JPH03214970A - Picture reader - Google Patents

Abstract

PURPOSE:To select an output in response to the propriety of confirmation or recognition for the content of processing applied to a picture data by providing an attribute storage means storing designated attribute for picture processing generated for each sectional area being each division of an original and an attribute discrimination means discriminating a specific attribute for each sectional area in matching with the read of the original and generating the discrimination attribute data. CONSTITUTION:A ROM 20a storing a processing program, a register 20b for storing tentatively various flags and status data in the execution of a program and a RAM 20c are incorporated in a CPU 20. A binarizing attribute discrimination section 251 discriminates the binarizing attribute for each divided area E, that is, discriminates whether the divided area E corresponds to a character picture or an intermediate tone picture and outputs a 1-bit discrimination attribute data alpha0 representing the result of discrimination. Moreover, a color attribute discrimination section 252 discriminates whether or not the divided area E corresponds to a specific color designated as an object of color edit and outputs a 1-bit discrimination attribute data alpha1.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image reading device that performs image processing on a signal obtained by reading a document and then outputs an image signal corresponding to the document. [Conventional technology] Conventionally, digital copying machines, facsimile machines, and OCR
(optical character reader), optical filing system,
As in various display devices, images are created, stored, and
An image reading device (Image League) that reads a document by dividing it into pixels is used as an image input means for display devices. In this type of image reading device, the read image information is quantized and treated as digital image data, so various image processing can be easily performed. Generally, it is configured to perform image processing to improve image quality, such as edge enhancement to make the image clearer, smoothing to make the image more natural, and image processing to edit the image. has been done.

Image editing includes trimming to extract only a specific area within a document, masking to whiten out a specific area, reverse to invert the negative and positive (black and white inversion), and color conversion to replace a specific color with another color. etc. are known.

For example, in an image reading device that allows image editing,
As shown in Japanese Unexamined Patent Publication No. 181570/1989, an attribute memory is provided for storing attribute data corresponding to the content of image processing, and this attribute memory stores information for each divided area in which the manuscript is divided into fine sections. Attribute data is stored in advance, and image processing based on the attribute data is executed in real time as the image is read.

According to this, the content of image processing can be specified for each segmented area, and complex image editing is possible.

By the way, as is well known, in image processing (hereinafter referred to as "binarization processing") for binarizing multi-valued data according to the density of each pixel of a document, from the quality of the reproduced image,
In general, so-called simple binary processing, which compares the multivalue level of one pixel with a certain darkness value level, is suitable for line images such as characters, and dithering is suitable for halftone images such as photographs. Pseudo gradation processing using the error diffusion method or the like is suitable.

Therefore, when reading a document containing a mixture of line images and halftone images, it is necessary to It is necessary to switch the binarization process. In order to automatically switch the 2W processing, image processing methods have conventionally been developed that incorporate discriminating means for discriminating binarized attributes using various methods disclosed in, for example, Japanese Patent Application Laid-Open No. 62-39978. There is a reading device. Note that an image reading device capable of color editing such as color conversion has a built-in means for determining color attributes of an image.

[Problem to be solved by the invention]

In conventional image reading devices, only image data after image processing is output, and attribute data indicating the content of the image processing is not output.

Therefore, there has been a problem in that operators or various devices that manually process image data cannot easily confirm or recognize what kind of image processing has been performed on image data from an image reading device. .

That is, especially when checking the state of image processing based on the discrimination results inside the image reading device, such as binarization processing, the operator should check the image data after obtaining the image based on the image data using the printer or display device. By carefully comparing and observing the image and the manuscript, it is necessary to indirectly check whether there are any errors in the classification results for each segmented area, which is extremely troublesome. For example, in OCR consisting of an image reading device and a host computer (host), especially when the value of image data is "0" indicating a blank area, the host determines whether the image data corresponds to a white area due to masking. It is necessary to recognize whether the image data corresponds to a blank area of the document or a blank area of the manuscript.For this recognition, the image data from the image reading device is once developed on the image memory, and the address (coordinates) of the image memory and the designated area for image editing are stored. (specified coordinates) must be calculated in order to associate them with (specified coordinates).

SUMMARY OF THE INVENTION In view of the above-mentioned problems, it is an object of the present invention to provide an image reading device that can select output depending on whether confirmation or recognition of the content of image processing performed on image data is necessary.

[Means to solve the problem]

In order to solve the above-mentioned problems, the present invention provides an image reading device configured to read a document by dividing it into pixels, perform image processing on the obtained signal, and output image data corresponding to each pixel. , an attribute storage means for storing designated attribute data for image processing generated for each segmented area into which the manuscript is divided; and determining a specific attribute for each segmented area in accordance with reading of the manuscript; Attribute discriminating means for generating discriminating attribute data; image processing means for performing image processing based on at least one of the specified attribute data and the discriminating attribute data; and output control means for outputting at least one of the data.

[For production]

The attribute storage means stores designated attribute data for image processing generated for each divided area of the document.

The attribute determining means determines a specific attribute for each segmented area as the document is read, and generates determined attribute data.

The image processing means performs image processing based on at least one of the specified attribute data and the discrimination attribute data. The output control means outputs at least one of image data corresponding to each pixel of the subdivided document and attribute data that defines image processing by the image processing means. [Example] Hereinafter, an example of the present invention will be described with reference to the drawings.

FIG. 2 is a cross-sectional front view showing the general configuration of the image reader 1. As shown in FIG.

The image reader 1 has a document table glass 2 attached to the top surface of a rectangular parallelepiped housing, on which a document of up to A3 size can be placed, and the placed document is held down by a document cover 3. ing. A reference pattern 14 consisting of a white image for shading correction is provided at the front end of the document platen glass 2 in the scanning direction. Inside the housing, there is an optical system that is arranged to scan the original image in the direction of arrow M5 (sub-scanning direction) below the original platen glass 2, and an electric system that generates image data according to the density of the original image. A circuit section 12 is provided. The optical system includes an exposure lamp 4, a reflecting mirror 5, a first slider 13 having a mirror 6, a second slider 13a having mirrors 7 and 8, a main lens 9, and the like. 1st
When the moving speed of the slider 13 is V, the second slider 13a is driven and controlled to move at a speed of v/2.

The scanning light that has passed through the main lens 9 is incident on the image sensor IO attached to the support member 11, and is converted into an electrical signal. The image sensor 10 is composed of a plurality of CCD chips arranged continuously in the main scanning direction, and is capable of reading an original at a resolution of 400 pixels/inch. Each CCD chip has a large number of light-receiving elements arranged in a row, and each light-receiving element is divided into three regions, one divided region being R (red), G (
A spectral filter is provided on the surface of each light-receiving element to receive light of one of three colors: green) and B (blue). One light-receiving element corresponds to one pixel of the original image, and each light-receiving element outputs a photoelectric conversion signal to the electric circuit section 12 in accordance with the intensity of reflected light for one color of the pixel. FIG. 1 is a block diagram of the electric circuit section 12.

The electric circuit section 12 converts the photoelectric conversion signal from the image sensor IO into R, G, . A color separation unit 21 that separates the signals of each color of B and performs predetermined amplification, and a digitization processing unit 22 that A/D converts (quantizes) the analog signals of each color and outputs 8-bin image data Dm7 to Dm0. , a shading correction unit 23 for correcting uneven light amount in the main scanning direction of image data Dm7 to Dm0 and variations between bits of the image sensor 10.
, an image discrimination section 25 that discriminates binarization attributes and color attributes, a density conversion section 26 that performs density level adjustment and γ conversion according to the density characteristics (T characteristics) of externally connected equipment, and an image editing process.
An image processing section 28 that performs image processing including value conversion processing, an output control section 29 for controlling data output, an attribute memory 30 that stores designated attribute data a2 to 0 (described later), an attribute data output control section 31, and a clock generation unit. Circuit 41: a line memory 24 that stores one line of image data Dml7 to Dml10 output from the shading correction section 23, a synchronization signal generation section 40 that outputs various synchronization signals to each section, and an exposure lamp 4.
The lamp controller 4a controls the lighting of the scanner, the driver 16a drives the scanner motor 16 for scanning, and the CPU 20 controls all of these components. CPtJ20 has ROM2 that stores the processing program.
0a, a register 20b and a RAM 20c are built-in for temporarily storing various flags and status data during program execution.

The CPU 20 communicates with an external host device equipped with operator operation means to exchange various commands (control instruction code data) and data indicating the operating state (status) of the image reader 1. Based on the received command, designated attribute data a2 to a0 defining image editing and binarization processing is generated prior to reading the document, and is stored in the attribute memory 30.

Note that the synchronization signals outputted by the synchronization signal generation section 40 include a horizontal synchronization signal Hsync outputted for each main scanning line, a pixel clock signal SYNCK that serves as a reference for data transmission timing for each pixel, and the image reader 1 There is an enable signal VD indicating the valid period of data output from the .

FIGS. 3(a) and 3(b) are diagrams showing the structure of the attribute memory 3o. As shown in FIG. 3(b), the attribute memory 30 uses the CPU 20 for each segmented region E, which is composed of an 8×8 donma matrix obtained by segmenting the image G into 8-pixel units in the main scanning direction and the sub-scanning direction. Specified attribute data a2 generated by
It is configured to be able to store ~0. In the image reader l, specified attribute data aO, al
．． a2 is data for specifying binarization processing, negative/positive inversion, and trimming, respectively. FIG. 4 is a block diagram of the image discrimination section 25.

The image discrimination section 25 includes a binarized attribute discrimination section 251, a color attribute discrimination section 252, a discrimination attribute data output section 253, and the like. The binarization attribute determination unit 251 determines the binarization attribute for each segmented area E, that is, determines whether the segmented area E corresponds to a character image or a halftone image. , 1 showing the discrimination result
Output the bit discrimination attribute data α0. Further, the color attribute determination unit 252 determines whether or not the classification nod area E corresponds to a specific color designated as a color editing target, and outputs 1-bit determination attribute data α1. In addition, in order to enable attribute discrimination for each segmented area E, the image discrimination unit 25 includes image data Dml7 to Dml7 for 8 lines.
A line memory (not shown) for storing 10 is provided. The discrimination attribute data αO is set to "0" when the classification target area E to be discriminated corresponds to a character image, and is set to "1" when it corresponds to a halftone image. In addition, the discrimination attribute data α
l is set to "1" when the divided area E to be determined corresponds to a specific color, and r■, when it corresponds to a color other than the specific color.
It is said that FIG. 5 is a block diagram of the image processing section 28. The image processing unit 28 includes a filtering unit 281, a trimming/masking unit 282, a simple binarization processing unit 283,
It is composed of a pseudo halftone processing section 284, a data selection section 286, a selector 285, a negative processing section 287, and the like.

The image processing unit 28 receives image data D from the density conversion unit 26.
m27 to m20 are input serially in the order of pixel arrangement. The manually generated image data Dm27 to Dm20 are first subjected to processing to improve image quality, such as edge enhancement and smoothing, in the filtering unit 281. Next, in the trimming/masking unit 282 according to the specified attribute data a2, when the data a2 is rlJ, rOj corresponding to the blank area is forcibly set as masking processing. data a
When 2 is "0", the data is output as is from the trimming/masking unit 282 (data through).

The image data output from the trimming/masking section 282 is binarized by a simple binarization processing section 283 and a false halftone processing section 284, respectively, and binary image data DA, D are obtained.
The selector 285 selects and outputs either binary image data DA or DB according to the output data DSELO from the data selection section 286.

The data selection unit 286 includes specified attribute data aO for controlling the binarization process along with the above-mentioned discrimination attribute data α0 obtained by automatically discriminating the binarization attribute, and the value of this data aO The value of output data DSELO is determined according to. That is, if data aO is "0",
The discrimination attribute data αO is output as is as the output data DSELO, and if the data aO is “1”, data obtained by inverting the discrimination attribute data αO is output. In other words, in the image processing unit 28, when the data aO is “O”, the designation of binarization processing from the outside is defaulted, and the binary image data is processed based on the automatic discrimination of the binarization attribute by the image discrimination unit 25. is output, and when data aO is “l”, 2
Binary image data that has been subjected to a binarization process that is the opposite of the result of automatic discrimination of digitized attributes will be output. If the designated attribute data a1 is "0", the negative processing unit 287 outputs the binary image data input from the selector 285 as is, and if the data a1 is "1", it outputs the binary image data with the values reversed. Output data. When outputting data from the negative processing section 287, it is sequentially outputted in the form of parallel image data V I DEO 1 to 8 for each 8 pixels. Note that the image processing unit 28 includes 8 lines of image data Dm2 in order to enable image discrimination for each segmented area E.
A line memory (not shown) for storing numbers 7 to 20 and 8-bit serial/parallel conversion means are provided.

FIG. 6 is a block diagram of the attribute data output control section 30.

The attribute data output control unit 30 outputs the discrimination attribute data α0. α
1 and specified attribute data a2 to 0 to output 5-bit attribute data a4 to O.

Attribute data a4 and a3 are discrimination attribute data αO, respectively.
, α1.

FIG. 7 is a block diagram of the output control section 29, and FIG. 8 is a timing chart showing the output operation of attribute data a4 to a0.

The output control unit 29 includes a frequency divider 401 that outputs a signal 1/8·Hsync obtained by dividing the signal SYNCK by 1/8, a latch circuit 402 that latches the signal VD according to the signal Hsync, and output control data CO from the CPU 20. The output is controlled by the selectors 403 and 404 which respectively output the output control signals SO and Sl in accordance with ~1, and the signals so and st, and each selects the image data V I DEO 1 to 8 or the attribute data a4 to 0 as the output data DATA. It consists of data latch circuits 405, 406, etc. The signal 1/8・Hsync is the data latch circuit 405
The clock terminal (CK) of 406 is input to the selector 404 as it is as a selection input, the selection input is inverted by an inverter 407 and added to the selector 403, and the output synchronization signal DSYNC is input via an inverter 408. It is output to the outside as

When signal VD becomes active ('LJ), frequency divider 40
1. Signal 1/8 from selectors 403 and 404, respectively
・HSync, So, SL are output, output data DA
TA output becomes possible.

With reference to FIGS. 8 and 1, the output data DATA
The output operation of is explained.

When scanning of the original is started, image data Dm17 to Dm10 are input to the image discrimination section 25 and the density conversion section 26 in order from those corresponding to the first pixel of the first line.

As mentioned above, in order to perform image discrimination for each segmented area E spanning eight lines, the image discrimination unit 25 stores image data Dm17 to lO for eight lines, so up to the eighth line. The output of the image discriminator 25 during scanning is invalid data. In Figure 8, invalid data is shaded. Therefore, the signal VD is set to "L" in synchronization with the signal Hsync indicating the start of scanning on the 9th line,
From the time when the image data Dm17 to Dm10 for the first 8 pixels of the 9th line are input from the image discrimination unit 25,
In synchronization with the signal 1/8.HSync, the discrimination attribute data α0. αl is output.

The attribute memory 30 outputs specified attribute data a2 to 0 in synchronization with the output of the discrimination attribute data α0 and α1. On the other hand, since a delay of 8 dots occurs in the density conversion section 26, the image data Dm27-20 of the 9th line is input to the image processing section 28 with a delay of 8 dots relative to the image discrimination section 25. The image processing unit 28 also processes 8 as described above.
Since a line delay occurs, the image processing for the I-th line image data Dm27-20 is completed when the 9th line image data Dm27-20 is input. Furthermore, in the C image processing unit 28, a delay occurs due to serial/parallel conversion, and the image data V I DEO 1 to 8 corresponding to the first line are processed every 8 pixels with a further delay of 8 dots from the input point of the 9th line. are output in parallel. The attribute data output control unit 3l outputs image data VIDEO1.
Attribute data a4 to 0 in synchronization with the output of ~8
Output.

Image data V I DEO 1~ output in this way
8 and attribute data a4 to a0 are each output control unit 29.
The data latch circuits 405 and 406 allow the signal 1/8
-Latched at the rising edge of Hsync. However, in the data latch circuits 405 and 406, the latch data is not always output from the output terminal (Q), and the output is the signals so and st according to the total 2-bit output control data CO to l from the CPU 20. (active low).

That is, when the output control data CO and CI are both "0", the signal So is set to "L" (ground level),
The signal SL is set to rH'' (Vcc level), the data latch circuit 405 is enabled to output, and the output terminal of the data latch circuit 406 is placed in a high impedance state.

As a result, the output control unit 29 outputs the output data DAT.
Image data V I DEO 1 to 8 are output as A in scanning order.

Further, when the output control data co is r1, and the output control data C1 is "0", the signals 50, 51 are respectively rH, , r L J, and the output data DA
Attribute data a4 to a0 are sequentially output as TA. Furthermore, when the output control data CO is "0" and the output control data Cl is "l", the signals so, st
is r}iJ or “L” according to the signal 1/8・Hsync.
”, and image data VID is output as output data DATA.
EO1-8 and attribute data a4-O are output alternately. That is, in the image reader 1, the output control data CO. By setting CI, image data VIDEO1 can be used as usual.
~8, output attribute data a4~0 for correcting the area specification in image editing, or display the image and its attributes at the same time to confirm image discrimination. The output can be selected depending on the purpose of use of the output data DATA and the data processing form of the receiving side, such as alternately outputting the data V I DEO 1 to 8 and the corresponding attribute data a4 to 0. be.

Note that, as shown in FIG. 8, in the image reader 1, image data V I DEO 1 to 8 and attribute data a
In the case of outputting 4 to 0, both data are output alternately at half the cycle of the case of outputting only image data V I DEO 1 to 8, so the attribute data a4 to
The output time of image data D1 to D8 is not extended by outputting 0. Further, the attribute data a4-0 are outputted via the same signal line as the image data VIDEO1-8,
There is no need to provide a dedicated signal line for outputting attribute data a4-0. Next, the operation of the image reader I will be explained with reference to the flowcharts shown in FIGS. 9 to 18. Figure 9 shows CPU20
This is a main flowchart schematically showing the operation of. When the power is turned on and the program starts, first,
Perform initial settings in step #1. In step #2, the presence or absence of a command from the host device is checked.

If there is a command, the type of command is determined (step #3), and depending on the type, reading processing (step #4) is performed.
Reading mode specification processing (step #5), attribute specification processing (step #6), output data specification processing (step #7)
).

Thereafter, other processing (step #8) is executed, the process returns to step #2, and the processing from step #2 to step #8 is repeated.

FIG. 10 is a flowchart of reception processing, and FIG. 11 is a flowchart of transmission processing.

These routines are interrupt routines and are executed at appropriate times in response to access from the host device. In the reception process shown in FIG. 10, first, code analysis of the received signal (step #11) is performed, and step #l2
If reception of the command is confirmed, the received command is stored in a predetermined area in the register 2Ob (step #13). If the received signal requests status notification (step #14), data indicating the status such as standby state or wait state is stored in the register 20.
b and send it to the host device. Furthermore, if the received signal does not correspond to any of the predefined commands and status requests, code data indicating a reception error is transmitted (step #16).

In the transmission process shown in FIG. 11, wait until the previous transmission is completed and the next transmission becomes possible (step #21), and set the code data to be transmitted in the register 20b (step #22).

Thereafter, in step #23, the presence or absence of code data to be subsequently transmitted, that is, whether or not transmission is necessary, is checked, and if transmission is necessary, the process returns to step #21. FIG. 12 is a flowchart of the initial setting of step #1 in FIG.

First, rWAIT, which indicates that preparation for read scanning is in progress, is set as the status. In other words, rWA is placed in the status area in register ZQb.
Store data corresponding to I TJ (step #31
).

Next, in step #32, a self-test is performed to check whether each part operates normally.

At step #33, a check is made to see if there is an abnormality, and if there is an abnormality, the process moves to step #37 and an inoperable code is sent to the host device.

If there is no abnormality, the process proceeds to step #34, where each part is initialized. At this time, "0" is written into the attribute memory 30 as designated attribute data aO, al, a2. Therefore, if the designated attribute data a2 to a0 are not rewritten thereafter, the image processing unit 28 will not perform image editing such as trimming and negative/positive reversal, and
The value conversion process will be performed based on the discrimination attribute data α0.

Further, the density converter 26 is set to have a standard density, and the output controller 29 is set so that image data V I DEO 1 to 8 and attribute data a4 to 0 are output alternately. The inputs of selectors 403 and 404 are selected.

After these initializations, the first slider 13 is moved to the home position in step #35, and when the movement is completed, the status is changed from "WAITJ" to rREADYJ indicating a standby state in step #36.

FIG. 13 is a flowchart of the above-mentioned reading process.

First, rBUs indicating that the status is being read.
Y" (step #41), and the exposure lamp 4 is turned on (step #42).

Next, turn on the scanner motor 16 (step #43L).
Wait until the first slider 13 reaches the shading position, that is, the position directly below the reference pattern 14 (step #
44).

When the first slider 13 reaches the reference pattern 14, the reference pattern 14 is read for shading correction;
The reference image data (white data) is stored in the line memory 24 (step #45). Next, in step #46, the first slider 13 waits for the first slider 13 to reach the leading edge position of the document, and in step #47, the synchronization signal generator 40 is turned on to output a synchronization signal. As a result, each part operates according to the synchronization signal, and as described above, after the start of scanning for the 9th line, valid image data V I DEO 1 to 8 and attribute data a4 to 0 are alternately output.

Wait until the scanning of the original is completed, that is, the first slider 13 reaches the rear end position of the original (step #48), turn off the synchronization signal generator 40 (step #49), and turn off the scanner motor l6 ( Step #50), and the exposure lamp 4 is turned off (Step #51).

Then, the scanner motor 16 is reversed so that both sliders 1
3. Return 13a (step #52) and wait for the first slider 13 to return to the home position (step #5
3), turn off the scanner motor 16 (step #54L)
Finally, in step #5 5 the f-task is 'READ'.
YJ and t. FIG. 14 is a flowchart of the reading mode designation process in step #5 of FIG.

In step #61, the status is set to rWAITJ, and in step #62, the parameters included in the command are checked, and the concentration is specified according to the parameters (step #62).
63) or other specifications such as specifying the output target device (step #64). And step #6
Step 5 returns the status to rREADYJ. FIG. 15 is a flowchart for specifying density in step #63.

In step #71, it is checked whether the density specification method is automatic. If the designation method is other than automatic, the process moves to step #74, where the density conversion unit 26 is set according to the parameters sent from the host device based on the designation operation by the operator. If automatic is determined in step #71, preliminary scanning is performed to detect the density of the document in step #72.
Image data Dm17 sequentially stored in line memory 24
Load ~10 into RAM20c at the appropriate time. and,
After detecting the density of the document based on the data in the RAM 20C in step #73, the process proceeds to step #74, where the density conversion section 26 is set according to the detection result.

FIG. 16 is a flowchart of the attribute specification process in step #6 of FIG. First, set the status to rWAITJ (step #81
), the correctness of the specification is checked (step #82).

If the specification is incorrect, for example, if an area outside the reading range is specified or if there is an error in the order of coordinate specification, the process moves to step #85 and an error code is sent to the host device. If the specification is correct, . Execute attribute data write processing to write designated attribute data ao, al, a2 to the attribute memory 30 (step #83), and set the status to rR.
EADY (step #84). FIG. 17 is a flowchart of the attribute data writing process in step #83. In step #91, the specified seed from the host device? Check and step #92 to step #98 depending on the type.
Execute each process.

That is, when automatic discrimination of binarization attributes is specified, in step #92, for the specified segmented area,
Let designated attribute data aO be r■.

If the binarization attribute is specified in advance, step #93
Then, for the specified classification nod area E, the specified attribute data a
Let O be "1".

When a blur designation, that is, a designation to not perform black and white reversal, is made, in step #94, the designated attribute data a1 is set to "0" for the designated segmented area E. On the other hand, if a negative designation, that is, a designation to perform black and white reversal, is made, in step #95, the designated attribute data al is set to "1" for the designated segmented area E. If trimming is specified, step #9
In step #97, the specified attribute data a2 is set to "1" for areas other than the specified segmented area E, and if masking is specified, the specified attribute data a2 is set to "1" for the specified segmented area E in step #97. "l". If trimming/unmasking is specified, in step #98, the specified attribute data a2 is returned to "0" for areas other than the specified segmented area E. FIG. 18 is a flowchart of the output data designation process in step #7 of FIG. In this routine, the type of output data DATA is first checked in step #101, and steps #102 to #104 are performed depending on the type. If output of only image data V I DEO 1 to 8 is selected, in step #102, output control data C
Let both o and CI be rQJ. When outputting only attribute data a4 to a0 is selected, step #103 is executed, output control data CO is set to "l", and output control data C1 is set to "O". Furthermore, if the output of both the image data VI DEO 1 to 8 and the attribute data a4 to 0 is selected, the output control data co is set to "0" in step #104, and the output control data CI is assumed to be r1. According to the embodiment described above, the discrimination attribute data α0 corresponding to the discrimination result of the binarized attribute is output, so when character recognition is performed on the host device, the character image is determined based on the discrimination attribute data αO. Recognition processing can be performed only on the segmented area E corresponding to , which improves recognition efficiency. Also,
When using the image reader l as an image reading means for a filing system or facsimile, image data V I DEO 1 to
8, it becomes possible to perform data processing such as data compression suitable for line images or halftone images. [Effects of the Invention] According to the present invention, output can be selected depending on whether confirmation or recognition of the content of image processing performed on image data is necessary.

[Brief explanation of drawings]

FIG. 1 is a block diagram of the electric circuit section according to the present invention, FIG. 2 is a cross-sectional front view showing the general configuration of the image league, and FIG.
The figure shows the configuration of the attribute memory, Figure 4 is a block diagram of the image discrimination unit, Figure 5 is a block diagram of the image processing unit, Figure 6 is a block diagram of the attribute data output control unit, and Figure 7 is the output FIG. 8 is a block diagram of the control section, FIG. 8 is a timing chart showing the output operation of attribute data, and FIGS. 9 to 18 are flow charts showing the operation of the image reader. DESCRIPTION OF SYMBOLS 1... Image reader (*image reading device), 25... Image discrimination part (attribute discrimination means), 28... Image processing part (image processing means), 29... Output control part (output control means) ), 30・
...Attribute memory (attribute storage means), aO. al, a'l
... Specified attribute data, a4-0... Attribute data, α0. α1...Discrimination attribute data, VI DEO 1-8...Image data, Dm27-
20...Image data (signal), E...Divided area.

Claims (1)

[Claims] (1) In an image reading device configured to read a document by dividing it into pixels, perform image processing on the obtained signal, and output image data corresponding to each pixel, each segmented area into which the document is divided. Attribute storage means for storing designated attribute data for image processing generated in each case; and Attribute determination means for determining a specific attribute for each of the divided areas in accordance with reading of the document and generating discriminating attribute data. and an image processing unit that performs image processing based on at least one of the specified attribute data and the discrimination attribute data, and an output that outputs at least one of the image data and the attribute data that defines the image processing. An image reading device comprising a control means.