JPH10126614A - Image processing method, device therefor and printer using it - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow an image processor to accurately discriminate whether or not an original is an output inhibited object even the image data does not give magnification information. SOLUTION: A prescribed mark such as a circle mark whose diameter is 16mm is printed on an original that is an output inhibited object, and e mark of image data going to be outputted is compared with masks 34a-34e with a different marks, and a magnification estimate section 35 estimates a magnification given to a mask with a highest similarity degree (mask 34a: 110%, mask 34c: 100%, and mask 34e: 90%) to be a magnification of the image data. Based on the estimated magnification, a data extract section 29 extracts an image of a prescribed area, a discrimination section 30 normalizes the extracted data to be 100% and takes matching with reference data stored in a dictionary 31 so as to discriminate whether or not the data are an output inhibited object.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method and apparatus and a printer using the same, and more specifically, the magnification of an image to be actually output (magnification relative to a document) is unknown. Even in such a case, the present invention relates to a device that can accurately perform matching with a reference pattern.

[0002]

2. Description of the Related Art With the development of an image forming apparatus such as a full-color copying machine in recent years, the quality of a copied image has reached a level indistinguishable from the original image and the naked eye, and such faithful copying can be easily obtained. Was. Along with this, although copying of paper money and securities is originally prohibited in society, it is necessary to consider that the risk of being misused for counterfeiting will increase. (For example, an image processing apparatus disclosed in Japanese Patent Application Laid-Open No. 2-210481).

[0003] These various devices register an image relating to a copy-prohibited material such as a banknote in advance, and determine whether or not the input image includes the image (a specific recognition method is as follows). Variously developed). When such an image is recognized, various types of copy prohibition means are employed, such as printing out the entire copy paper in black, or stopping the copy processing itself.

[0004] By the way, a copying machine is basically to faithfully copy and output an object (original) placed on an original platen at its original size or at a predetermined magnification except for a copy prohibited material such as a bill. When an image similar to (i.e., an image that can be legally copied instead of a bill) is input, it is necessary to recognize that the image is not a bill and perform the normal copying process as it is. Therefore, the degree of coincidence (threshold of the determination criterion) when performing the recognition determination is inevitably increased.

[0005] When the copying process is performed by reducing or enlarging the image by the scaling function of the copying machine, the size (shape) of the image data given to the image processing apparatus from the copying machine main body side becomes the reference data registered in advance. (1: 1 magnification: 100%), the degree of coincidence is reduced, and it may be recognized as not a bill.

[0006] When the deviation of the scaling ratio is about several percent to several tens percent, the size is different if compared directly with a genuine banknote, so that the size is different. It is difficult to understand when looking at only (differential) things alone, and even more so for foreign banknotes, which are not usually used.

Therefore, as an apparatus for coping with the conventional magnification, for example, a plurality of types of reference patterns corresponding to each magnification are prepared, and magnification information is acquired from the copier main body during actual copying processing. In some cases, matching processing is performed based on a reference pattern corresponding to the obtained magnification.

Further, as shown in JP-A-6-237379, a thinning process is performed on image data based on scaling information (the thinning number increases as the enlargement ratio increases).
In some cases, the image after the thinning is made to have a fixed size, and the image after the thinning and the reference pattern are matched. As a result, the size of the image to be compared (after thinning) becomes constant irrespective of the scaling factor, so that only one type of reference pattern needs to be prepared. Then, a thinning pattern at the time of thinning processing is generated using a cyclic shift register.

On the other hand, recently, the resolution of a printer has been improved, and the resolution of a computer such as a personal computer and a scanner connected thereto has also been improved.
For example, by reading a bill or the like with a scanner and storing it in a memory in a personal computer, and then outputting the image data stored in the memory to a printer, there is a possibility that bills, securities, etc. may be forged. . Therefore, it has become necessary to take countermeasures against counterfeiting and the like in a printer which finally forms an image and outputs a counterfeit product.

[0010]

However, in the case of the above-mentioned printer, the image processing apparatus for preventing forgery in a copying machine in which various countermeasures have been conventionally taken is directly applied for the following various reasons. I can't.

That is, in the case of a copying machine, as shown in FIG. 1, a reading section 1 for reading a document and an image forming section 2 for forming and outputting a read image are incorporated in the same apparatus. When performing the copying process, first, the reading unit 1 reads a document placed on a document table, sends the read image data to the recognition unit 3 together with the image forming unit 2, and performs a predetermined recognition process in real time. Is recognized as a specific document such as a copy-prohibited material, the above-described predetermined copy prohibition process is performed. The operation of each unit is controlled by the control unit 4.

When a forgery is performed using such a copying machine, the "genuine article" for the forgery object is often placed on the platen. Then, as described above, since the image is read and the image is formed and output, the copying machine also provides information as to whether the original was copied at the same magnification or the magnification processing was performed and, in the case of the magnification, what percentage was enlarged / reduced. Since the control unit 4 of the main body is known, the recognizing unit 3 can perform the recognition process according to the magnification by acquiring the information from the control unit 4, and the recognition can be performed easily and with high accuracy. It becomes possible.

However, in the case of a printer, although the image data to be output is already stored in a memory of a computer or the like, it is not known what percentage of the original document is relative to the original document. However, as described above, it is not possible to recognize that the predetermined image is being output by the recognizing process corresponding to the scaling.

On the other hand, since the printer is a final output device in a counterfeiting operation, there is no actual harm even if output is not prohibited unless the object (image to be output) is difficult to distinguish from the real one. Therefore, at first glance it is difficult to identify, for example, 90%
Recognizing that it is output with a delicate magnification of about 110%,
It suffices if output can be prohibited.

Therefore, in addition to the recognition processing unit corresponding to 100%, a plurality of recognition processing units corresponding to N% (N is an arbitrary numerical value in the range of 90 to 110) are provided, and the plurality of recognition processing units are used. By performing recognition processing on the image data to be output respectively and determining the recognition processing result in an integrated manner, it is determined whether or not the image should be output-prohibited. Similarly, it is conceivable to perform a predetermined output prohibition process.

However, it takes too much time to perform recognition processing a plurality of times while switching the recognition algorithm by using one recognition processing unit, so that real-time processing becomes impossible. On the other hand, it is also conceivable to prepare as many recognition processing units as the number of scaling factors (N%) to perform recognition processing, and to configure a circuit so that each of the recognition processing units performs parallel processing on image data to be output. Can be With such a configuration, the time required for the recognition process can be the same as that of a single device. However, if the recognizing unit 3 corresponding to one type of scaling factor (actual magnification) has a configuration as shown in FIG. 2, the recognizing unit 3 'corresponding to the scaling factor.
As shown in FIG. 3, the circuit configuration becomes large and the cost is increased.

That is, when only the same magnification is considered, as shown in FIG. 2, the recognizing unit 3 includes the resolution converting unit 5 on the input side, and reduces the resolution of the input image data to a predetermined low resolution and blurs. An image is formed. And
By performing recognition processing based on such blurred image data, it is resistant to print misalignment and noise,
High-speed recognition processing is performed with a small circuit scale. The blurred image data is temporarily stored in the memory 6, and the image data of the predetermined area stored in the memory 6 is subjected to pattern matching by the candidate pattern extracting unit 7, and the candidate pattern which is considered to be a specific pattern is obtained. Is extracted, and when a candidate pattern is detected, necessary information (detection signal and position information, etc.) is sent to the data extraction control unit 8, and the data extraction unit 9 outputs the extracted signal from the data extraction control unit 8. The image data of a predetermined area stored in the memory 6 is extracted based on the extracted data, and the extracted (cut out) image data is provided to the determination unit 10 at the next stage. In the determination unit 10, the dictionary 1
The reference pattern stored in No. 1 is matched, and it is determined whether or not the reference pattern is a specific pattern. If the specific pattern is determined, a detection signal is output.

On the other hand, the size of the specific pattern to be recognized is different in order to cope with the magnification change. Therefore, as shown in FIG. 3, the candidate pattern extraction units 7a to 7e are The data candidate extraction control units 8a to 8e, the data extraction units 9a to 9e, the determination units 10a to 10e, and the dictionaries 11a to 11e are required. An integrating unit 12 for making a determination is also required. Thus, the circuit scale increases. In the illustrated example, for clarity of comparison with the embodiment of the present invention, it is assumed that five types of magnifications are supported for convenience. When the number is increased, such a problem appears more prominently.

Further, in each case, since the magnification of the output image with respect to the real document is unknown, each magnification (N%)
Needs to be given to the integration unit 12 to perform a comprehensive judgment process. Since it is determined whether or not one of the given recognition results matches the scaling factor, many recognition results differ from the actual scaling factor of the image data. Since the (recognition result) is also used as the determination material, the recognition range is widened, and even an image forming action that is not forgery may be recognized as a specific document (false recognition: too much) and output prohibition processing may be performed. .

The present invention has been made in view of the above background, and has as its object to solve the above-mentioned problems, to have a simple structure, and to be capable of real-time recognition processing.
In addition, it is possible to provide an image processing method and apparatus capable of minimizing the possibility of erroneous recognition, prohibiting output of an image to be prohibited, and permitting output of a normal image, and a printer using the same. is there.

[0021]

In order to achieve the above-mentioned object, an image processing method according to the present invention detects a predetermined mark present in image data to be output by an image forming apparatus, and detects the mark. A scaling factor of the image data to be output is estimated from a size of a mark, and matching of the image data with reference data is performed based on the scaling factor (claim 1).

Here, the predetermined mark corresponds to the circular pattern portion P1 in the embodiment. In the embodiment, a part of the specific pattern is also used as the predetermined mark. However, the present invention is not limited to this, and a mark suitable for specifying the magnification is separately provided from the specific pattern. It is of course also possible to provide such a detection and detect it.

In the present invention, the "magnification" refers to the magnification of an image to be finally output with respect to a real (original). Therefore, it does not matter whether the image is being deformed in the middle of the process.

The matching performed based on the scaling factor is as follows:
For example, after extracting a predetermined area in the image data, the image data is normalized to a reference magnification that constitutes reference data based on the estimated magnification, and the normalized data is matched with the reference data. (Claim 2). In addition, a plurality of types of reference data corresponding to different scaling factors are prepared, and reference data to be used is selected from the plurality of types based on the estimated scaling factor, and the selected reference data and the image data are compared. Matching can be performed (claim 3).

An apparatus suitable for carrying out the above-mentioned method includes a mark detecting means for detecting a predetermined mark existing in the image data to be output by the image forming apparatus (in the embodiment, a window 33 and a mark detecting means). Mask 34a ~
34e), scaling factor estimating the scaling factor of the image data from the size of the mark detected by the mark detecting device, and the scaling factor estimated by the scaling factor estimating means. Matching means for matching image data with reference data (in the embodiment,
(Corresponding to a determination unit, etc.).

The matching processing in the matching means can perform various kinds of processing in addition to pattern matching / template matching, extracting and comparing feature amounts, and includes what is generally called recognition processing.

The image processing apparatus further includes an area extracting means (corresponding to a data extracting unit in the embodiment) for extracting a predetermined area in the image data, and a dictionary storing reference data corresponding to a reference magnification. Means for normalizing the area extracted by the area extracting means to the reference magnification based on the estimated magnification, and matching the normalized data with reference data stored in the dictionary. (Claim 5). This specific configuration is described in detail in the embodiment shown in FIG. 5. In the embodiment, the reference magnification is set to 100%, but the present invention is not limited to this value.

Further, the image processing apparatus further comprises an area extracting means for extracting a predetermined area in the image data, and a dictionary storing reference data corresponding to different scaling factors, wherein the matching means is adapted to store the reference data based on the estimated scaling factors. The reference data to be used may be selected from the plurality of types, and the selected reference data may be matched with the image data. This specific configuration is described in the embodiment shown in FIG. 12, and in this embodiment,
Although five types of dictionaries were prepared, the number is arbitrary. Although the number of masks and the number of dictionaries are made equal, the present invention does not need to make them equal, and they may be different. As an example, assuming that there are five types of masks and the scaling factor estimated using the masks is as described in the modified example,
More than the type of mask (for example, it is determined comprehensively so that an intermediate scaling factor of the mask scaling factor can be obtained)
In such a case, the number of dictionaries may be equal to or greater than the number of masks corresponding to the type of the magnification.

On the other hand, the mark detection means has means for setting a window of a predetermined size, and a plurality of masks having a shape corresponding to the predetermined mark based on a plurality of magnifications. The means may compare the image data in the area extracted in the window with the plurality of masks, and determine a scaling factor based on the comparison result.

In the embodiment, "calculating the scaling factor based on the comparison result" means that the scaling factor of the mask having the maximum matching degree is estimated as the scaling factor of the image data. The present invention is not limited to this, and the scaling ratio may be obtained by comprehensively judging the results obtained with a predetermined mask.

In the present invention, when a predetermined mark is detected and the size is the size when the magnification is 100, the magnification of the image data to be processed can be estimated to be 100%. When the size of the predetermined mark is smaller than that of the same size, it can be estimated that the mark has been reduced, and it is possible to know the scale of the reduction from the specific size. Conversely, when the size of the predetermined mark is larger than that of the same mark, it can be estimated that the mark has been enlarged, and its magnification can also be estimated.
Therefore, in the present invention, a predetermined mark is detected based on such a principle, and the magnification of the image data to be output is estimated based on the size of the mark.

If the scaling factor can be estimated in such a manner, the size of the specific pattern can be known, so that, for example, the area (size) to be extracted can be set almost without excess or deficiency.
Also, since the scaling ratio is known for comparison / recognition / judgment processing with reference data, only one recognition algorithm needs to be used, and the possibility of entering unnecessary information can be suppressed as much as possible. Can be detected, and no erroneous detection / oversight occurs.

Further, when the scaling factor is known, the type of the reference data can be reduced and the memory capacity can be reduced by normalizing the scaling factor to a predetermined scaling factor. Further, when a plurality of types of reference data are provided as in claims 3 and 6, although the amount of use of the memory is increased by that amount, only one reference data is actually used because the scaling factor is known. As shown in FIG. 3, the integration processing becomes unnecessary, and the accuracy is increased.

According to the present invention, there is provided a printer comprising an image forming means for forming and outputting an image based on given image data, wherein the image processing device according to any one of claims 4 to 7 is provided. Along with mounting, the image data is input to the image processing apparatus in parallel with the image forming means, and the image processing apparatus transmits information regarding whether or not the image data being processed is a detection target. The output is performed, and when the image data is recognized as the detection target object from the output information, an output prohibition process is performed (claim 8).

With this configuration, even if the information on the scaling ratio of the image data to be output is not given, the scaling ratio is estimated by internal processing, and it is determined whether or not the image data can be output. In the case of a product, a predetermined output prohibition process is performed to prevent forgery and the like.

[0036]

FIG. 4 shows an example of an embodiment of a printer according to the present invention. As shown in FIG. 1, an image forming unit 20 receives image data provided from an external device such as a personal computer, and forms and outputs the image data on a predetermined sheet. This image forming unit 20
A predetermined printing process is performed in response to a control signal from the control unit 21. Since such a configuration is basically the same as that of a conventional printer, detailed description of each unit will be omitted.

Here, in the present invention, image data to be provided to the image forming unit 20 is also provided to the recognizing unit 22 in parallel, and the recognizing unit 22 determines whether or not a specific pattern exists in the image data. When it is recognized that the image data exists, a predetermined output prohibition process is performed on the image forming unit 20 via the control unit 21. As the output prohibition processing, the output itself is stopped, the whole or a part of the image is painted with black or another color, or a predetermined mark or character (for example, “sample”, “copy”) is printed in an overlapping manner. And so on.

FIG. 5 shows an example of the internal configuration of the recognition unit 22. First, in this example, a specific pattern (circle P) existing in the output image as shown in FIG.
1 and a star mark P2 existing in the circle), and if there is such a specific pattern, it is recognized as an output prohibited object.

The recognizing unit 22 has a resolution converting unit 25 on the input side, and reduces the resolution of the input image data to a predetermined low resolution to form a blurred image. As an example, in the case of a printer, for example, image data is generally given at 600 DPI,
I try to drop it to 100 DPI.

By performing recognition processing based on such blurred image data, it is possible to perform high-speed recognition processing with a small circuit scale while being resistant to printing deviation and noise. Also, by converting to 100 DPI in this manner, assuming that the specific pattern has a diameter of 16 mm as shown in FIG.
%), As shown in FIG. 7, the diameter is represented by 64 dots. When converting the resolution to 100 DPI, a binarization process may be performed.

The image data having the blurred resolution is temporarily stored in the memory 26.
The memory 26 is connected to a candidate pattern extraction unit 27 and a data extraction unit 29, and can output an image of a predetermined area to each of them. In the recognition processing according to the present embodiment, the candidate pattern that is likely to be a specific pattern is extracted by performing pattern matching on the image data of the predetermined area stored in the memory 26, and the candidate pattern is detected. In this case, necessary information (detection signal, position information, etc.) is sent to the data extraction control unit 28, and the data extraction unit 29 transmits the predetermined information stored in the memory 26 based on the extraction signal from the data extraction control unit 28. The image data of the region is extracted, and the extracted (cut out) image data is provided to the determination unit 30 at the next stage. The determination unit 30 matches the reference pattern stored in the dictionary 31 and determines whether or not the pattern is a specific pattern. If the determination unit 30 determines that the pattern is a specific pattern, it outputs a detection signal. The flow of such processing is the same in the apparatus of FIG. 2 for which a specific description is omitted.

Here, in the present invention, the candidate pattern extracting unit 2
In performing the extraction process in step 7, the scaling ratio of the image data being processed with respect to the original document is estimated, and scaling information is also output. That is, in the present example, the candidate pattern extraction unit 27 detects a circular pattern portion P1 that can be relatively easily detected among the specific patterns, and if there is, detects that there is a candidate pattern, and determines that there is a candidate pattern.
Internal data including 1 is cut out and subjected to a determination process. That is, the circular pattern portion P1 corresponds to a predetermined mark for estimating the magnification in the present invention.

Therefore, assuming that the image data to be recognized is the same magnification (magnification: 100%), the circular pattern portion P1 is shown in FIG.
Are compared with the image data using a mask as shown in (1) (a portion painted in black in the figure corresponds to the circular pattern portion P1), and the degree of circular coincidence is calculated based on the ratio of pixels that have coincided. When the degree of circle coincidence is equal to or more than a certain value, it is basically assumed that the circle pattern portion P1 is detected and that there is a candidate pattern.

Further, in this example, since the magnification corresponds to a magnification of 90 to 110%, the diameter of the specific pattern is about 14 mm (56 dots) at the magnification of 90%, and the magnification of 110% is obtained. In this case, the diameter of the specific pattern is about 18 mm (72 dots).

Therefore, as a result of the detection processing of the circle pattern portion P1, if a circle having a diameter of, for example, 64 dots is detected, it is the size of the mark that appears when the size is the same size. Can be estimated. Further, as a result of the detection processing of the circular pattern portion P1, for example, when a circle having a diameter of 56 dots is detected, it is the size of the mark that appears when the magnification is 90%, and the image data being processed is It can be estimated that the original document has been scaled to 90%, and if the circle pattern portion P1 is similarly detected as a result of the detection processing of the circle pattern portion P1, for example, if a circle with a diameter of 72 dots is detected, it is determined that the scale ratio has been changed. It can be estimated that the magnification has been reduced to 110%. Therefore, in the present example, the magnification is estimated from the size of the detected candidate pattern while performing the detection processing of the candidate pattern based on the principle.

Specifically, the candidate pattern extracting unit 27
Window part 33 and plural types of masks 34a to 34e
And a scaling estimating unit 35. Then, the window unit 33 extracts an image in the window area from the image data stored in the memory 26 using a 72-dot square window in order to be able to extract a circular pattern portion having a maximum diameter of 72 dots. It is supposed to.

Further, a plurality of types of masks 34a to 34e
In this example, since the range of the diameter of the circular pattern portion described above can be detected in increments of 1 mm, as shown in the figure, a circular mask 34a for a diameter of 18 mm (magnification of 110%) and a diameter of 17 mm (magnification of 105%) Circular mask 34b, diameter 16 mm (equal size) circular mask 34c, diameter 15 mm
(Differential magnification 95%) circular mask 34d, diameter 14mm
Five types of circular masks 34 e for (magnification: 90%) were prepared and connected to the window 33 in parallel.

As a result, as shown in FIG. 9, five masks M1 to M concentric and having different diameters are used for the image data in the 72 × 72 window W extracted by the window section 33.
5, and in each of the masks 34a to 34e,
Each circle matching degree is obtained, and the obtained circle matching degree is sent to the scaling estimating unit 35.

The scaling estimating unit 35 extracts the largest one of the given circle coincidences, and when it is equal to or greater than a predetermined value, determines that there is a candidate pattern and determines the maximum circle coincidence. The specified mask is specified, and the magnification of the mask is estimated as the magnification of the image being processed. Then, when such a candidate pattern is detected, not only the detection signal but also the magnification information is sent to the data extraction control unit 28.
Output to

When the candidate pattern is detected as described above, the data extraction control unit 28 needs to extract the detected circular pattern portion P1, that is, the inside of the mask having the maximum circular coincidence degree. Data (center coordinates and its diameter, etc.) to the data extraction unit 29,
The variable magnification is also sent to 0.

The data extraction section 29 extracts an area having a size corresponding to the magnification of the image to be output. That is to say, referring to FIG. 9 as an example, the mask M1
画素 M5, the pixel data existing in the inner part of the circle (which may include the circumference (circle pattern part) or may be inside the circle) having the highest degree of circle coincidence is extracted and sent to the determination unit 30 become. Thereby, for example, a magnification of 100%
(Fixed) to extract (diameter 6 for given center coordinates)
(4 dots are extracted), the outer peripheral part cannot be extracted for a pattern with a magnification of 110%, or a specific pattern more than necessary for a pattern with a magnification of 90%. There is no risk of extracting pixel data around the specific pattern portion, and the specific pattern portion can be extracted without excess or deficiency, so that the accuracy of the subsequent determination processing is improved.

The decision unit 30 normalizes the image data (circle) sent from the data extraction unit 29 based on the estimated scaling factor, and then compares the image data (circle) with the reference pattern stored in the dictionary 31. I try to take a match. That is, in this example, the dictionary 31 has a magnification of 100% (actual magnification).
Since the data relating to the reference pattern is registered, the data is normalized to a magnitude of 100% regardless of the magnification to be actually output. Note that 100
The reason for normalizing to% is to reduce the deformation ratio from the actual image data, and may be 90% or 110%, and further, normalized to any scaling factor outside or within the range. You may do it.

Since the determination processing in the determination unit 30 is performed by rotation matching, the normalization processing is also performed in units of radius. That is, if the scaling factor is estimated to be 110%, the extracted image data is
As shown in (A), a circle having a diameter of 18 mm is obtained. Therefore, this is normalized to a circle having a magnification of 100% (diameter 16 mm) as shown in FIG.

The radius of a circle having a diameter of 18 mm is 36 dots, the radius of a circle having a diameter of 16 mm is 32 dots, and the difference is 4 dots. Therefore, for example, four dots may be simply thinned out, but in this example, pixels arranged in the radial direction are divided into four as shown in FIG. Then 18
In the case of mm, one section is 9 dots. On the other hand, 16
In the case of mm, one section is 8 dots. Therefore, one dot is reduced for each section. And
The normalization also takes into account, for example, the arrangement of pixels in one section so that the arrangement of black and white does not change much before and after normalization. As an example, if there are seven blacks and two whites, it is possible to use seven blacks and one white. The conversion rule to be normalized can be created by, for example, creating a table that associates before and after the conversion with reference to the table at the time of normalization. Of course, simple thinning may be performed for each section, and various types of specific processing for normalization can be used.

Although illustration is omitted, in the case of 90%, the radius is 28 dots and one section is 7 dots.
One dot is complemented for each section. And
The middle can be dealt with by performing normalization in units of two sections.

Then, if a circle having a deformation rate of 100% (radius of 32 dots) is generated by normalization in this manner, as shown in FIG. 11, a predetermined angle from an arbitrary reference angle (0 degree) is obtained. By cutting out the pixel data in the radial direction at every time (for example, at intervals of 6 degrees) and arranging the pixel data, a strip-shaped matching pattern is generated as shown in FIG. Then, the matching pattern is compared with the reference data to determine the degree of coincidence, thereby determining whether or not the pattern is a specific pattern. In the case of the specific pattern, a detection signal is output to the control unit. It should be noted that the determination as to whether or not the pattern is the final specific pattern may be made on the control unit side only by obtaining the degree of coincidence (similarity).

By adopting the above-described configuration, the scaling factor can be estimated, so that a specific pattern portion can be accurately cut out, and one dictionary (in the embodiment, a scaling factor of 10).
0%), and the recognition processing (matching) in the determination unit 30 may be performed based on one type of scaling factor, so that the memory capacity and the circuit scale are reduced, and the luminous flux and accurate recognition processing are performed. I can do it.

The structure of the recognizing unit is not limited to the one described above, and may be, for example, as shown in FIG.
In this example, different from the above-described embodiment, a plurality of dictionaries 31a to 31e corresponding to the magnification are prepared. Then, according to the scaling factor estimated by the scaling estimating unit 35, the determination unit 3
At 0 ', a dictionary is selected as a comparison reference when performing the determination process. Then, the determination unit 30 'compares (references) the reference data stored in the selected dictionary with the given image data (rotation matching) to obtain a degree of coincidence. In this embodiment, since the dictionary having the same scaling ratio as that of the image data is selected, it is not necessary to perform normalization as in the above-described embodiment, and the number of processing steps is reduced, thereby enabling faster and more accurate processing. Becomes However,
Since a plurality of dictionaries are prepared, a large amount of memory is required. Note that the other configuration and operation and effect are the same as those of the embodiment shown in FIG. 5 described above, and thus detailed description thereof will be omitted.

In the above embodiment, an example in which five types of circular masks are provided has been described. However, the present invention is not limited to this, and a large number of circular masks may be provided. Then, by creating a fine magnification, a more accurate recognition process can be performed.

In each of the above embodiments, when estimating the scaling factor, the scaling factor corresponding to the mask having the maximum degree of coincidence is used. However, the present invention is not limited to this. The determined magnification may be comprehensively determined to determine the magnification. In such a case, for example, although the scaling factor corresponding to the mask is discrete, various processes can be performed on the scaling factor in the middle portion between adjacent scaling factors, and the recognition process can be performed with higher accuracy. I can do it. With such a configuration, the number of masks can be reduced,
Since a more accurate scaling ratio can be obtained, the recognition rate can be further improved.

[0061]

As described above, the image processing method and apparatus according to the present invention and the printer using the same can estimate the scaling factor even if the scaling factor of the output image data is not given. By performing image processing based on the estimated scaling factor, real-time recognition processing can be performed with a simple structure, and the possibility of erroneous recognition is suppressed as much as possible. However, normal images can be allowed to be output.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating an example in which a virtual processing device (recognition unit) is mounted on a conventional copying machine.

FIG. 2 is a diagram illustrating an example of an internal configuration of the recognition unit.

FIG. 3 is a diagram illustrating a conventional problem.

FIG. 4 is a diagram illustrating one embodiment of a printer according to the invention.

FIG. 5 is a diagram showing an internal structure of the recognition unit.

FIG. 6 is a diagram illustrating an example of a specific pattern.

FIG. 7 is a diagram illustrating a state of expression of a specific pattern in an image.

FIG. 8 is a diagram illustrating an example of a mask.

FIG. 9 is a diagram illustrating an operation.

FIG. 10 is a diagram illustrating normalization.

FIG. 11 is a diagram illustrating rotation matching.

FIG. 12 is a diagram showing another internal structure of the recognition unit.

[Explanation of symbols]

 Reference Signs List 20 image forming unit 22, 22 'recognizing unit 29 data extracting unit (region extracting unit) 30, 30' determining unit (matching unit) 31, 33a-33e dictionary 33 window unit (mark detecting unit) 33a-33e circular mask (mark) Detecting means) 35 Conversion estimating unit (magnification changing means)

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Shinya Sonoda 10 Okayama Todocho, Ukyo-ku, Kyoto, Kyoto Prefecture (72) Inventor Jin Nakamura 10 Okayama Todocho, Ukyo-ku, Kyoto, Kyoto Within (72) Inventor Mitsuru Matsumura 10 Ohanado Dodocho, Ukyo-ku, Kyoto-shi, Kyoto

Claims (8)

[Claims] An image forming apparatus detects a predetermined mark existing in image data to be output, estimates a magnification of the image data to be output from the size of the mark, and calculates a magnification based on the magnification. Wherein the image data is matched with reference data. 2. The matching performed based on the scaling factor, after extracting a predetermined area in the image data, normalizing to a reference scaling factor constituting reference data based on the estimated scaling factor, 2. The method according to claim 1, wherein data is matched with the reference data.
The image processing method according to 1. 3. A plurality of types of reference data corresponding to different scaling factors are prepared, reference data to be used is selected from the plurality of types based on the estimated scaling factors, and the selected reference data and the image are selected. 2. The image processing method according to claim 1, wherein matching with data is performed. 4. A mark detecting means for detecting a predetermined mark existing in image data to be output by an image forming apparatus; and a scaling factor of the image data based on a size of the mark detected by the mark detecting means. An image processing apparatus, comprising: a scaling estimating means for estimating; and a matching means for matching the image data with reference data based on the scaling factor estimated by the scaling estimating means. 5. An apparatus according to claim 1, further comprising: an area extracting means for extracting a predetermined area in said image data; and a dictionary storing reference data corresponding to a reference magnification, wherein said matching means is extracted by said area extracting means. The area is normalized to the reference scaling factor based on the estimated scaling factor, and the normalized data is matched with reference data stored in the dictionary. The image processing apparatus according to claim 4, wherein 6. An area extracting means for extracting a predetermined area in the image data, and a dictionary storing reference data corresponding to different scaling factors, wherein the matching means is configured to execute the processing based on the estimated scaling factors. 5. The image processing apparatus according to claim 4, wherein reference data to be used is selected from the plurality of types, and the selected reference data is matched with the image data. 7. The variable-magnitude estimation, wherein the mark detecting means includes a means for setting a window of a predetermined size, and a plurality of masks having a shape corresponding to the predetermined mark based on a plurality of magnifications. 7. The apparatus according to claim 4, wherein the means compares the image data in the area extracted by the window with the plurality of masks, and obtains a scaling factor based on the comparison result. The image processing apparatus according to claim 1. 8. A printer provided with image forming means for forming and outputting an image based on given image data, wherein the image processing apparatus according to claim 4 is mounted, and Data is input to the image processing apparatus in parallel with the image forming means, and the image processing apparatus outputs information on whether or not the image data being processed is a detection target, A printer wherein output prohibition processing is performed when image data is recognized as a detection target from output information.