JP3333894B2 - Image processing device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for reading an image of a document by photoelectrically converting the image, and performing image processing such as background adjustment from an obtained digital image signal.

[0002]

2. Description of the Related Art An image processing apparatus for reading and processing a document image is provided in a full-color digital copying machine or the like and is connected to a digital printer. Documents to be processed by the image processing apparatus are various, and include a full-color document, a single-color document, a character / line drawing, a photograph, a map, a halftone document, a photographic paper photograph, a printed photograph, and the like. In order to read these various originals and copy them appropriately, it is desirable to switch the image processing content according to the types of the originals. For this reason, conventionally, a user selects a document type on the operation panel, and the image processing apparatus performs appropriate image processing according to the selection.

[0003]

However, manual selection of the document type by the user has a problem that the user is troublesome. Further, when various originals are stacked on the automatic original feeder, manual selection cannot be used. Also, both a full-color original and a monochrome color original must be able to be processed.

An object of the present invention is to provide an image processing apparatus capable of automatically determining the type of a document.

[0005]

An image processing apparatus according to the present invention has a three-dimensional image obtained by reading an original with a photoelectric conversion element.
A lightness signal generator for converting the R, G, B signals of the primary colors into lightness signals, first histogram generating means for generating a histogram of achromatic pixels in the document based on the lightness signals, and Second histogram creating means for creating a histogram of chromatic pixels in the original, original type determining means for determining the original type from the histograms created by the first and second histogram creating means, Processing switching means for switching image processing contents based on the determination result of the determination means. Also, preferably,
The image processing apparatus further includes an analysis unit that analyzes a distribution state of the histogram, and the document type determination unit determines a document type based on an analysis result by the analysis unit. In the image processing apparatus, for example, the analysis unit counts the number of the achromatic pixels and the chromatic pixels. Further, in the image processing device, for example, the analysis unit may store the histogram
Divide into multiple blocks for each brightness level,
The number of achromatic pixels and chromatic pixels is counted for each block.

[0006]

In the image processing apparatus, the type of the original is automatically determined for various originals, and the optimum image processing is selected. For this reason, the original is preliminarily scanned to obtain a lightness (VH) close to human relative luminous efficiency (brightness) from the read data, a lightness histogram is created, and the histogram is analyzed to determine the original type (preferably). Is a character line drawing,
(Including full color and monochrome color for photos and each). This determination is made based on the histogram. This determination result is preferably displayed by the display means. Then, image processing (for example, background adjustment, presence or absence of black character discrimination or gradation reproduction switching) suitable for each document type is selected and switched. As a result, image processing such as background adjustment can be uniquely performed in both the full color mode and the monochrome color mode.

[0007]

Embodiments of the present invention will be described below with reference to the accompanying drawings. (1) Overall Configuration of Digital Color Copier FIG. 1 shows the overall configuration of a digital full color copier. The original is read by the image scanner unit 30, and signal processing is performed by the digital signal processing unit 10. Printer unit 20
Prints out an image corresponding to the document image read by the image scanner unit 30 on paper in full color. In the image scanner section 30, the original placed on the original platen glass 31 is pressed by the pressing plate 39, but this is replaced when an automatic original feeder (not shown) is mounted. The original on the platen glass 31 is a lamp 32
And is guided to mirrors 33a, 33b, 33c, and forms an image on a linear full-color sensor (CCD) 36 by a lens 34. The full-color information red (R), green (G),
The signal is converted into a blue (B) component and sent to the signal processing unit 10.
By driving the scanner motor 37, the first
The slider 35 moves at a speed V, and the second slider 40 mechanically moves at a speed of V / 2 in a direction perpendicular to the electrical scanning direction of the color sensor, and scans the entire original. A white plate 38 for shading correction is provided at an end of the document table glass 31.

[0008] The signal processing section 10 electrically processes the read signal, and outputs magenta (M), cyan (C), yellow (Y),
It is separated into black (Bk) components and sent to the printer unit 20. With respect to one original scan in the image scanner unit 30, one component of C, M, Y, and Bk is sequentially sent to the printer unit 20, and one printout is completed by a total of four original scans. Sequential transfer method). Signal processing unit 10
The image signals of C, M, Y, and Bk sent therefrom are driven and modulated by the laser diode drive (PHC unit) in accordance with the image signal level. The laser beam scans the photosensitive drum 206 via the polygon mirror 215, the f-θ lens 216, and the return mirrors 217a and 217b.

The developing unit is composed of developing units 208a, 208b, 208c, 208d of C, M, Y, Bk. The developing units 208a, 208b, 208c, 208d come in contact with the photosensitive drum 206, The electrostatic latent image formed on the photosensitive drum 206 charged by the toner is developed with toner. On the other hand, the paper feed units 201a, 201
b, 201c to transfer drum 202
The transfer roller 205 conveys the image developed on the photosensitive drum 206 to the sheet by the transfer roller 205. After the four colors C, M, Y, and Bk are sequentially transferred in this manner, the separation chargers 209a and 209b
The paper is separated and conveyed by the fixing roller 210a,
The sheet passes through 210b and is discharged to the sheet discharge tray 211. 218 and 219 are reference position sensors for the transfer drum,
Reference numeral 20 denotes a separation claw for separating the sheet from the transfer drum.

(2) Processing of Image Signal in Signal Processing Unit FIGS. 2 and 3 show the overall configuration of image processing in the signal processing unit 10. The image scanner unit 30 forms an image of the reflected light from the document surface on the linear CCD sensor 36 by the micro optical system, and obtains an analog signal that is photoelectrically converted into R, G, and B color separation information. These signals are sent to the signal processing unit 10. The A / D converter 100 converts the 400 DPI image data photoelectrically converted by the CCD sensor 36 into 8 bits (256 gradations) for each of R, G, B color information by the A / D converter.
To digital data. Shading correction unit 1
02, the data obtained by reading the white plate 38 for shading correction before reading the document is independently stored for each of R, G, and B in order to eliminate light amount unevenness in the main scanning direction of the R, G, and B data. Is stored in a shading memory (not shown) as reference data, converted to a reciprocal number, and then multiplied by read data of document information to correct shading. The line-to-line correction unit 104 uses an internal field memory (not shown) in accordance with the scan speed (magnification on the sub-scanning side) to match the reading position of each of the R, G, and B sensor chips in the scan direction. Then, delay control is performed on the white data in line units, and R, G, and B data are output. The timing control unit 106 controls the CCD sensor 3
6. A / D converter 100, shading corrector 102
And the timing of the line-to-line correction unit 104 is controlled.

R, output from the line-to-line correction unit 104,
For the G and B data, the scaling / movement control unit 108
Using two line memories for scaling for each of G and B data,
Input / output is alternately performed on one line, and the writing timing and the reading timing are independently controlled to perform scaling / movement control in the main scanning direction. In this control, interpolation processing is performed before writing on the reduction side and after reading on the enlargement side in accordance with the scaling factor, thereby preventing image loss and rattling. Further, by this control, image repeat processing, enlarged continuous shooting processing, and mirror processing are performed. The histogram generation unit 110 generates a lightness signal from the inter-line corrected R, G, B data obtained by preliminary scanning the document information, and generates a histogram of the document. Based on the obtained histogram information, an automatic color selection judgment for judging whether the original is color / black and white, an original background level judgment for automatically skipping the original background level, and an original mode for copy operation
(Standard / Photo mode) settings are made automatically.

The HVC converter 114 converts the R, G, B data from the scaling / movement controller 108 into a lightness signal (V data) and a color difference signal (Cr, Cb data). Editing unit 1
Reference numeral 16 receives the V, Cr, and Cb data and performs an editing operation such as color change or coloring by detecting a closed loop area based on an optional specification of an editor. The bill recognizing unit 118 determines whether or not the document loaded on the document glass 31 is a bill, a securities, or the like, and instructs copy prohibition based on the result. Image interface 1
20 is V, sent through the first image selector 122,
Upon receiving the Cr and Cb data, the image data is transferred to an external device. In order to support various image data color signal interfaces, this block converts R, G, B signals from V, Cr, Cb signals, X, Y, Z signals and L ^* , a ^* , b, which are general color spaces. ^{* It} has a function of converting it into a signal or the like and outputting it to an external device, or conversely converting image data transferred from the outside into V, Cr and Cb signals. Further, C, M, Y, and Bk data to be transferred to the printer unit 20 are transferred to an external device,
There is also a function of receiving C, M, Y, and Bk signals from an external device and transferring them to the printer unit 20 side.

The image synthesizing section 124 includes a second image selector 1
After selecting one of the V, Cr, and Cb data output from the image interface unit 120 or the editing processing unit 116 via the H.26, image synthesis (inset / character synthesis) with the original data from the HVC conversion unit 114 is performed. Do. The HVC adjustment unit 128 controls the lightness (V: brightness), hue (H: hue), and saturation (C: V) of the V, Cr, Cb data from the image synthesis unit 124.
For the purpose of performing image adjustment corresponding to three human senses of "visibility", adjustment is performed independently for each of H, V, and C based on the designation of the operation panel. The AE processing unit 130 controls the background level of the document based on the brightness component based on the information obtained by the histogram generation unit. Inverse HVC converter 132
Performs data conversion again from V, Cr, Cb data to R, G, B data.

In the color correction section 134, first, the LOG correction section 1
36 converts the reconverted R, G, B data into density data (D
(R, DG, DB), while the monochrome data generation unit 138 generates tone data (DV) for monochrome reproduction after creating brightness data from the R, G, B data. UCR ・
The BP processing unit 140 uses the difference (MAX (R, G, B) -MIN (R, G, B)) between the maximum value and the minimum value of the R, G, B data as document saturation information, and outputs the DR, DG, DB minimum value (MIN (DR, DG,
DB)) as the under color components of the document, and performs under color removal and black printing in accordance with those values, and converts the DR, DG, and DB data to C
Create O, MO, YO, Bk data. The masking calculation unit 142 performs a masking calculation process for color correction,
The C, M, Y data (CO, MO, YO) after the R processing is converted into CMY data for color reproduction according to the color toner of the printer unit 20. The color data selection unit 144 outputs monochrome DV data as a monochrome copy mode when it is determined that the monochrome is determined by the designation of the operation panel or the ACS determination.
In the full color mode, according to the reproduction process signal (CODE), the masking operation processing data (C, M, Y data) is selected during the C, M, Y reproduction process, and the BP processing data (Bk data) is selected during the Bk reproduction process. Output.

On the other hand, based on the R, G, B data, the area discriminating section 146 calculates a difference between the minimum value (MIN (R, G, B)) and the maximum value and the minimum value (MAX (R, G, B) -MIN). (R, G, B)), the discrimination such as black character discrimination and halftone dot discrimination is performed, and the result (JD signal)
And the correction data (USM signal). Further, in order to achieve both the reproducibility of the image character area and the granularity of the image, an LIMOS signal for changing the image reproduction cycle is output to the printer. The MTF correction unit / sharpness adjustment unit 148
The input C, M, Y, and Bk data are subjected to edge correction, color fringing correction, smoothing processing, and the like based on the area determination result, thereby performing optimal correction of a copy image. Further, the gamma correction / color balance adjustment unit 150 adjusts the gamma curve and the color balance of C, M, Y, and Bk according to the density level information input from the operation panel. Thus, the C, M, Y, and Bk data subjected to various corrections are transferred to the printer together with the LIMOS signal for switching the gradation reproduction method, and a full color copy image of 400 DPI and 256 gradations is obtained. Here, the CPU 152 controls the signal processing unit 10, and the operation panel 154 performs input / output and display of data.

(3) Copy Mode Next, a copy operation mode of the full-color copying machine will be described. FIG. 4 shows a basic screen on the operation panel 154, and the user can set various modes. (A) Background processing (AE processing and manual setting) First, whether automatic exposure (AE) processing should be performed for the background processing.
You can choose to select one level of manual designation of the stage. In the AE process, five types of originals are determined from the original histogram information by the preliminary scan operation.
(Color standard (base white, with background color) / photo original and black-and-white standard / photo original), color standard (base white) or black-and-white standard original, perform brightness gradation complementarity as shown in FIGS. For other types (photos), the manual center level is automatically selected. At the time of manual designation, the contents are as shown in Table 2.

(B) Original mode An ACS (automatic color mode selection) mode or four types of original modes can be selected. When the ACS mode is selected, one of the four original modes is automatically selected based on the determination of the original type by the preliminary scanning operation. If it is determined that the document is a black-and-white document, one of the monochrome standard / photograph mode is automatically selected, and a monochrome mode copy operation in the black one-color reproduction process is performed. For color originals, color standard originals /
Automatically select one of the photo modes and select C, M, Y, Bk
A copy operation is performed by a full-color reproduction process using colors.
The same applies to the manual mode. However, when the monochrome standard / photograph mode is selected, the operation screen is changed to a monochrome mode operation screen (not shown), and the R, G, and B data for determining the grayscale data for monochrome as original parameters are set. Choose a mixing ratio. (In the ACS mode, the R, G, and B average sensitivity distributions are set as defaults, and in the manual mode, the relative luminous efficiency distributions are set as defaults.) The reproduction color can be selected from 16 colors including black. In addition, C, M, Y,
Color separation mode to reproduce Bk data on paper, negative / positive inversion, base color, image erase
Create function to adjust image, hue (hue) / saturation
Image quality adjustment function to simultaneously display five or more monitor images by simultaneously changing five types (brightness) / sharpness / gamma correction (contrast contrast) / color balance (RC / GM / BY / copy density) There is. In each case, detailed description is omitted.

(4) Preliminary Scan Operation In the copying machine of this embodiment, a preliminary scan operation is performed to
Analyze the results to determine the type of original, and
(AE) processing and automatic color selection (ACS) processing are performed. The document scanning unit is stopped on the side of the shading correction plate 38 opposite to the document reference position at the time of the main scan before copying. When the copy button is pressed on the operation panel 154,
After the lamp is turned on, it moves to read the correction data, scans the shading correction plate 38, and returns to the original reference position while creating histogram data of the original. The automatic exposure process and the automatic color selection process are determined from the created histogram data, and the main scan operation is started.
This shortens the first copy time.

(5) Generation of Histogram Next, generation of a histogram in the preliminary scanning operation will be described. FIG. 5 is a block diagram of the histogram detection unit 110. The histogram detection unit 110 obtains a histogram of R, G, and B data in the document area during the preliminary scan operation. Before starting the preliminary scan, the inside of the histogram memories 202 and 204 is initialized in advance.
The PU 152 writes “0” to the histogram memories 202 and 204 at addresses of all gradation levels (0 to 255). From the R, G, B data (8 bits) input from the line-to-line correction unit 104, the brightness creation unit 200
Calculates a lightness signal (VH) based on the following equation, and this is input to the first histogram memory 202 and the second histogram memory 204 as an address.

## EQU00001 ## VH = 0.31640625 * R + 0.65625 * G + 0.02734375 * B (1) The brightness signal obtained by this equation is approximated to human relative visibility (brightness). Here, the reason why the histogram is created is not the R, G, and B data, but the brightness data VH, in order to correct the data separated into the brightness / color difference signals by the automatic exposure process. This will be described in detail.

Based on the sampling interval setting value from the CPU 152, the sampling interval determining circuit 206 acquires intervals (thinning rates) into the histogram memories 202 and 204.
To determine. This is because creating a histogram of all dots of the maximum original size (A3) requires a maximum memory capacity of 32 Mbits, and the sampling interval is set to an appropriate value (main scanning direction: 1/8, sub-scanning direction: 1). / 4) Decimate,
It is 1 Mbit. FIG. 6 is a diagram illustrating an example of a sampling situation in generating a histogram. When a document (hatched portion) placed on the platen glass 31 is read, data at the position indicated by the triangle is sampled. Further, the document size is detected before the preliminary scan, and various signals are input from the timing control unit 106 to the sampling interval determination circuit 206. Here, a / HD signal (main scanning direction) indicating a document size area and a / VD signal
(Sub-scanning direction) is input to the sampling interval determination circuit 206, and the generation of the histogram is permitted only within the valid document area. Note that / TG is a main scanning synchronization signal, and has a cycle for each line. (In this specification,
A signal preceded by "/" means a negative logic signal. VCLK is a synchronous clock of image data.

When an address ADR is input to the histogram memories 202 and 204, the histogram data at that address is read out, the adders 208 and 210 add +1 to the data, and write the data again to the same address.
That is, the operation of the histogram is as follows.
The cycle is a read-modify-write cycle, and the addresses of the histogram memories 202 and 204 indicate the gradation level (lightness), and the data is the frequency (number) of each gradation level.
Is represented. When the preliminary scan is completed, the CP
U152 reads the frequency data of each gradation from the histogram memories 202 and 204. As will be described later, the contents of the automatic exposure operation, the automatic color selection operation, and the like are determined from the histogram data obtained by the preliminary scan operation.

The two types of histogram memories 202 and 204 are provided for automatic color selection processing. The first histogram memory 202 simply calculates the brightness histogram of the document, and the second histogram memory 204 calculates the histogram of achromatic dots in the document. Therefore, the minimum value circuit 212 and the maximum value circuit 21
4 detects the MAX value and the MIN value of the input R, G, B data, and obtains the difference between them by the subtraction circuit 216. Then, the comparator 218 determines that the difference is a predetermined level (SRE).
F) When it is determined that it is smaller, the writing of the brightness VH data into the second histogram memory 204 is permitted.
A small (MAX value−MIN value) of the R, G, B data indicates that the document data is achromatic data. Therefore, the second histogram memory 2
04 means that a histogram is calculated only for achromatic data. In the first histogram memory 202, / WE is always at the “L” level, and writing is possible for all pixels.

(6) Automatic Color Selection Process (ACS) The automatic color selection mode is to determine whether the original loaded on the original platen glass 31 is a black-and-white original or a color original, and automatically determine the copy mode. Mode. As a result, the copy speed of a black-and-white document can be increased because the image needs to be reproduced in the reproduction process of only Bk. In particular, when the automatic document feeder is used, even if a black-and-white document and a color document are mixed, an appropriate copy can be obtained without the operator's awareness.

The determination of automatic color selection will be described below. h1 (n) represents frequency data at the lightness level n of the first histogram created in the first histogram memory 202, and h2 (n) represents the second histogram memory 2
The frequency data at the lightness level n of the second histogram created in step 04 is shown. The CPU 152 calculates the second histogram 2 from each frequency (h1 (n)) of the first histogram 202.
04 is subtracted from each frequency (h2 (n)), and the third histogram
(h3 (n) = h1 (n) -h2 (n)) is created. this is,
4 shows a histogram of a chromatic portion of a document. As shown in FIG. 7, two types of histogram memories 202 and 204 are provided.
For example, the following amounts can be analyzed from the two histograms created in (1). From the first histogram memory 202, W = the number of dots in the background (white) area (n = μ1 to 255) in the document, M = monochrome halftone (gray) area (n = μ2 to μ)
1) the number of dots, B = the number of dots in the black area (n = 0 to μ2), and S = the sum of the total frequencies of the first histogram = the total number of pixels in the document size. Further, the second histogram memory 204
, C = the number of dots in the color area (n = σ1 to σ2).
That is,

(Equation 2)

In the automatic color selection, based on the brightness data,
The document type is determined from the ratio between the achromatic color and the chromatic color. Specifically, using S and C obtained from the histogram, the ratio of chromatic color dots in the document is obtained, and color copy is performed.
Judge whether to make black and white copy. As described above, C is the number of dots in the color area (n = σ1 to σ2), and Sn is the total number of pixels in the document size. Therefore, C / S corresponds to the ratio between chromatic color and (chromatic color + achromatic color). That is, if the judgment formula (C / S) is equal to or less than the reference value,
Since there are few chromatic colors, a black-and-white copy mode is set. If it is larger than the reference value, a full-color copy mode is set because there are many chromatic colors. Here, by using S as the denominator, the determination of the automatic color selection, in particular, the influence of the document size can be ignored.

FIG. 8 shows a flow of the automatic color mode selection of the CPU 152. First, the histogram creation unit 110
As a result, the first and second histogram memories 202 and 20
Then, a brightness histogram is created in step S4 (step S10).
0). Next, the above C and S are obtained from the histogram (step S102), and the ratio C / S is calculated (step S1).
04). If C / S is larger than a predetermined threshold (YES in step S106), it is determined that the document is a color document (step S108). Otherwise (NO in step S110), it is determined that the document is a monochrome document. (Step S112).

(7) Judgment of Document Type Further, the CPU 152 first determines the histogram memories 202 and 204 as an initial stage of the automatic exposure (AE) process.
And the results of the automatic color selection (ACS) (see FIG. 8), the following five types of originals (see Table 1) are determined. (a) Black-and-white photo originals (black-and-white photographs, black-and-white high-definition dot printing, etc.) (b) Black-and-white standard originals (black-and-white characters and line drawings, etc., with relatively white background) (c) Full-color photo originals (color silver halide photographs) , Color high-definition halftone printing, etc.) (d) Color standard original (white background) (Original relatively white original including color characters and color line drawings) (e) Color standard original (with background color) Manuscript with)

Automatic color selection is also obtained from the histogram, and the document type is determined by analyzing the histogram (see FIG. 9). The concept of determining the document type is as follows. For color documents and black-and-white documents, as described above for automatic color selection, the number of achromatic pixels and the number of chromatic pixels are obtained from the histogram, and the number of chromatic pixels and the entire achromatic color are calculated. If the ratio of the number of pixels is larger than a certain reference value, it is determined that the document is a color document. Otherwise, it is determined that the document is a monochrome document. The photographic original and the standard original can be determined from the distribution of the histogram. The standard original is an original mainly composed of characters, and the histogram shows a binary distribution as shown in FIGS. Here, the case where the background is not white is also considered. If it shows a binary distribution, it is determined that the document is a standard document; otherwise, it is determined that it is a photographic document. More specifically, the number of pixels in the density range (black side) is compared with the number of pixels in the vicinity of white from the histogram. In the case of a color standard document, it can be determined from this that the background is white. In the case of a color original, it is necessary to distinguish between a standard original with a base color and a color photo original, so that a histogram whose distribution in the histogram is wide and average is determined to be a color photo original.
Otherwise, it is determined that the original is a color standard original with a base color. Specifically, the determination is made based on the difference between the maximum value and the minimum value in the histogram.

FIG. 9 shows a flow of the document type discrimination by the CPU 152. First, from the data h1 (n) and h2 (n) of the first and second histogram memories 202 and 204, various frequency sums G25, G24, G23, G22, G
21, G20, G35, G34, G33, G32, G31, G
30 and furthermore, the background level a (the gradation level indicating the maximum frequency when the output data ID is 0.4 or less in the second histogram memory 204) and the character level b (the output data ID 0.6 or more in the second histogram memory 204) To obtain the gradation level indicating the maximum frequency (Step S)
200).

(Equation 3) As shown on the left side of FIG.
5 corresponds to the output data ID, and these values correspond to the six ranges of the output data (0.2 or less, 0.2 to 0.4, 0.4).
~ 0.6, 0.6-0.8, 0.8-1.1, 1.1 or more), h2 (n) or h2 (n) depending on whether the brightness data is larger or smaller than the threshold value SREF. h3 (n) (= h1
(n) -h2 (n)). In FIG. 10, the ranges indicated by C, M, Y, R, G, and B indicate the ranges of VH in the corresponding colors.

Next, photos (a) and (c) and standard manuscripts (b) and
In order to distinguish between (d) and the standard original with a background color (e), a determination is made between a photographic original and an original with a background color. First, based on the processing result of the aforementioned automatic color selection (ACS),
(B) and the color originals (c), (d), and (e) can be determined (step S202). If the determination result of the automatic color selection is color (Y in step S202)
ES), and proceeds to step S204 to determine the type of the color document (here, α2 represents a threshold). First, the output data (ID) achromatic color of 0.4 or more and the output data (ID)
When the sum of the frequencies with 0.2 or more chromatic colors (corresponding to a portion other than the white background) accounts for a small percentage of the total document frequency (Sn) (YES in step S204), there are many white background portions. It is determined that the document is white (base white) (b) (step S206). Then, for image processing, the background adjustment
Automatic exposure (AE) processing is set, black character discrimination processing is set, gradation reproduction switching processing is set, and the original mode is set to the color standard mode (step S208). If output data
If the sum of the frequencies of the achromatic color of (ID) 0.4 or more and the chromatic color of the output data (ID) 0.2 or more accounts for a large proportion of the total document frequency (S) (NO in step S204), It is determined whether the ratio occupied by the frequency sum in the frequency block having a chromatic color is very high (step S210, where α
3 represents a threshold value). Specifically, the ratio between the difference between the maximum frequency sum and the minimum frequency sum in the chromatic frequency blocks G30 to G34 and the total document frequency is calculated. If this ratio is not very high (step S210: NO), since the image data is not average over all lightness gradations, it is determined that the base of the color standard document is colored (a) (step S212). In the image processing, the background adjustment is set at the center of the standard manual setting, the black character discrimination processing is set, the gradation reproduction switching processing is set, and the original mode is set to the color standard mode (step S21).
4). Otherwise (YES in step S210),
Since the image data is average over all brightness gradations,
It is determined that the original is a color photograph original (c) (step S21).
6) Regarding the image processing, the background adjustment is set at the center of the manual photograph setting, the black character discrimination processing is not set, the gradation reproduction switching processing is not set, and the original mode is the color photograph mode (step S218).

On the other hand, if the determination result of the automatic color selection (ACS) is not a color (NO in step S202),
The process proceeds to step S220 (where α1 represents a threshold value). If the sum of the frequencies of the achromatic colors having the output data (ID) of 0.4 or more in the total document frequency (S) is small (YES in step S220), it is determined that the document is a monochrome photograph (e) (step S222). For the image processing, the background adjustment sets the center of the photo manual, does not set the black character discrimination processing, does not set the gradation reproduction switching processing, and sets the original mode to the monochrome standard mode (step S224). Otherwise (NO in step S220), since there are many white background portions, it is determined that the original is a black-and-white standard original (d) (step S22).
6) Regarding image processing, background adjustment is performed by automatic exposure (AE).
The process is set, the tone reproduction switching process is set without setting the black character discrimination process, and the original mode is set to the monochrome standard mode (step S228).

Finally, the respective document type determination results are displayed on the basic operation screen (FIG. 4) of the operation panel 154 (step S230). Without this display, the user may be uneasy because the result of determination of the document type is not known. Therefore, by displaying the document type on the operation panel 154, the user can immediately understand the determination result. Through the above processing, the determination of the document types (a) to (e) and the setting of the image processing corresponding thereto have been made. Table 1 shows the contents of the automatic color selection (ACS), the image processing mode, and the document mode for each document type. Table 2
Indicates the setting of the background processing in various modes.

[0033]

[Table 1]

[0034]

[Table 2]

(8) HVC Conversion and HVC Adjustment The copying machine of this embodiment converts image data into HVC data. The HVC conversion unit 114 includes a matrix calculator that converts R, G, and B data into a signal in a color space including a lightness signal (V) and two types of color difference signals (Cr, Cb).

(Equation 4) The three attributes of the color including hue, lightness, and saturation are obtained as follows using the V, Cr, and Cb signals.

(5) Lightness (Value) = V Saturation (Chroma) = (Cr ² + Cb ² ) ^1/2 (5) Hue (Hue) = arctan (Cb / Cr) Conversion into such signals V, Cr, Cb The reason for this is that by performing processing similar to human sensation, high image quality is realized, and the processing (image synthesis, automatic exposure processing, HVC adjustment) performed later is facilitated.

The output of the HVC conversion unit 114 is transferred to a processing unit after the image synthesizing unit 124,
In step 6, image editing such as color change is performed. On the other hand, the bill recognizing unit 118 determines whether the original is a bill, a securities, or the like, and instructs prohibition of the copy operation. Then, in the image synthesizing unit 124, the V, Cr, and Cb signals output from the HVC conversion unit 114 are temporarily input to the delay memory 115, and synchronized with the image signal from the editing processing unit 116.
Then, the image synthesizing unit 124 synthesizes an image based on the output data (V, Cr, Cb) of the delay memory 115 and the output data (V, Cr, Cb) of the image selector 126. Typical synthesizing methods include watermark synthesizing, inlay synthesizing, and character synthesizing, but a detailed description thereof will be omitted.

As shown in FIG. 11, the HVC adjustment unit 128
Is provided for the image quality adjustment mode. The HVC adjustment unit 128 includes a matrix calculator 128a that receives the V, Cr, and Cb data and performs the following matrix calculation processing so that image adjustment can be performed independently for each of the H, V, and C signals. .

(Equation 6) Here, q is a saturation adjustment coefficient, and θ is a hue adjustment coefficient. These coefficients are output from the HVC adjustment control unit 129 and sent from the image quality monitor control unit to Mdata (3).
Bit) as a switching signal, the image signal and the eight kinds of coefficient groups are selected in real time. In this way,
Adjustments similar to human sensations are made, facilitating image adjustments according to operator preferences.

(9) Automatic Exposure Processing (AE) The background processing will be described as an example of image processing according to the document type. Conventionally, in a full-color copying machine, the automatic exposure process is operated only in a monochrome mode because there is a possibility that the copy may have a color balance different from that of the original. However, even in the full-color mode, processing such as black character discrimination is introduced, and when black-and-white and color mixed originals are reproduced clearly, the original background level is automatically set like automatic exposure processing to prevent show-through etc. The need for optimal control has arisen. In the present embodiment, a brightness signal (VH) approximated to the relative luminous efficiency (brightness) of a human is created, and a histogram is generated and a document type is determined. As a result, the color tone of the full-color original is not changed, and
An automatic exposure process is performed to copy the image by automatically controlling the background level of the original without considering the color portion. That is,
Once the image signal is converted from R, G, B signals to V, Cr, Cb signals, the data is subjected to automatic exposure processing,
By converting the signals into G and B signals, the full-color mode and the monochrome mode are uniquely processed, so that the background level can be optimized. Further, even in the full-color mode, no processing is applied to the color component signals Cr and Cb, so that the color balance does not change due to the automatic exposure processing.

More specifically, the background adjustment is as follows.
This is performed by automatic automatic exposure processing or manual setting. In the basic operation screen of FIG. 4, the user can select whether to perform the automatic exposure process or to select one of eight levels of manual specification. In the automatic exposure process,
5 from original histogram information by preliminary scan operation
Different types of originals (color standard (base white, with background color) / photo original and black and white standard / photo original) are performed (see FIG. 9). As shown in FIG. 9, if the original is a color standard original (white background) or a black-and-white standard original, the brightness gradation complementarity as shown in FIGS. 12 and 13 is performed. Is automatically selected.

The automatic exposure (AE) processing unit 130 removes the background based on the document type information obtained by the histogram generation unit 110. Here, the document type (b) and
For (d), a look-up table memory (AE table) 131a for brightness V is used to convert a brightness signal (V _in ) before automatic exposure processing to a brightness signal after automatic exposure processing.
The (V _{o ut)} determined by the following correction formula, performs the brightness correction.
That is, for a monochrome standard manuscript,

V _out = 256 * (V _in −b−8) / {(a−8) −b} (7) Further, for a color standard original (base white),

V _out = 256 * (V _in -8) / (a-8) (8)

Here, a indicates the background level, and b indicates
At the character level. In other words, as shown in FIG. 12, for a black-and-white standard original, the background is skipped and, at the same time, light characters such as pencils are darkened and copied in a legible manner. Therefore, lightness between a + 8 and b is expanded to 0-255, data V _in the above from a + 8 from below and b is discarded. On the other hand, in a color standard document, the background is simply skipped. For color standard document, as shown in FIG. 13, data V _in between 8 and b is expanded to 0-255, data on the b is discarded. In this embodiment, the level at which the background is skipped is set to 0 to 8.

In a monochrome / color standard original, the background level a and the character level b are obtained as follows. The following values are obtained from the brightness histogram h1 (n) of the entire document in the first histogram memory 202. First, in order to determine the background level of the document, n = 136 to 255 (ID 0.
4 or less), a gradation level m at which h1 (n) obtains the maximum frequency is obtained. Then, a = m−8 and the background brightness is set to 255. Similarly, only in the case of a black and white document, n = 0 to 120 (ID 0.
(4 or more), a gradation level 1 at which h1 (n) obtains the maximum frequency is obtained. Then, b = 1 + 8, and the brightness of the character portion = 0. The reason for setting a = m−8 is that the histogram distribution is a normal distribution having a certain variation near the level m, so that the variation is ± 8 and the gradation near the level m is reliably skipped. is there. Similarly,
The reason for setting b = l + 8 is to ensure that the gradation near level 1 is black. The reason why the control is not performed by b in the color original standard mode is that characters are not always black.

Here, since the color components of Cr and Cb are passed through (in the AE tables 131b and 131c, D _in =
D _out ), only the density information (V) is controlled without changing the color information of the original. Therefore, the background level can be automatically adjusted without changing the color information of the color original. Color difference signal C
For the r and Cb components, no correction is performed (V _out =
V), the color balance does not collapse.

Further, in the manual setting (cancellation of the automatic exposure processing) set on the operation panel 154, the brightness correction for changing the background level value can be performed. This mode differs between monochrome / color and photo / standard modes, and has seven manual setting values. -1 to -4 are centered around ± 0, and the bases are +1 and +2. (See FIG. 4). As for detailed setting contents, in the color standard mode, the monochrome standard mode and the photograph mode, as shown in Table 2, each level of manual setting +2 to -4 is set.

As another image processing corresponding to the document type, there is a black character determination processing (color blur correction) in the MTF correction unit 148. As shown in Table 1, this process is performed on a color standard document. This is performed to optimize the image reproduction of black characters when a color image and a monochrome image are mixed. First, in the portion determined by the region determination unit 146 as an edge portion of a black character, C, M, Y
The component data is attenuated, and the Bk component data is Bk1
The edge emphasis is performed while slightly thickening the character by adding the data of the lightness V to the data of 00%.

(10) Inverse HVC Conversion In the inverse HVC conversion unit 132, the V, Cr, and Cb signals are again
In order to convert the signals into R, G, and B signals, the inverse matrix operation of the above-described matrix is performed as follows, and R, G, and B are output.

(Equation 9) The above-described processing of the scan data is performed by converting the data into brightness data. Here, by performing the inverse HVC conversion, the subsequent data processing such as color correction can be performed on the data of the three primary colors.

[0047]

According to the present invention, for both a color original and a black and white original,
Since various document types can be automatically determined, the user does not need to select a document type. Further, even if various originals are mixed and loaded on the automatic original feeder, it is possible to automatically determine the original type and perform the optimum image processing for each of them.

[Brief description of the drawings]

FIG. 1 is a schematic sectional view showing an entire digital color copying machine.

FIG. 2 is a block diagram of a part of a signal processing unit.

FIG. 3 is a block diagram of a remaining part of the signal processing unit.

FIG. 4 is a block diagram of a histogram generation unit.

FIG. 5 is a diagram showing a state of sampling in histogram generation.

FIG. 6 shows various quantities obtained from a histogram.

FIG. 7 is a flowchart of automatic color selection.

FIG. 8 shows various quantities obtained from a histogram.

FIG. 9 is a flowchart for determining an image type.

FIG. 10 is a diagram for explaining the relationship between a brightness signal and various signals (G25 to G35).

FIG. 11 is a diagram illustrating an HVC adjustment unit, an automatic exposure (AE) processing unit, and an HVC inverse conversion unit.

FIG. 12 is a graph showing a change in original brightness distribution before and after AE processing for a monochrome color standard original.

FIG. 13 is a graph showing a change in original brightness distribution before and after AE processing for a color standard original (base white).

[Explanation of symbols]

110 Histogram generator, 152 CPU, 20
0 lightness creation unit, 202 first histogram memory for all pixels, 204 second histogram memory for chromatic colors, 208, 210 adder for integration, 212 to 218
Circuit for chromatic color identification.

Continuation of front page (56) References JP-A-2-144567 (JP, A) JP-A-2-884 (JP, A) JP-A-6-152962 (JP, A) JP-A-2-2055984 (JP, A) JP-A-6-284281 (JP, A) JP-A-6-205204 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04N 1/40-1/409 H04N 1/46 H04N 1/60

Claims (4)

(57) [Claims] 1. A brightness signal generator for converting R, G, and B signals of three primary colors obtained by reading a document by a photoelectric conversion element into a brightness signal, and an achromatic color pixel in the document based on the brightness signal. First histogram creating means for creating a histogram, second histogram creating means for creating a histogram of chromatic pixels in the document based on the lightness signal, and first and second histogram creating means. An image processing apparatus comprising: a document type determining unit configured to determine a document type based on the histogram; and a processing switching unit configured to switch image processing content based on a determination result of the document type determining unit. 2. The apparatus according to claim 1, further comprising an analyzing unit configured to analyze a distribution state of the histogram, wherein the document type determining unit determines a document type based on an analysis result by the analyzing unit. The image processing apparatus according to any one of the preceding claims. 3. The image processing apparatus according to claim 2, wherein the analysis unit counts the number of the achromatic pixels and the chromatic pixels. Wherein said analyzing means, said each histogram
Divide into multiple blocks for each brightness level,
4. The image processing apparatus according to claim 3, wherein the number of achromatic pixels and chromatic pixels is counted for each block.