is described, but applicable object of the present invention is not limited thereto, for example as shown in figure 23, also go for color digital compounding machine (MFP:Multi-Function Printer) 100.This color digital compounding machine 100 has copy function, printer function, facsimile function, scan function, be scanned up to mail (scan to e-mail) function, filing function etc.

In addition, in Figure 23, the member that has same function for the part with explanation in digital color copier 1 is given identical symbol, and omits the explanation about them.

Here, dispensing device 5 for example is made of modulator-demodular unit or network interface card.Dispensing device 5 carries out data communication by network interface card, LAN cable etc. with other devices (for example, PC, server unit, other digit apparatus, facsimile unit etc.) that are connected on the network.

When dispensing device 5 sends view data, carry out transmission formality with the other side, if guarantee the state that can send, then read view data with prescribed form compression (view data of reading in by scanner) from storer, carry out the necessary processing such as change of compressed format, and send to the other side successively by communication line.

In addition, when dispensing device 5 receives view data, communicate formality, and receive the view data of sending here from the other side, and be input to color image processing apparatus 3.The view data of receiving is upheld predetermined process such as processing, rotation processing, resolution changing processing, the correction of output gray level grade, grayscale reproduction processing by color image processing apparatus 3, and by 4 outputs of coloured image output unit.In addition, also the view data that receives can be kept in the memory storage (not shown), color image processing apparatus 3 is read as required, and implements the processing of afore mentioned rules.

In addition, in compounding machine 100, can use guidance panel 6, import the user to the processing requirements of input image data (for example tupe (duplicating, printing, transmission, editor etc.), handle the transmission address of number (duplicate number, print number), input image data etc.), in document control treatment portion 14, when control part 7 has been judged to be when similar, not only handle duplicating, also outputs such as printing, transmission, editor processing is limited.

For example, select to send the pattern of fax, in document control treatment portion 14,, carry out the deletion of the view data of being stored in the storer, the transmission of not faxing when being judged to be when forbidding exporting.Perhaps, even the registration original copy, when also allowing the transmission of fax (will represent register the index ID of original copy and send target explicitly, storage in advance), the data of reference transmission address send.

Use Figure 24, illustrate that the fax in the above-mentioned color digital compounding machine 100 is handled.In addition, in Figure 24, dot and handle uncared-for handling part.Though here use, in the fax processing, regional separating treatment portion 17 not necessarily.In addition, when sending fax, after the grayscale reproduction handling part 22, handle by resolution changing portion, compression/extension handling part.

In transmission, 2 that read in from the coloured image input media, for example 8 input image data color image processing apparatus 3, be implemented above-mentioned each handle, for example by matrix operation etc., rgb signal is transformed to luminance signal (in Figure 24, the K signal) by input gray level grade correction portion 15.The view data that is transformed to luminance signal is carried out predetermined process by regional separating treatment portion 17, spatial filtering handling part 20, in grayscale reproduction handling part 22, for example handles by binaryzation by error diffusion.Be rotated processing as required by the view data of binaryzation, and be compressed, and store in the not shown storer by the form of compression/extension handling part with regulation.

Utilize dispensing device (for example modulator-demodular unit) 5 to carry out transmission formality with the other side, if guarantee the state that can send, then view data is read from above-mentioned storer under the compressed state of form with regulation, and implement the necessary processing such as change of compressed format, send by communication line to the other side successively.

And in reception, if send image here by communication line from the other side, then control part communicates formality on one side, receives the image from the other side sent on one side, and the reception view data of state of form that will be compressed to regulation is to color image processing apparatus 3 inputs.

To the reception view data of color image processing apparatus 3 input, by being upheld by the processing of compression/extension handling part, and the original image that the reconstruction of image is sent as page or leaf.The original image that is reproduced, the ability according to the coloured image output unit is rotated processing, and carries out resolution changing by resolution changing portion and handle.As the reproduced view data of the image of page or leaf unit, be output from the coloured image output unit.

In addition, in above, the situation of handling monochromatic view data is represented as example, but also can use for example between regional separating treatment portion 17 and spatial filtering handling part 20, has color correct portion 18 and the black background color that generates is removed the color image processing apparatus of portion, this color correct portion 18 is for the rgb signal that reads in from coloured image input media 2, for the loyalty realization of color reproduction, do not need to absorb CMY (C: dark green composition, ink powder or printing ink and remove to comprise; M: peony; Y: the yellow) processing based on the look muddiness of dichroism of look material, this is black to generate that background color removes that 3 chrominance signals of the CMY of portion after the color correct generate the black generation of black (K) signal and deducts by black from original C MY signal and generate the K signal that obtains and generate new CMY signal.

And then, in color digital compounding machine 100, as above-mentioned input image data, have by scanner and read original copy and view data of importing and the electronic data that uses a computer (software) generation, for example for the form of electronic data, the electronic data that can consider to use a computer (software) necessary item of input and generate, be about to based on the data electronization of paper spare data, utilize 2 kinds of data (electronic application etc.) that electronic data directly generates etc.

In addition, in the formation of Figure 23, carry out the similarity determination processing in the document control treatment portion 14 that in color digital compounding machine 100, has, but be not limited thereto.For example, also can with the connected communicatedly external device (ED) of color digital compounding machine 100 in carry out part or all of function that control part 7 and document control treatment portion 14 had.

In addition, the present invention for example also goes for monochromatic compounding machine.In addition, be not limited to compounding machine, for example also go for facsimile communication apparatus, duplicating machine, image read-out of monomer etc.

For example, Figure 25 is illustrated in the flat bed scanner 101 block diagram that is suitable for the configuration example when of the present invention.

As shown in the figure, flat bed scanner 101 has coloured image input media 2 and color image processing apparatus 3 '.Color image processing apparatus 3 ' is made of A/D transformation component 11, shade correction portion 12, document control treatment portion 14, control part 7 (not shown among Figure 25), storer 8 (not shown among Figure 25), be connected coloured image input media 2 with it, as all, composing images data output processing device.In addition, the function of the A/D transformation component 11 in the color image processing apparatus 3 ', shade correction portion 12, document control treatment portion 14, control part 7, storer 8 and above-mentioned digital color copier 1 are roughly same, therefore omit explanation here.

In addition, in the respective embodiments described above, constitute the document control treatment portion that had in digital color copier 1, compounding machine 100, the flat bed scanner 101 and each one (each member) of control part, utilize processor such as CPU and realize by software.That is, digital color copier 1, compounding machine 100, flat bed scanner 101 have the memory storages (recording medium) such as CPU (centralprocessing unit), the ROM (read only memory) that stores said procedure, the RAM (random access memory) that launches said procedure, storage said procedure and various memory of data of the order of carrying out the control program of realizing each function etc.And, can utilize computing machine to record the software of realizing above-mentioned function is the recording medium of the program code (execute form program, intermediate code program, source program) of the control program of digital color copier 1, compounding machine 100, flat bed scanner 101 with reading, supply with to digital color copier 1, compounding machine 100, flat bed scanner 101, the program code that is write down in this computing machine (perhaps CPU or MPU) playback record medium is also carried out, and realizes purpose of the present invention thus.

As above-mentioned recording medium, for example can use tape classes such as tape or magnetic tape cassette, comprise the card class or mask ROM (Mask the ROM)/semiconductor memory classes such as EPROM/EEPROM/ flash rom of the dish class, IC-card (comprising storage card)/light-card etc. of CDs such as disks such as floppy disk (registered trademark)/hard disk or CD-ROM/MO/MD/DVD/CD-R.

In addition, digital color copier 1, compounding machine 100, flat bed scanner 101 constitute joinably with communication network, supply with the said procedure code by communication network.As this communication network, do not limit especially, for example can utilize the Internet, intranet, extranet, LAN, ISDN, VAN, CATV communication network, Virtual Private Network (virtual private network), telephone wire road network, mobile radio communication, satellite communication link etc.In addition, as the transfer medium that constitutes communication network, do not limit especially, for example can utilize IEEE1394, USB, line of electric force transmission, catv line, telephone wire, adsl line etc. wired, also can utilize infrared rays such as IrDA or remote control, bluetooth (registered trademark), 802.11 wireless, HDR, mobile telephone network, satellite circuit, ground wave digital network etc. wireless.In addition, the present invention also can realize by the transmission form that specialize, that be embedded in computer data signal in carrier wave of said procedure code with electronics.

In addition, the each several part of digital color copier 1, compounding machine 100, flat bed scanner 101 is not limited to use software to realize, also can constitute, also can make up the hardware of a part of handling and carry out the control carry out this hardware or the formation of the arithmetic unit of the software of remaining processing by hardware logic.

Computer system of the present invention also can be by image-input devices such as flat bed scanner, film scanner, digital cameras, carry out image display devices such as the computing machine of various processing such as above-mentioned similar degree computing or similarity determination processing, the CRT monitor that shows the result of computing machine, LCD, the result of computing machine is constituted to the image processing systems such as printer of outputs such as paper by loading regulated procedure.And then, also can have by network be connected with server etc. as the network interface card of communication component or modulator-demodular unit etc.

As mentioned above, image data output processing apparatus of the present invention, as mentioned above, input image data is exported processing, and comprise: the characteristic quantity that extracts image from input image data, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, and judge the image of above-mentioned input image data and register the homophylic image comparing part of the image of original copy; Result of determination based on above-mentioned image comparing part, control the output processing controls portion that the output of above-mentioned input image data is handled, in this image data output processing apparatus, above-mentioned image comparing part is when distributing the distribution original copy of each image of a plurality of original copys at above-mentioned input image data, to distribute the image that comprises in the original copy is that unit is judged and the similarity of registering original copy, above-mentioned output processing controls portion controls output with image as unit and handles based on the result of determination of distributing each image that comprises in the original copy.

Thus, input image data is to distribute original copy, and when including the original image that restriction output handles, can provide with high accuracy detection it, and the view data output that restriction output is handled is handled.

In image data output processing apparatus of the present invention, and then from the image extract minutiae, based on each unique point of extracting, calculate above-mentioned characteristic quantity by each unique point, above-mentioned image comparing part is utilized the distribution of the degree of the characteristic quantity unanimity unique point consistent with the characteristic quantity on the input image data, judge similarity, when above-mentioned output processing controls portion was similar to the image of registration original copy in the images category of input image data, the control output that is positioned at the locational image of being determined by the distribution of the unique point of above-mentioned characteristic quantity unanimity was handled at least.

Also can realize aforesaid operations in the following way.Promptly, in image data output processing apparatus of the present invention, and then from the image extract minutiae, based on each unique point of extracting, calculate above-mentioned characteristic quantity by each unique point, above-mentioned image comparing part is utilized the positional information of the degree of the characteristic quantity unanimity unique point consistent with the characteristic quantity on the input image data, judge similarity, when above-mentioned output processing controls portion was similar to the image of registration original copy in the images category of input image data, the control output that is positioned at the locational image of being determined by the positional information of the unique point of above-mentioned characteristic quantity unanimity was handled at least.

In view of the above, the image comparing part is not only considered the degree of characteristic quantity unanimity, also adds the positional information of the unique point of the characteristic quantity unanimity on the input image data, and promptly distribution is judged similarity.

Input image data is to distribute original copy, when including the image of registering original copy in a plurality of original images, with the consistent unique point of characteristic quantity of this registration original copy, concentrates to be distributed in the part that disposes the registration image.Therefore, from the distribution of this unique point, be position information, can judge which of distributing original copy and partly dispose the registration image.

For example, as mentioned above, close 1 original copy 2, and when including the registration original copy respectively in 2 zones, 2 the characteristic quantity consistent unique points of registering original copys high with the consistent degree of characteristic quantity separately concentrate in 2 zones, therefore can judge to include 2 registration original copys.

That is, the consistent degree of characteristic quantity not only also adds the distribution of the unique point of the characteristic quantity unanimity on the input image data, judges similarity, and the result can be a unit with each original image that distributes, and takes a decision as to whether the registration original copy.

And, in output processing controls portion, at least the control output that is positioned at the locational image of being determined by the positional information of the unique point of such characteristic quantity unanimity is handled, thereby when being the image of registration original copy with the original image that distributes, can precision limiting its output well and handle.

In image data output processing apparatus of the present invention, and then from the image extract minutiae, based on each unique point of extracting, calculate above-mentioned characteristic quantity by each unique point, above-mentioned image comparing part, when the images category of input image data is similar to the image of a plurality of registration original copys, quantity according to similar image, with the image segmentation of input image data is a plurality of, the consistent degree of the characteristic quantity that utilization is calculated by the unique point that comprises in each zone of cutting apart, the information of the position of the unique point consistent with the characteristic quantity on the input image data, judge similarity, above-mentioned output processing controls portion, when the images category of input image data was similar to the image of a plurality of registration original copys, the control output that is positioned at the locational image of being determined by the positional information of the unique point of above-mentioned characteristic quantity unanimity was handled at least.

In view of the above, the image comparing part when the images category of input image data is similar to the image of a plurality of registration original copys, according to similar registration picture number, is a plurality of with the image segmentation of input image data.As mentioned above, be to distribute original copy at input image data, and when including the image of a plurality of registration original copys, it is similar to be judged as a plurality of registration original copys under a lot of situations.

At this moment, the image comparing part, the image of cutting apart input image data with the quantity of similar registration image, the information of the position of the unique point that the characteristic quantity on the input image data in the consistent degree of the characteristic quantity that unique point that utilization comprises from each zone of cutting apart is calculated and each zone of being cut apart is consistent is judged similarity.Thus, can precision be unit with each original image that distributes better, take a decision as to whether the registration original copy.

Then, in output processing controls portion, at least the control output that is positioned at the locational image of being determined by the positional information of the unique point of such characteristic quantity unanimity is handled, thereby when the original image that distributes is the image of registration original copy, can precision limits its output well and handle.

In image data output processing apparatus of the present invention, preferably constitute and also have following compulsory mode, be that above-mentioned output processing controls portion is when the images category of above-mentioned input image data is similar to the image of registration original copy, not only the restriction output that is positioned at above-mentioned definite locational image is handled, also limiting the output of the integral image of this input image data handles, and be preferably, when the restriction that output is handled is set each registration original copy, select the highest restriction of priority.

In addition, as image data output processing apparatus of the present invention, comprise: the image formation processing that is based on input image data and forms image on recording materials is handled in output, and has the image data output processing apparatus that forms the image formation unit of image based on view data on recording materials; Perhaps the output processing is that the filing that input image data is stored in the memory storage is handled, and has the image data output processing apparatus with the filing unit of image data storage in memory storage; Perhaps the output processing is that input image data is handled by the transmission that network sends to receiving trap, and has the image data output processing apparatus that view data is sent to the transmitting element of receiving trap by network.

The control device of image data output processing apparatus of the present invention, as mentioned above, be arranged on input image data is exported in the image data output processing apparatus of processing, and comprise: the image comparing part, it is implemented output for input image data and handles, and extract the characteristic quantity of image from input image data, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, judge the image of above-mentioned input image data and register similarity between the image of original copy; Output processing controls portion, it is based on the result of determination of above-mentioned image comparing part, control the output of above-mentioned input image data is handled, in this control device, above-mentioned image comparing part is when distributing the distribution original copy of each image of a plurality of original copys, to be unit to distribute the image that comprises in the original copy at above-mentioned input image data, judges the similarity with the registration original copy, above-mentioned output processing controls portion is that unit control output is handled based on the result of determination of distributing each image that comprises in the original copy with the image.

In addition, image data output processing method of the present invention, as mentioned above, input image data is exported processing, and comprise: the characteristic quantity that extracts image from input image data, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, judge that the homophylic image between the image of the image of above-mentioned input image data and registration original copy contrasts step; Result of determination based on above-mentioned image contrast step, control the output processing controls step that the output of above-mentioned input image data is handled, in this image data output processing method, above-mentioned image contrast step is when distributing the distribution original copy of each image of a plurality of original copys at above-mentioned input image data, to distribute the image that comprises in the original copy is unit, judge similarity with the registration original copy, above-mentioned output processing controls step is that unit control output is handled based on the result of determination of distributing each image that comprises in the original copy with the image.

In addition, above-mentioned image data output processing apparatus also can be realized by computing machine, at this moment, by computing machine is moved as above-mentioned each parts, the recording medium of utilizing the program of the above-mentioned image data output processing apparatus of computer realization and recording the embodied on computer readable of this program is also contained in the scope of the present invention.

Claims (10)

1. image data output processing apparatus, input image data is implemented output to be handled, and comprise: the image comparing part, it is a characteristic quantity from the amount constant with respect to geometric distortion that input image data extracts image, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, judge the image of above-mentioned input image data and register similarity between the image of original copy; Output processing controls portion, it is based on the result of determination of above-mentioned image comparing part, and control is handled the output of above-mentioned input image data, and this image data output processing apparatus is characterised in that:

Above-mentioned image comparing part is when being assigned the distribution original copy of each image of a plurality of original copys, to be unit to distribute the image that comprises in the original copy at above-mentioned input image data, judges the similarity with the registration original copy,

Above-mentioned output processing controls portion is unit based on the result of determination of distributing each image that comprises in the original copy with the image, and control output is handled.

2. image data output processing apparatus according to claim 1 is characterized in that:

From the image extract minutiae, based on each unique point of extracting, and calculate above-mentioned characteristic quantity by each unique point, this unique point is meant the center of gravity of each connecting area that links from the black pixel that will extract the view data after the input image data binaryzation,

Above-mentioned image comparing part is utilized the information of the position of the degree of the characteristic quantity unanimity unique point consistent with the characteristic quantity on the input image data, judges similarity,

Above-mentioned output processing controls portion, when the images category of input image data was similar to the image of registration original copy, the control output that is positioned at the locational image of being determined by the positional information of the unique point of above-mentioned characteristic quantity unanimity was handled at least.

3. image data output processing apparatus according to claim 1 is characterized in that:

From the image extract minutiae, based on each unique point of extracting, and calculate above-mentioned characteristic quantity by each unique point, this unique point is meant the center of gravity of each connecting area that links from the black pixel that will extract the view data after the input image data binaryzation,

Above-mentioned image comparing part, when the images category of input image data is similar to the image of a plurality of registration original copys, quantity according to similar image, with the image segmentation of input image data is a plurality of, and the positional information of the degree of unanimity of utilizing the characteristic quantity that calculates by the unique point that comprises in each zone of the cutting apart unique point consistent with the characteristic quantity on the input image data, judge similarity

Above-mentioned output processing controls portion, when the images category of input image data was similar to the image of a plurality of registration original copys, the control output that is positioned at the locational image of being determined by the positional information of the unique point of above-mentioned characteristic quantity unanimity was handled at least.

4. according to claim 2 or 3 described image data output processing apparatus, it is characterized in that:

Above-mentioned output processing controls portion has following compulsory mode, promptly when the images category of above-mentioned input image data is similar to the image of registration original copy, not only the control output that is positioned at above-mentioned definite locational image is handled, and also controls the output of the integral image of this input image data and handles.

5. image data output processing apparatus according to claim 4 is characterized in that: be set with the control that output is handled by each registration original copy, in above-mentioned compulsory mode, select the highest control of priority.

6. according to any described image data output processing apparatus in the claim 1～3, it is characterized in that:

Above-mentioned output is handled and is based on above-mentioned input image data, and the image that forms image on recording materials forms to be handled,

This image data output processing apparatus has the image formation unit that forms image based on view data on recording materials.

7. according to any described image data output processing apparatus in the claim 1～3, it is characterized in that:

It is the filing processing of above-mentioned input image data being stored into memory storage that above-mentioned output is handled,

This image data output processing apparatus has the filing unit of image data storage to memory storage.

8. according to any described image data output processing apparatus in the claim 1～3, it is characterized in that:

It is that above-mentioned input image data is handled by the transmission that network sends to receiving trap that above-mentioned output is handled,

This image data output processing apparatus has the transmitting element that view data is sent to receiving trap by network.

9. control device, be arranged on input image data is exported in the image data output processing apparatus of processing, and export processing for input image data, and this control device comprises: the image comparing part, it is a characteristic quantity from the amount constant with respect to geometric distortion that input image data extracts image, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, judge the image of above-mentioned input image data and register similarity between the image of original copy; Output processing controls portion, it is based on the result of determination of above-mentioned image comparing part, and control is handled the output of above-mentioned input image data, and this control device is characterised in that:

Above-mentioned image comparing part is when being assigned the distribution original copy of each image of a plurality of original copys, to be unit to distribute the image that comprises in the original copy at above-mentioned input image data, judges the similarity with the registration original copy,

Above-mentioned output processing controls portion based on the result of determination of distributing each image that comprises in the original copy, is a unit with the image, and control output is handled.

10. image data output processing method, input image data is exported processing, and comprise: the amount constant with respect to geometric distortion of extracting image from input image data is characteristic quantity, and this characteristic quantity that relatively extracts and the characteristic quantity of the image of the registration original copy of registration in advance, judge that the homophylic image between the image of the image of above-mentioned input image data and registration original copy contrasts step; Based on the result of determination of above-mentioned image contrast step, the output processing controls step that control is handled the output of above-mentioned input image data, this disposal route is characterised in that:

Above-mentioned image contrast step is when being assigned the distribution original copy of each image of a plurality of original copys, to be unit to distribute the image that comprises in the original copy at above-mentioned input image data, judges the similarity with the registration original copy,

Above-mentioned output processing controls step based on the result of determination of distributing each image that comprises in the original copy, is a unit with the image, and control output is handled.

Images