JP4257172B2 - Digital image recognition apparatus and method - Google Patents

Description

  The present invention relates to an apparatus for recognizing a character pixel area from an image signal and performing image processing based on the recognition result, and more particularly to a digital image recognition apparatus and method capable of outputting to a plurality of types of recording apparatuses including an inkjet printer.

  Conventionally, representative methods for forming an image include an electrophotographic method and an inkjet method. In the electrophotographic system, a latent image is formed on a charged photosensitive drum based on input image information, developed, and transferred and fixed on a recording medium to reproduce the input image. In the inkjet method, an inkjet drive pulse having a plurality of ink ejection nozzles is generated based on input image data, and an input image is reproduced by controlling the presence or absence of ink ejection from each nozzle.

  There are pros and cons to electrophotography and inkjet. First, the recording speed is overwhelmingly higher for electrophotography. The image quality of the inkjet is excellent in dot reproducibility and dot position accuracy of 1 dot because there is no process such as latent image, transfer, and conveyance. For example, when outputting an image shot with a digital camera, the inkjet is higher. It is image quality. Furthermore, it is a great merit of inkjet that special colors can be used to form brightly colored images. However, with regard to character images, particularly black character images that frequently appear in a document, electrophotography is often superior. In inkjet printers, water-soluble dye inks that are difficult to dry are often used to prevent clogging of the nozzles due to ink drying, but images formed on recording media tend to bleed and have a sufficient black density compared to electrophotography. It has the disadvantages of being inferior in image storage stability (light resistance, water resistance, moisture resistance, etc.).

  A recording apparatus described in Patent Document 1 is a recording apparatus that records an image read from a document on a recording material, and includes an electrophotographic recording unit that performs recording by an electrophotographic method, and an inkjet recording unit that performs recording by an inkjet method And an apparatus that performs recording by an electrophotographic method and / or an ink jet method according to a user's request, and its purpose is to make use of the respective merits described above. For example, electrophotographic recording is performed for black and white copying, and ink jet recording is performed for color copying. As another example, it may be effective to perform switching such as recording by an electronic photograph in the case of a manuscript almost occupied by a character image and recording by an ink jet method in the case of a manuscript occupied by a picture image. As another example, there is an example in which a black portion is reproduced by an electrophotographic method and a color portion is reproduced by an ink jet method in one image.

By the way, when an image is read by a color scanner, predetermined image processing is performed on the read image, and image formation is performed by electrophotography, ink jet, or the like, the black character portion in the image is identified and the black character portion is made of only a black color material. The black character processing to be reproduced is a well-known technique. This is to prevent the color dots from appearing outside the black characters when the color shift of the output device occurs. A device that detects a black edge by detecting an edge and determines a black edge in combination with the result of color determination, and performs black character processing on the black edge (results in a reproduced image such as 1 after black character processing in FIG. 10), a result of edge detection Based on the above, there is an apparatus for determining the inside character as a character and processing the black character (the reproduced image is as shown in 2 after the black character processing in FIG. 10). An apparatus that processes black characters in a character has a problem of determination accuracy and a hardware restriction of a reference area size, and it is inevitable that a determination result is mixed in somewhere and becomes a defect. However, since the color difference between the edge and the inner side that occurs when only the character edge is reproduced with black color material and the inside of the character is reproduced with three-color color material can be eliminated. The device is more useful. In the apparatus described in Patent Document 1, this corresponds to the case where the black portion in one image is electrophotographic / color portion is ink-jet reproduced.
JP-A-8-104034 “Image area separation method for mixed images of characters / pictures (dots, photos)”, IEICE Transactions D-II Vol.J75-D-II No.1 pp.39-47 January 1992

  As mentioned earlier, inkjet has insufficient black density and inferior image storage compared to electrophotography, so regenerating only the edges with electrophotographic black toner and the inside of characters with inkjet color ink produces a large color difference. The gap is noticeable. Even when an image is reproduced only by inkjet, there are devices that use pigment ink for black ink and dye ink for cyan, magenta, and yellow color inks. The color difference is conspicuous.

  Pigment inks are suitable for character images because they can form sharp dots with less blurring than dye inks, but there is a problem with respect to the surface strength that the ink peels off when the surface is rubbed or bent. In order to take advantage of the benefits of pigment inks in black text images, black pigment inks and color dye inks are used in combination, but black regenerated with black pigment ink is more than black regenerated with black dye ink. However, since the black density reproducibility is good and the image preservability is relatively high, the color difference becomes conspicuous many times as much as the black dye ink / color dye ink. Even in the case of black dye ink / color dye ink, compared with the case where black is regenerated with only black ink, when black is regenerated with three-color ink, the change over time for three colors occurs. In many cases, image storability is poor.

  In the above description, only the color difference (related to density and color) was mentioned, but the glossiness differs greatly between the black color material and the color color material, and they may be used in combination in one image. Gap due to the difference in gloss occurs and becomes a problem.

  SUMMARY OF THE INVENTION An object of the present invention is to provide characteristics of a recording apparatus (especially characteristics of a coloring material) in an apparatus that switches recording formation in an electrophotographic system and / or an ink jet system in accordance with user's request, user instruction / automatic detection of document color and document type ) To switch whether to determine the character pixel area including the inside of the character, and to form one image using both electrophotography and inkjet, or one image using pigment ink and dye ink In the case of forming a character, the character is judged to the inside of the character to reproduce a high-quality image without the edge of the formed image and the gap in the character, while on the other hand, when reproducing one image only with an electrophotography, etc. In the case of color materials that are relatively difficult to generate gaps, high quality is unlikely to cause major degradation by judging only the character edges to avoid image degradation due to misjudgment. In addition, when recognizing characters inside the character, the recognition accuracy of the character pixel region inside the character is improved. When only character edges are determined, the character pixel region is not increased unnecessarily. In other words, the character pixel area is recognized so that a natural reproduced image that is difficult to identify erroneous detection is obtained.

In order to solve the above problems, a digital image recognition apparatus according to claim 1 is provided.
In a digital image recognition apparatus for inputting an image signal and recognizing a character pixel area based on the image signal, detection means for detecting a character edge pixel candidate area in the document based on the input image signal, and the detection A first correction unit that corrects the result based on the pixel density of the character edge pixel candidate region and detects the character edge pixel region, and the character inside the character with respect to the correction result corrected by the first correction unit Image processing is performed on the image signal based on the recognition result by the digital image recognition device based on any one of output speed, document color, and document type, and a second correction unit that corrects the pixel region . A recording device selection means for selecting a plurality of or one recording device from a plurality of recording devices each having a different recording method as a result output destination;
Depending on the characteristics of the recording device that is the output destination of the result of performing image processing on the image signal selected by the recording device selection means based on the recognition result by the digital image recognition device, the character pixel region is formed as far as the character inside. Control means for determining whether to determine only a character edge as a character pixel region, and controlling to switch the first correction means and the second correction means on or off simultaneously based on the determination And the control means turns on the second correction means and increases the first correction means in the direction of increasing the character pixel area. In the case where only the character edge is determined as the character pixel region, the second correction unit is turned off and the first correction is performed. Correcting means is characterized by, turning off so as not to perform correction in a direction to increase the character pixel area.

According to a second aspect of the present invention, there is provided a digital image recognition apparatus that receives an image signal and recognizes a character pixel area based on the image signal. Detecting means for detecting a character edge pixel candidate area, first detection means for correcting the detection result based on a pixel density of the character edge pixel candidate area, and detecting a character edge pixel area; A second correction unit configured to correct the correction result corrected by the correction unit so that the inside of the character is a character pixel region; and the image signal is output based on any one of an output speed, a document color, and a document type. A plurality of or one recording device among a plurality of recording devices each having a different recording method, which is an output destination of an image processing result based on a recognition result by the digital image recognition device A recording device selection means for selecting the location, the image signal selected by the recording device selection means to the characteristics of the recording apparatus as the output destination of the result was carried out image processing based on the recognition result by the digital image recognition device Accordingly, it is determined whether to determine the character pixel region up to the inside of the character or only the character edge as the character pixel region, and based on the determination, the first correction unit and the second correction unit are simultaneously turned on, Or a control unit that performs switching control to be turned off at the same time, and when the control unit determines that the character pixel area is within the character, the second correction unit is turned on and the first correction is performed. When the means is turned on to perform correction in the direction of increasing the character pixel area, and only the character edge is determined as the character pixel area, the second supplement The first correction means as well as turning off the device, characterized in that, to allow to reduce the correction amount in the direction to increase the character pixel area.

According to a third aspect of the present invention, in the digital image recognition apparatus according to the first aspect, when the recording apparatus forms an image by electrophotography, the second correction unit is turned off and the first correction unit is turned off. The correction means controls the first and second correction means so as not to perform correction in the direction of increasing the character pixel area.

According to a fourth aspect of the present invention, in the digital image recognition apparatus according to the second aspect, when the recording apparatus forms an image by electrophotography, the second correction unit is turned off and the first correction unit is turned off. The correction means controls the first and second correction means so as to reduce the correction amount in the direction of increasing the character pixel area.

According to a fifth aspect of the present invention, in the digital image recognition apparatus according to the first or second aspect , when the recording apparatus forms an image by an ink jet method, the second correction unit is turned on and the first correction unit is turned on. The correction means controls the first and second correction means so as to perform correction in the direction of increasing the character pixel area .

According to a sixth aspect of the present invention, in the digital image recognition apparatus according to the first or second aspect , when the recording apparatus forms an image by an electrophotographic method or an inkjet method, the second correction unit is turned on. At the same time, the first correction means controls the first and second correction means so as to perform correction in the direction of increasing the character pixel area .

The invention of claim 7 is a digital image recognition apparatus according to claim 1 or 2, wherein, if the recording apparatus is that to form an image by the pigment ink and dye ink, to select the second correction means In addition, the first correction means controls the first and second correction means so as to perform correction in the direction of increasing the character pixel area.

The invention according to claim 8 is the digital image recognition device according to claim 1 or 2 , wherein the color determination means for determining whether the pixel region is a chromatic pixel region or an achromatic pixel region based on the image signal, A black character pixel region determining unit that determines whether or not the character pixel region is a black character pixel region based on the correction result determined by the second correcting unit and the determination result by the color determining unit; and the black character pixel region determination Image processing means for performing image processing based on a determination result by the means.

The invention according to claim 9 is a digital image recognition method for inputting an image signal and recognizing a character pixel area based on the image signal, and based on the input image signal, a character edge pixel candidate area in the document. A detection step for detecting the character edge, a first correction unit for correcting the detection result based on a pixel density of the character edge pixel candidate region, and a second correction step for detecting the character edge pixel region; A second correcting step for correcting the correction result corrected by the first correcting step by the correcting unit so that the inside of the character is a character pixel region; and an output speed and an original color by the recording device selecting unit. Each of the output destinations of the result of image processing of the image signal based on the recognition result of the digital image recognition device based on one of the document types has a different recording method. A recording device selection step of selecting a plurality of or one recording device from the plurality of recording devices, and image processing of the image signal selected by the recording device selection means based on a recognition result by the digital image recognition device. In accordance with the characteristics of the recording device that is the output destination of the result of the determination, it is determined whether to determine the character pixel area up to the inside of the character or only the character edge as the character pixel area, and based on the determination, the first A control step of controlling the correction means and the second correction means to be turned on or off at the same time, and when the control step determines the character pixel area to the inside of the character, And the first correction unit is turned on so as to perform correction in the direction of increasing the character pixel area, and only the character edge is When determining as a character pixel region, the second correction unit is turned off and the first correction unit is turned off so as not to perform correction in the direction of increasing the character pixel region. And

According to a tenth aspect of the present invention, in the digital image recognition method for inputting an image signal and recognizing a character pixel region based on the image signal, the character edge pixel candidate region in the document based on the input image signal. A detection step for detecting the character edge, a first correction unit for correcting the detection result based on a pixel density of the character edge pixel candidate region, and a second correction step for detecting the character edge pixel region; A second correcting step for correcting the correction result corrected by the first correcting step by the correcting unit so that the inside of the character is a character pixel region; and an output speed and an original color by the recording device selecting unit. A recording method in which each of the output destinations of the image processing result of the image signal based on the recognition result by the digital image recognition device is different based on one of the document types. A recording device selection step of selecting a plurality or one of the plurality of recording devices, and image processing of the image signal selected by the recording device selection means based on a recognition result by the digital image recognition device. In accordance with the characteristics of the recording device that is the output destination of the result of the determination, it is determined whether to determine the character pixel area up to the inside of the character or only the character edge as the character pixel area, and based on the determination, the first A control step of controlling the correction means and the second correction means to be turned on or off at the same time, and when the control step determines the character pixel area to the inside of the character, And the first correction unit is turned on so as to perform correction in the direction in which the character pixel area is increased. When determining as a character pixel area, the second correction unit is turned off, and the first correction unit reduces the correction amount in the direction of increasing the character pixel area. To do.

  According to the present invention, when detecting the character edge pixel candidate region, the first correction unit corrects the detection result based on the pixel density based on the characteristics of the recording apparatus, and the first correction unit corrects the detection result. Since the second correction means for correcting the correction result so as to make the inside of the character a character pixel area based on a predetermined condition is controlled, appropriate character pixel area recognition can be performed for each printing apparatus.

  Further, by simultaneously turning on / off the first correction and the second correction, the recognition accuracy is improved when the inside of the character is recognized as the character pixel region, and the false detection is performed when the inside of the character is not recognized as the character pixel region. It can be controlled not to increase.

  In addition, since it is controlled whether or not the inside of the character is recognized as a character pixel area in accordance with the recording device or the color material of the recording apparatus, it is possible to perform optimum character pixel area recognition for each apparatus.

  The recognition of the character pixel area according to the characteristics of the color material in the output recording apparatus was realized as follows.

  FIG. 1 is a block diagram of the present embodiment. Information related to the recording device selection is acquired from the operation panel 26 by a user instruction, and the recording device selection 27 selects and determines the recording device. In this embodiment, it is assumed that the apparatus has two recording apparatuses of EP (electrophotography) and IJ (inkjet), and an image is formed by EP19, an image is formed by IJ22, or CMY is IJ22 and K is EP19. The user can select one of three methods for image formation. The operation panel 26 also has buttons for directly specifying, for example, output speed selection, document color selection, document type selection, and recording device, and assigns recording devices according to a table as shown in FIG. For speed priority, EP, for color photographic paper, IJ, for color text / photo, EP and IJ, and so on. An apparatus for automatically determining and selecting original color selection and original type selection is also known as an existing technology, and it is of course possible to select a recording apparatus by automatic selection.

  Image data is acquired from the scanner 11 and is subjected to LOG conversion 12 to convert the reflectance linear signal into a density linear signal. On the other hand, color determination 23 and character pixel region determination 24 are performed from the image data after the scanner input. In color determination 23, it is determined whether the pixel region is a chromatic pixel region or an achromatic pixel region. For example, when Max (R, G, B) −Min (R, G, B) <Th (Th is a predetermined threshold value), an achromatic pixel region is set, and otherwise, a chromatic pixel region is set. The character pixel area determination 24 will be described in detail later. In the black character pixel area determination 25, when it is an achromatic pixel area and a character pixel area, it is determined to be a black character pixel area.

The filter 13 performs a smoothing filter for removing noise in the pattern portion and an edge enhancement filter process for correcting the sharpness of the character portion on the image after LOG conversion. In the color correction 14, the R′G′B ′ signal is converted from the R′G′B ′ signal into the C′M′Y ′ signal corresponding to the toner color of the printer by the following formula:
C ′ = a0 + a1 × R ′ + a2 × G ′ + a3 × B ′
M ′ = B0 + B1 × R ′ + B2 × G ′ + B3 × B ′
Y ′ = C0 + C1 × R ′ + C2 × G ′ + C3 × B ′.

a0 to a3, B0 to B3, and C0 to C3 are color correction parameters. When R ′ = G ′ = B ′, achromatic colors are guaranteed so that C ′ = M ′ = Y ′. At this time, referring to the recording device selection signal, color correction parameters suitable for the recording device are appropriately selected and applied. In the UCR / black generation 15, a K signal is generated according to the C′M′Y ′ data (black generation), and the amount corresponding to K is reduced from C′M′Y ′ (under color removal (UCR)):
C = C′−K,
M = M′−K,
Y = Y′−K,
K = α × Min (C ′, M ′, Y ′).
In the UCR / black generation, the output signal of the black character pixel area determination 25 is referred to, 100% UCR (α = 1) is set for the black character pixel area, and α = 0.5 is set for the other areas.

  In output destination switching 16, the output destinations of CMY and K are respectively switched according to the recording device selection signal, and γ correction 17 and pseudo halftone 18 conforming to EP19, or γ correction 20 and pseudo halftone 21 conforming to IJ22 are selected. And an image is formed by a recording apparatus.

<Character pixel area determination>
FIG. 2 is a block diagram of the character pixel area determination 24. The minimum RGB value (where 0 is black and 255 is white) is selected by the minimum value selection 241, and the character edge pixel candidate area is detected by the character edge pixel candidate detection 242 based on the RGB minimum value signal. This is performed in the same manner as the edge region detection using the continuity of black pixels and white pixels described in Non-Patent Document 1. Next, the first correction is performed on the character edge pixel candidate area to detect the character edge pixel area, and the second correction is further performed to detect the character pixel area. The first correction and the second correction are configured to be simultaneously switched ON / OFF by a switch 245 in accordance with a recording device selection signal. It should be noted that the ON / OFF switching is the same in the control in which the switch is substantially turned OFF by parameter switching instead of direct switch switching.

<First correction>
In the first correction 243, correction according to the pixel density of the character edge pixel candidate is performed. In this embodiment, as shown in FIG. 3, 5 × 5 pixels centered on the target pixel are referenced, and if there are N or more (for example, N = 1) character edge pixel candidates, the target pixel is the character edge pixel. The correction which judges is carried out. This is correction in the direction of increasing the character pixel area.

<Second correction>
FIG. 4 is a block diagram of the second correction 244. From the character edge pixel region signal, the character candidate detection 2441 detects the inside of the character (inside the character edge pixel region) as a character candidate pixel region. Further, binarization is performed with a predetermined threshold value based on the RGB minimum value, and a high-density pixel area is detected, and an AND operation 2444 determines whether the area that is a high-density pixel area and a character candidate area is a character. It detects as a pixel area. Finally, the logical sum operation 2442 outputs the character edge pixel area or the character pixel area as the character pixel area. ON / OFF of the second correction can be controlled by ON / OFF of the pixel candidate detection 2441 in the character.

  FIG. 5 is a diagram for explaining pixel candidate detection 2441 among characters. M pixels (for example, M = 27) are referred to in the vertical and horizontal directions. If each of the upper, lower, left, and right regions with respect to the target pixel is AR1 to AR4, for example, both AR1 and AR3 have pixels in the character edge pixel region in order to determine whether the region is surrounded by the character edge pixel region. In this case, the target pixel may be determined to be in an area surrounded by the character edge pixel area. Similarly, when there are character edge pixel regions in both AR2 and AR4, or when there are character edge pixel regions in three directions in all four directions of AR1 to AR4 and in all directions, the target pixel is the character edge pixel region. It can be determined that it is in the area surrounded by. When the target pixel is a non-character edge pixel region and is a region surrounded by the character edge pixel region, the target pixel is determined as a pixel candidate region within a character. By increasing or decreasing the value of M representing the reference pixel range, it is possible to control which thickness of the character corresponds to the character determination.

Here, the meaning of simultaneously turning ON / OFF the first correction and the second correction will be described with reference to FIGS.
FIG. 6 shows a case where the first correction is turned off and only the second correction is turned on. In this way, the character edge pixel candidate region is relatively often interrupted (disconnected) in an image having a crossed shape. If the second correction is performed based on the character edge pixel candidate area where the line is broken, and the character pixel area is determined to be a character area, large detection errors will occur frequently as shown in the figure, which is unsightly. It becomes a playback image. Therefore, when performing the second correction, the first correction is performed as shown in FIG. 7 as a pre-processing to correct the line break, thereby preventing the detection defect in the character pixel area. However, when the second correction is turned off, on the contrary, if only the first correction is performed, erroneous detection at a portion (in the pattern) other than the original character edge is further expanded to be a character pixel region. Therefore, it is not preferable because the misdetection is only emphasized unnecessarily. Therefore, when turning off the second correction, control is performed to turn off the first correction as well. Even if the first correction is not turned OFF, the correction amount of the first correction is made minute (by controlling the correction amount so as to be smaller than the correction amount of the first correction when performing the second correction). You may not make it stand out.

  As described above, according to the present embodiment, when switching whether to determine the character pixel area including the inside of the character according to the characteristics of the recording apparatus in the apparatus for switching the recording formation in the electrophotographic system and / or the inkjet system, When the character pixel area determination (second correction) inside the character is turned ON / OFF, the first correction, which is the preprocessing, is also turned ON / OFF. To improve the recognition accuracy of the pixel area, when character determination is performed only on the character edge, the character pixel area should be recognized so that a natural reproduction image that is difficult to visually detect erroneous detection is obtained without increasing the character pixel area unnecessarily. it can.

  In the first embodiment, an example in which both EP and IJ recording apparatuses are provided in one apparatus is shown. However, one scanner and a plurality of printers (including EP and IJ) are connected via a network as shown in FIG. For example, the processing up to UCR / black generation is processed by a common processing unit, and the CMYK signal is transmitted and output to any printer. Region recognition can be performed.

  In the first embodiment, there are a plurality of recording apparatuses. However, the recording apparatus is not the plurality of recording apparatuses, and is the only recording apparatus. The second correction is performed according to the image processing mode and key operation (character priority, picture priority). Even in the device that controls ON / OFF, by controlling the first correction in conjunction with the ON / OFF of the second correction, it is possible to improve the accuracy of determination of characters according to each case, or to suppress the defect. The effect of is obtained.

  As shown in the second embodiment, the character pixel area recognition according to the present invention is applied to a network system, and an already owned printer can be used effectively.

It is a block diagram which shows the structure of the digital image recognition apparatus of this invention. It is a block diagram which shows the detail of the character pixel area | region determination 24. FIG. It is a figure which shows the surrounding pixel and reference pixel which are referred in order to determine whether an attention pixel is a character edge pixel. 12 is a block diagram showing details of a second correction 244. FIG. It is a figure explaining pixel candidate detection 2441 among characters. It is a figure explaining the meaning which turns off 1st correction | amendment and 2nd correction | amendment simultaneously. It is a figure explaining the meaning which turns off 1st correction | amendment and 2nd correction | amendment simultaneously. It is a figure which shows the criteria for selecting a recording device in the operation panel. 1 is a diagram showing a system in which a plurality of recording devices cooperate via a network. It is a figure which shows the conventional black character process.

Explanation of symbols

24 Character pixel area determination 25 Black character pixel area determination 243 First correction 244 Second correction

Claims (10)

In a digital image recognition apparatus that inputs an image signal and recognizes a character pixel region based on the image signal,
Detection means for detecting a character edge pixel candidate area in the document based on the input image signal;
Correcting the detection result based on a pixel density of the character edge pixel candidate region, and detecting a character edge pixel region;
Second correction means for correcting the inside of the character to be a character pixel region with respect to the correction result corrected by the first correction means;
A plurality of recording systems each having a different recording system, each of which is an output destination of an image processing result of the image signal based on a recognition result by the digital image recognition device based on any one of output speed, document color, and document type Recording device selection means for selecting a plurality of or one recording device from among the recording devices;
Depending on the characteristics of the recording device that is the output destination of the result of performing image processing on the image signal selected by the recording device selection means based on the recognition result by the digital image recognition device, the character pixel region is formed as far as the character inside. Control means for determining whether to determine only a character edge as a character pixel region, and controlling to switch the first correction means and the second correction means on or off simultaneously based on the determination When,
Have
The control means includes
When the character pixel area is determined as far as the inside of the character, the second correction unit is turned on and the first correction unit is turned on so as to perform correction in the direction of increasing the character pixel region. In the case where only the character pixel region is determined, the second correction unit is turned off and the first correction unit is turned off so as not to perform correction in the direction of increasing the character pixel region.
A digital image recognition apparatus. In a digital image recognition apparatus that inputs an image signal and recognizes a character pixel region based on the image signal,
Detection means for detecting a character edge pixel candidate area in the document based on the input image signal;
Correcting the detection result based on a pixel density of the character edge pixel candidate region, and detecting a character edge pixel region;
Second correction means for correcting the inside of the character to be a character pixel region with respect to the correction result corrected by the first correction means;
A plurality of recording systems each having a different recording system, each of which is an output destination of an image processing result of the image signal based on a recognition result by the digital image recognition device based on any one of output speed, document color, and document type Recording device selection means for selecting a plurality of or one recording device from among the recording devices;
Depending on the characteristics of the recording device that is the output destination of the result of performing image processing on the image signal selected by the recording device selection means based on the recognition result by the digital image recognition device, the character pixel region is formed as far as the character inside. Control means for determining whether to determine only a character edge as a character pixel region, and controlling to switch the first correction means and the second correction means on or off simultaneously based on the determination When,
Have
The control means includes
When the character pixel area is determined as far as the inside of the character, the second correction unit is turned on and the first correction unit is turned on so as to perform correction in the direction of increasing the character pixel region. In the case where only the character pixel region is determined, the second correction unit is turned off and the first correction unit reduces the correction amount in the direction of increasing the character pixel region.
A digital image recognition apparatus. When the recording device forms an image by electrophotography,
The second correction unit is turned off, and the first correction unit controls the first and second correction units so as not to perform correction in the direction of increasing the character pixel area. The digital image recognition apparatus according to 1. When the recording device forms an image by electrophotography,
The second correction unit is turned off, and the first correction unit controls the first and second correction units so as to reduce the correction amount in the direction of increasing the character pixel area. Item 3. The digital image recognition apparatus according to Item 2. When the recording apparatus forms an image by an inkjet method,
2. The first correction means controls the first and second correction means to turn on the second correction means and to perform correction in a direction in which the character pixel area is increased. Or the digital image recognition apparatus of 2. When the recording apparatus forms an image by an electrophotographic method and an inkjet method,
2. The first correction means controls the first and second correction means to turn on the second correction means and to perform correction in a direction in which the character pixel area is increased. Or the digital image recognition apparatus of 2. When the recording device forms an image with pigment ink and dye ink,
2. The first correction means controls the first and second correction means to turn on the second correction means and to perform correction in a direction in which the character pixel area is increased. Or the digital image recognition apparatus of 2. Color determination means for determining whether a chromatic pixel region or an achromatic pixel region based on the image signal;
A black character pixel region determining unit that determines whether or not the character pixel region is a black character pixel region based on the correction result determined by the second correcting unit and the determination result by the color determining unit;
Image processing means for performing image processing based on a determination result by the black character pixel region determination means;
The digital image recognition apparatus according to claim 1, further comprising: In a digital image recognition method for inputting an image signal and recognizing a character pixel region based on the image signal,
A detection step of detecting a character edge pixel candidate region in the document based on the input image signal;
A first correction step of correcting the detection result based on a pixel density of the character edge pixel candidate region by a first correction unit, and detecting a character edge pixel region;
A second correction step of correcting the inside of the character as a character pixel region with respect to the correction result corrected by the first correction step by a second correction unit;
Depending on the output speed, the document color, and the document type, the recording device selection unit varies the output destination of the result of image processing of the image signal based on the recognition result of the digital image recognition device. A recording device selection step of selecting a plurality or one recording device from a plurality of recording devices having a recording method;
Depending on the characteristics of the recording device that is the output destination of the result of performing image processing on the image signal selected by the recording device selection means based on the recognition result by the digital image recognition device, the character pixel region is formed as far as the character inside. A control step of determining whether to determine only a character edge as a character pixel region, and controlling to switch on or off simultaneously the first correction unit and the second correction unit based on the determination When,
Have
The control step includes
When the character pixel area is determined as far as the inside of the character, the second correction unit is turned on and the first correction unit is turned on so as to perform correction in the direction of increasing the character pixel region. In the case where only the character pixel region is determined, the second correction unit is turned off and the first correction unit is turned off so as not to perform correction in the direction of increasing the character pixel region.
A digital image recognition method characterized by the above.   In a digital image recognition method for inputting an image signal and recognizing a character pixel region based on the image signal,
  A detection step of detecting a character edge pixel candidate region in the document based on the input image signal;
  A first correction step of correcting the detection result based on a pixel density of the character edge pixel candidate region by a first correction unit, and detecting a character edge pixel region;
  A second correction step of correcting the inside of the character as a character pixel region with respect to the correction result corrected by the first correction step by a second correction unit;
  Depending on the output speed, the document color, and the document type, the recording device selection unit varies the output destination of the result of image processing of the image signal based on the recognition result of the digital image recognition device. A recording device selection step of selecting a plurality or one recording device from a plurality of recording devices having a recording method;
  Depending on the characteristics of the recording device that is the output destination of the result of performing image processing on the image signal selected by the recording device selection means based on the recognition result by the digital image recognition device, the character pixel region is formed as far as the character inside. A control step of determining whether to determine only a character edge as a character pixel region, and controlling to switch on or off simultaneously the first correction unit and the second correction unit based on the determination When,
  Have
  The control step includes
  When the character pixel area is determined as far as the inside of the character, the second correction unit is turned on and the first correction unit is turned on so as to perform correction in the direction of increasing the character pixel region. In the case where only the character pixel region is determined, the second correction unit is turned off and the first correction unit reduces the correction amount in the direction of increasing the character pixel region.
  A digital image recognition method characterized by the above.

Images