JP3411821B2 - Image reading 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 reading apparatus which irradiates an image recorded on a film with light and reads the image using a line sensor.

[0002]

2. Description of the Related Art Conventionally, in this type of image reading apparatus, an exposure measurement is performed prior to a prescan and a main scan for detecting image data to be output to a personal computer, and an optimum exposure which is a charge accumulation time of a line sensor is performed. The time is calculated. That is, in the exposure measurement, a scan is performed at a coarser pitch than the main scan, and the optimum exposure time is obtained by analyzing the image data detected by exposing the line sensor in a relatively short exposure time. This image data analysis is performed based on a histogram showing the frequency (pixel number) distribution of the pixel value levels of the image data obtained by the line sensor.

[0003]

The histogram is stored in memory. That is, in the memory, the number (frequency) of pixels having the pixel value is stored in the address corresponding to the pixel value (for example, the brightness value). However, in the calculation of the optimum exposure time, not all the histogram is used, but only the data in a predetermined range.

It is an object of the present invention to save the memory capacity by reducing the amount of histogram data stored in the memory.

[0005]

An image reading apparatus according to the present invention is a pixel data reading unit for reading pixel data forming one image, and pixel data within a predetermined range from the pixel data read by the pixel data reading unit. A histogram creating means for creating a histogram showing the pixel value distribution of each pixel data and a pixel data reading means
An optical sensor provided to detect the image and a histogram
The maximum value that is smaller than the maximum pixel value in the distribution range by a predetermined amount.
Optimal exposure time for optical sensor based on large effective value
Select the desired exposure measurement method and histogram distribution range.
The minimum effective value that is greater than the minimum pixel value
Pixels output from the optical sensor based on the large effective value
Color correction parameters for performing color correction on data
Equipped with a color correction parameter calculation unit that calculates
The specified range of the histogram used in the
Predetermined range of histogram used by parameter calculation means
And are in different ranges from each other.

In the above-mentioned image reading apparatus, the histogram creating means is, for example, a color correction parameter computing means.
When calculating the correction parameter, pixel data from the maximum value of the range of pixel values to the first boundary value smaller than the maximum effective value and the pixel data larger than the minimum effective value and larger than the first boundary value. A histogram showing the distribution of the values of each pixel data may be created for the pixel data from the second boundary value that is smaller than the minimum boundary value to the minimum value of the range of possible pixel values. Exposure measurement by means of exposure measurement means
Is determined based on the proper output value of the optical sensor
Third regarding only a small pixel data than the boundary value, even if a histogram showing the distribution of the value of each pixel data
Good .

The image reading apparatus may include a data storage means for storing in the memory the data relating to the histogram created by the histogram creating means.

[0008] In the image reading apparatus, histogram creation unit, the maximum chromatic from the maximum value of the range of possible pixel values
Pixel data up to the first boundary value smaller than the effective value and the minimum
When creating a histogram showing the distribution of pixel values of each pixel data from the second boundary value that is larger than the effective value and smaller than the first boundary value to the minimum value of the range that the pixel value can take The data storage means stores the data relating to the histogram of the pixel data from the maximum value to the first boundary value and the data relating to the histogram of the pixel data from the second boundary value to the minimum value at consecutive addresses of the memory. May be configured as follows.

In the above-mentioned image reading apparatus, the most preferable
The maximum valid value is the degree from the maximum pixel value in the histogram.
It is the maximum pixel value when the total number is greater than or equal to a predetermined threshold.
Yes, the smallest valid value is the smallest pixel value in the histogram
Is the minimum when the sum of frequencies from
It is a pixel value.

[0010]

[0011]

DETAILED DESCRIPTION OF THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an image reading apparatus according to an embodiment of the present invention.

The read original M used in this image reading apparatus is a transparent original (film), and a color image is recorded on this film. The document M to be read is intermittently transported in the direction of arrow A by the document transport mechanism 10. A light source 20 is provided above the passage of the document M to be read, and an imaging lens 31 and a line sensor 30 are provided below the light source 20.
Is provided. The light source 20 is controlled by the light source drive circuit 41, and the line sensor drive circuit 42 controls the image detection operation by the line sensor 30. The document transfer mechanism 10, the light source drive circuit 41, and the line sensor drive circuit 42 operate according to a command signal output from the system control circuit 40.

The image data read from the line sensor 30 is amplified by the amplifier 43 and converted into a digital signal by the A / D converter 44. The digital image data is temporarily stored in the memory 46 after being subjected to shading correction in the image processing circuit 45. This image data is read from the memory 46 and subjected to predetermined arithmetic processing such as color correction and gamma correction. And the image data is
The signal is converted into a signal in a predetermined format in the interface circuit 47, and is output to the computer 60 provided outside the image reading apparatus via the output terminal 48. Image processing circuit 45 and interface circuit 4
7 is controlled by the system control circuit 40.

In this embodiment, all the operations of the image reading apparatus are controlled by the computer 60, but by connecting the switch 49 to the system control circuit 40,
The operation of the image reading apparatus may be controlled by operating the switch 49.

FIG. 2 shows the document transfer mechanism 10, the light source 20 and the line sensor 30. The document M to be read is a frame 1
The frame 11 is a film-shaped transparent original supported by 1, and a frame 11 is fixed to a plate-shaped stage 12 by a stopper 13. An opening (not shown) is formed on the stage 12 at a position corresponding to the document M to be read. Stage 12
A rack 14 is formed on the side end surface of the pinion 1. The rack 14 has a pinion 1 mounted on the output shaft of a document feed motor 15.
It meshes with 6. The document feeding motor 15 is driven under the control of the system control circuit 40, and the position of the document M to be read is controlled.

The light source 20 is located above the stage 12,
Light emitting elements 21R, 21G, and 21B that emit blue (B), green (G), and red (R) light are periodically arranged in this order, and this arrangement is arranged according to the purpose. It can be changed. Although only six light emitting elements are shown in FIG. 2, more light emitting elements may be provided or less light emitting elements may be provided. These light emitting elements 21R, 21G, 21B are supported by an elongated support member 22 extending in the width direction of the stage 12, and a cylindrical lens 23 extending in parallel with the support member 22 is arranged between the support member 22 and the stage 12. ing. That is, the light emitted from the light emitting element 21 is condensed by the cylindrical lens 23 and is linearly irradiated onto the document M to be read.

The line sensor 30 is located below the light source 20 with the stage 12 interposed therebetween, and is provided in parallel with the light source 20 and the cylindrical lens 23. That is, the line sensor 30 extends in a direction substantially orthogonal to the direction in which the document M to be read is transported. An imaging lens 31 is provided between the line sensor 30 and the stage 12. Imaging lens 3
1 extends in parallel with the line sensor 30 and is constituted by a rod lens array. Therefore, when the light source 20 irradiates the document M to be read, the image recorded on the document M to be read is imaged on the light receiving surface of the line sensor 30 via the imaging lens 31.

FIG. 3 shows the configuration of the light source 20, the line sensor 30 and the like when a reflective original is used as the original M to be read. In this configuration, the light source 20 and the cylindrical lens 23 are arranged below the document M to be read together with the line sensor 30 and the imaging lens 31. That is, the light emitted from the light source 20 is applied to the lower surface of the document M to be read via the cylindrical lens 23, and the light reflected by the document M is imaged on the line sensor 30 via the imaging lens 31. .

FIG. 4 is a flow chart showing an image reading routine executed in the image reading apparatus.

At step 110, an exposure measurement is performed. That is, in the state where the light source 20 is turned on, the document M to be read is intermittently transported by the document transport mechanism 10 at a pitch coarser than that of the main scan executed in step 160. During this intermittent transfer, the line sensor 30 is exposed for a certain exposure time to detect image data for one screen. In this exposure measurement, the light source 20
The light emitting elements 21R and 21R
G and 21B are controlled to be turned on in a predetermined order,
R, G, and B image data are detected. Then, the exposure time, that is, the optimum exposure time for obtaining the optimum output level for each of the R, G, and B images is calculated by a conventionally known method.

Next, in steps 120 to 136, processing for calculating the color correction parameter is performed. The image signal read from the read document M is a negative /
Positive conversion is applied. At this time, since the background color of the document M (negative film) to be read is colored, color correction is necessary.
That is, the color correction parameter is used for color correction performed when negative / positive conversion is performed on the image signal.

In step 120, a rough scan is performed. Similar to the exposure measurement, the rough scan is performed with the light source 20 turned on. While the document M to be read is intermittently transported at a coarser pitch than the main scan, the line sensor 30 is used.
Is exposed according to the optimum exposure time, and pixel data for one screen is detected. Then, based on this pixel data, a histogram H1 showing the distribution of pixel values as shown in FIG. 5 is obtained for each of the R, G, and B color components. That is, the horizontal axis of FIG. 5 shows the pixel value (image signal level), and the vertical axis shows the appearance frequency (frequency) of the pixel value.

In step 130, the maximum effective value D, which is a pixel value smaller than the maximum pixel value Q2 by a predetermined amount, is obtained for the histogram H1 of each of the R, G, and B color components. In step 133, the minimum effective value d, which is a pixel value that is larger than the minimum pixel value Q1 by a predetermined amount, is obtained for each histogram H1. Next, at step 136, the color correction parameter is calculated using the maximum effective value D and the minimum effective value d. The maximum effective value D and the minimum effective value d
The method of obtaining and the calculation of the color correction parameter will be described later.

At step 140, prescan is performed. The reading pitch of the image in the pre-scan is, for example, finer than that of the coarse scan and coarser than that of the main scan. In the pre-scan, the line sensor 3 is exposed for each of the R, G, and B color components by exposure for the optimum exposure time.
Pixel data is detected by 0, and this pixel data is A /
It is converted into digital pixel data by the D converter 44.

The pixel data of each color component is image processing circuit 45.
In, the color correction and the negative / positive conversion are performed using the color correction parameters obtained in step 136.
The normalized data thus obtained is gamma-corrected and then output to an external display device or the like via the input / output terminal 48, and a color image is displayed on the screen of this display device.

In step 150, it is determined whether or not the main scan is started. The user can determine whether to start the main scan by looking at the image displayed on the screen of the display device. For example, by clicking a button for starting the main scan using the mouse, the process moves from step 150 to step 160, and the main scan is executed. When the main scan is not started, error processing and the like are performed, and processing for changing the conditions of the main scan and prescan and the like are performed.

In step 160, the main scan is executed, the image of the document M to be read is read at a predetermined fine pitch, the pixel data is detected as in the prescan, and
Converted to digital pixel data. This digital pixel data is subjected to color correction and negative / positive conversion, and converted into normalized data. The normalized data is converted into a preset value by referring to, for example, a look-up table (LUT) stored in the memory 46, and subjected to correction such as gamma correction. The gamma-corrected data is output from the input / output terminal 48 to the display device via the interface circuit 47, and the program ends.

Next, color correction and negative / positive conversion will be described. FIG. 5 is a histogram H showing the distribution of pixel data values (pixel values) obtained by the line sensor 30.
It is 1. This pixel data is subjected to color correction and negative / positive conversion according to the following equation (1), and converted into normalized data. The histogram showing the distribution of the signal level in the normalized data is as shown in FIG. 8, and the meaning of the equation (1) will be described with reference to the histograms in FIGS.

[0029]

[Equation 1]

In the equation (1), D is the maximum effective value of the histogram, and d is the minimum effective value of the histogram. L1 and L2 are a lower reference value and an upper reference value of a look-up table (LUT) for performing gamma correction and the like. That is, the lower reference value L1 is the minimum pixel value that can be referenced in the LUT, and the upper reference value L2 is the maximum pixel value that can be referenced in the LUT.

The pixel value (input value) in the histogram H1 of FIG. 5 is subjected to offset subtraction according to the following equation (2).

[0032]

[Equation 2]

That is, the pixel value is subtracted by the minimum effective value d by the term of (input value-d), and the histogram H1 is shifted to the left side of FIG. By adding the lower reference value L1 to the pixel value after the shift, the histogram H2 shown in FIG. 6 is obtained.

Although the substantial distribution width W1 of the histogram H2 is L1 to D, it is converted into the distribution width W2 according to the equation (3).

[0035]

[Equation 3]

That is, a value (X1-L) obtained by subtracting the offset L1 from the pixel value X1 in the histogram H2 of FIG.
The distribution width W2 shown in FIG. 7 is expanded by multiplying 1) by the coefficient (L2-L1) / (D-d). By adding the lower reference value L1 to this, the human gram H3 shown in FIG. 7 is obtained.

Next, from the upper reference value L2, X2 in the equation (3) is calculated.
By subtracting and adding the lower reference value L1 (see the equation (1))), the histogram is inverted left and right in the figure. That is, the pixel data is subjected to negative / positive conversion, and the histogram H4 of the normalized data shown in FIG. 8 is obtained.

As described above, in the histogram of the pixel data, the distribution range of the minimum effective value d to the maximum effective value D is converted from the upper reference value L2 to the lower reference value L1 by the equation (1). The transformation of this distribution range is R, G and B
By independently performing each color component, the color components are balanced and can be reproduced in a natural color tone originally possessed by the read image. That is, the pixel data is color-corrected by adjusting the distribution range of the histogram of each color component. In the equation (1), the shift amount (that is, the minimum effective value) d and the coefficient (L2-
L1) / (D-d) is a color correction parameter, and these color correction parameters are obtained for each of the R, G, and B color components.

The normalized data obtained by performing the color correction and the negative / positive conversion is subjected to the gamma correction and the like by referring to the LUT as described above, and is transferred to the display device to display the color image on the screen. Displayed above. Figure 9
Indicates a histogram H5 corresponding to the gamma-corrected pixel data. As shown in this figure, the histogram H5 is distributed within the range of LUT [L1] to LUT [L2]. LUT [L1], LUT [L2]
Are table values corresponding to the lower reference value L1 and the upper reference value L2, respectively. Further, FIG. 10 shows a configuration example of the LUT.

FIG. 11 is a flowchart of the rough scan subroutine executed in step 120 of FIG. In step 200, the histogram data stored in the memory until then is cleared. In step 210, the stage 12 is moved to set the end of the document M to be read to the initial position corresponding to the light source 20, and the initial value of the parameter Y corresponding to the position of the document M to be read is set to zero. It

In step 220, the light emitting element 21R of the light source 20 is turned on to irradiate the document M to be read. The line sensor 30 is exposed for the optimum exposure time of R by the light transmitted through the read document M. As a result, one line of pixel data is detected. This pixel data is A /
It is converted into digital pixel data by the D converter 44 and is temporarily stored in the memory 46.

In step 230, the subroutine for creating the histogram H1 (see FIG. 13) is executed, and the data of the histogram H1 is generated based on the pixel data of one line obtained in step 220 and stored in the memory 46. It

In the present embodiment, when the histogram H1 is created, all pixel data obtained for one image is not used, but the maximum value MX (1023 for 10-bit data) to the first boundary value BX is used. Pixel data up to and the pixel data from the second boundary value BI to the minimum value MI (0) are used. The first and second boundary values BX and BI are preset values, respectively, so that the first boundary value BX is statistically smaller than the maximum effective value D, and the second boundary value BI is It is set to be statistically larger than the minimum effective value d.

FIG. 12 shows an example of the memory map of the histogram H1 in the comparative example and this embodiment. In this example, the gradation of the input pixel data is 10 bits.

In the comparative example in which a histogram is created using all the pixel data, the frequency K [0] of pixel values at addresses (0) to (1023) is assigned as indicated by reference numeral P1.
[1023] is stored. On the other hand, in this embodiment, the frequencies HK [0] to [387] forming the upper histogram P2 are stored at the addresses (253) to (640). That is, the maximum value M is assigned to the address (640).
The frequency HK [387] of X (= 1023) is stored, and the frequency HK [0] of the first boundary value BX is stored at the address (253).
Is stored. Also, addresses (0) to (252)
And the frequencies LK [0] to the lower histogram P3.
[252] is stored. Ie address (0)
The frequency LK [0] of the minimum value MI (= 0) is stored in, and the frequency LK [25 of the second boundary value BI is stored in the address (252).
2] is stored.

As described above, in the present embodiment, data relating to the histogram of pixel data from the maximum value MX to the first boundary value BX and data relating to the histogram of pixel data from the second boundary value BI to the minimum value MI. Are stored at consecutive addresses [0] to [640] of the memory.

FIG. 13 is a flowchart of a subroutine for creating the histogram H1. By executing this subroutine once, a histogram for one line is created. In this example, the maximum value MX is 102
3, the first boundary value BX is 636, and the second boundary value BI is 2
52 and the minimum value MI is 0.

In step 310, the initial value of the parameter X is set to 0. The parameter X is the line sensor 30.
Corresponding to the position of the photodiodes (that is, the pixels) arranged in the longitudinal direction of, the initial value 0 indicates the position of the pixel at the end on one line.

In step 320, the pixel value N of one pixel is read from the memory 46. In step 330, it is determined whether the pixel value N is larger than the second boundary value BI (= 252). When the pixel value N is equal to or smaller than the second boundary value BI, that is, when the pixel value N corresponds to the lower histogram P3, the frequency LK of the pixel value is determined in step 340.
[N] is counted. That is, in step 340, the lower histogram P3 is generated and the process proceeds to step 380.

On the other hand, when it is determined in step 330 that the pixel value N is larger than the second boundary value BI, step 350 is executed and the pixel value N is larger than the first boundary value BX (= 636). Is also small.
When the pixel value N is equal to or larger than the first boundary value BX, that is, when the pixel value N corresponds to the upper histogram P2, in step 360, the pixel value N is changed to the first boundary value BX (= 6).
36) is subtracted, and the frequency HK [N] of pixel values is counted in step 370. For example, the pixel value 1023
In the case of, the frequency HK [387] is counted. That is, in step 370, the upper histogram P2 is generated and the process proceeds to step 390.

When it is determined in step 350 that the pixel value N is smaller than the first boundary value BX, that is, when the pixel value N takes a value between 253 and 635, this pixel value N is displayed in the upper histogram P2. Since neither is included in the lower histogram P3, steps 340, 360 and 370 are skipped and the process proceeds to step 390.

In step 380, the parameter X is incremented by 1. In step 390, it is determined whether the parameter X is equal to or larger than the total number of pixels of the line sensor 30.
When the parameter X is smaller than the total number of pixels, step 3
Returning to step 20, the above-mentioned processing is repeated.

When it is determined in step 390 that the parameter X has reached the value of the total number of pixels or more, that is, when the histogram is created for the pixel values of one line of the color component of R, this subroutine ends, Returning to step 230 of FIG. 11, step 240 is executed next.

In steps 240 to 270, step 2
In the same manner as 20, 230, G and B exposure and histogram creation are performed. That is, in step 240, the light emitting element 21G of the light source 20 is turned on, and the G pixel data is detected by the exposure of the optimum G exposure time. In step 250, a G histogram is created for one line. In step 260, the light emitting element 21B of the light source 20 is turned on, and the pixel data of B is detected by the exposure of the optimum exposure time of B. In step 270, the histogram of B is created for one line.

In step 280, the stage 12, that is, the document M to be read is moved by one pitch, and step 290
Then, the parameter Y is incremented by 1. Step 300
In, it is determined whether or not the parameter Y is equal to or larger than the end value (that is, the number of scanning lines of the rough scan). When the parameter Y is smaller than the end value, step 220-
The processes up to step 300 are executed again, and the histogram of the next line is created. When it is determined in step 300 that the parameter Y is greater than or equal to the end value, 1
The histogram for the image has been completed, and this rough scan subroutine ends.

FIG. 14 is a flow chart of a subroutine for obtaining the maximum effective value D executed in step 130 of FIG. The maximum effective value D is the upper histogram P2
Required from.

In step 510, the pixel value E is set as the upper limit value of the upper histogram P2 as the initial value of the maximum effective value D. The upper limit value of the upper histogram P2 is 640 in the example of FIG. In step 520, the sum L of frequencies is set to an initial value 0. In step 530, the threshold TH
Is set to a frequency of 0.5% of the total number of pixels, for example. The maximum effective value D is determined according to this threshold value TH. That is, the maximum effective value D is the maximum pixel value when the sum L of frequencies is equal to or greater than the threshold value TH.

At step 540, the sum L of frequencies is obtained based on the upper histogram P2. That is, from the address (640) of the memory shown in FIG. 12, the frequency HK [3
87] is read and added to the total sum L of the frequencies so far. At step 550, the sum L of frequencies is the threshold TH.
It is determined whether or not the above. The sum L of frequencies is the threshold T
If it is not H or more, in step 560, the pixel value E
Is subtracted by 1. Next, at step 570, it is judged if the pixel value E is larger than the lower limit value of the upper histogram P2. The lower limit value is 253 in the example of FIG.
When the pixel value E is larger than the lower limit value of the upper histogram P2, that is, when the pixel value E is included in the upper histogram P2, step 540 is executed again, and the upper limit value 640 of the upper histogram P2 is changed to the pixel value E. The sum L of the frequencies up to is obtained.

In this way, when it is determined in step 550 that the sum L of frequencies is equal to or greater than the threshold value TH,
In step 580, the pixel value E is subtracted by 1,
Step 590 is executed. When it is determined in step 570 that the pixel value E is less than or equal to the lower limit value of the upper histogram P2, step 590 is executed. At step 590, the maximum effective value D is obtained according to D = E + 1 + (1023-upper histogram upper limit) (5). That is (5)
By the formula, the pixel value E is converted into the maximum effective value D in the complete histogram in which the pixel data in the range from the first boundary value to the second boundary value is not omitted.

FIG. 15 is a flow chart of a subroutine for obtaining the minimum effective value d executed in step 133 of FIG. The minimum effective value d is the lower histogram P3.
Required from.

In step 610, the pixel value e is set to 0 as the initial value of the minimum effective value d. In step 620, the total sum L of frequencies is set to an initial value 0. Step 63
At 0, the threshold TH is set to a frequency of 0.5% of the total number of pixels, for example. The minimum effective value d is determined according to this threshold value TH. That is, the minimum effective value d is the minimum pixel value when the sum L of frequencies is equal to or greater than the threshold value TH.

At step 640, the sum L of frequencies is obtained based on the lower histogram P3. That is, the frequency LK of the lower limit value from the address (0) of the memory shown in FIG.
[0] is read and added to the total sum L of the frequencies up to that point. At step 650, the sum L of frequencies is the threshold TH.
It is determined whether or not the above. The sum L of frequencies is the threshold T
If it is not H or more, in step 660, the pixel value e
Is incremented by 1. Next, at step 670, it is judged if the pixel value e is smaller than the upper limit value of the lower histogram P3. The upper limit value is 253 in the example of FIG.
When the pixel value e is smaller than the upper limit value of the lower histogram P3, that is, when the pixel value e is included in the lower histogram P3, step 640 is executed again, and the lower limit value 0 of the lower histogram P3 is changed to the pixel value e. The sum L of the frequencies up to is obtained.

In this way, when it is determined in step 650 that the sum L of frequencies is equal to or greater than the threshold value TH,
The pixel value e is incremented by 1 in step 680,
Step 690 is executed. When it is determined in step 670 that the pixel value e is smaller than the lower limit value of the lower histogram P3, step 690 is executed and 1 is subtracted from the pixel value e to obtain the minimum effective value d.

As described above, according to this embodiment, instead of storing all the data of the histogram in the memory,
Only data used for color correction and the like is stored in the memory. Therefore, it is possible to reduce the memory capacity for storing the histogram.

In the above-described embodiment, the histogram is used in the calculation of the color correction parameter.
The histogram can also be used in the exposure measurement performed in step 110 of FIG. In exposure measurement,
The exposure time of the line sensor 30 is very short as compared with the prescan and the main scan. Therefore, as shown in FIG. 18, the histogram is created only for the pixel values smaller than the third boundary value BE. That is, the routine for creating a histogram is as shown in FIG.
It is sufficient to omit 370 and change so as to determine in step 330 whether the pixel value is smaller than the third boundary value BE. The third boundary value BE has a value that is, for example, about ½ of the upper reference value Vmax. Next, the process of exposure measurement will be described.

16 and 17 are flowcharts of a subroutine for performing a rough scan in exposure measurement.

In step 700, the memory 46
The histogram data stored in is cleared. The histogram is created for each of the R, G, and B color components. In step 710, the stage 12 is moved to set the end of the document M to be read to the initial position corresponding to the light source 20, and the initial value of the parameter Y corresponding to the position of the document M to be read is set to 0. It

In step 720, the light emitting element 21R of the light source 20 is turned on to irradiate the read original M, and the line sensor 30 is exposed by the light transmitted through the read original M for the first exposure time for R. . As a result, one line of pixel data is detected. This pixel data is A
The data is converted into digital pixel data by the / D converter 44 and stored in the memory 46. In step 730, FIG.
Similar to step 230 of step 1, the subroutine for creating a histogram is executed.

In step 740, the light emitting element 21G of the light source 20 is turned on, and the line sensor 30 makes the first G
Is exposed for the exposure time of. At step 750, a histogram of G is created. In step 760, the light emitting element 21B of the light source 20 is turned on and the line sensor 30 is exposed for the first exposure time for B. Step 7
At 70, a histogram of B is created.

In this way, for each of the R, G, and B color components, a histogram is created for each line using the first exposure time.

Next, in steps 780 to 830, a histogram is created for each color component using the second exposure time as well as the first exposure time. The second exposure time of R, the second exposure time of G, and the second exposure time of B are determined based on the input / output characteristics of the line sensor 30 and the like and the input range of the A / D converter, respectively. Also, these second exposure times are set longer than the first exposure times.

The processing contents of steps 780 to 830 are basically the same as those of steps 720 to 770.

In step 840, the stage 12, that is, the document M to be read is moved by one pitch, and step 850
Then, the parameter Y is incremented by 1. Step 860
In, it is determined whether or not the parameter Y is equal to or larger than the end value (that is, the number of scanning lines of the rough scan). When the parameter Y is smaller than the end value, steps 720 to 720
The processes up to step 860 are executed again, and the histogram of the next line is created. When it is determined in step 860 that the parameter Y is greater than or equal to the end value, 1
The histogram for the image is completed, and the exposure measurement subroutine ends. In FIG. 18, reference numeral H6 shows a histogram according to the first exposure time, and reference numeral H7 shows a histogram according to the second exposure time.

Next, a program similar to the subroutine for calculating the maximum effective value D shown in FIG.
And the maximum effective values D1 and D2 of the histogram according to the second exposure time are obtained for each of the R, G, and B color components.

The optimum exposure time t is obtained by the equation (4).

[Equation 4]

In FIG. 18, the histogram H6 is the first
Corresponding to the pixel data obtained by the exposure of the exposure time t1 and the histogram H7 corresponds to the pixel data obtained by the exposure of the second exposure time t2. The upper reference value Vmax is an appropriate output value of the line sensor 30, and is determined by the input range of the A / D converter 44, in particular.

A for the amount of light incident on the line sensor 30
In the output characteristics of the A / D converter 44, the A / D converter 44
The output signal of is highly non-linear in the range where the amount of incident light is relatively small. Therefore, the optimum exposure time is the line sensor 30.
Is calculated in consideration of the input / output characteristics of the amplifier 43 and is determined so as to eliminate the influence of nonlinearity as much as possible.

That is, while the input / output characteristics of the line sensor 30 and the amplifier 43 are non-linear during the first exposure time t1, the exposure time is longer than the first exposure time t1, that is, the output of the line sensor 30. Pixel value range (V
In max-D1), the input / output characteristics of the line sensor 30 and the amplifier 43 are substantially linear, and the line sensor 30
Output is approximately proportional to the exposure time (t2-t1). Therefore, the proportional coefficient (Vmax- is related to the exposure time (t2-t1).
By multiplying D1) / (D2-D1), the output range of the line sensor 30 is changed from (D2-D1) to (Vmax
The optimum exposure time t can be obtained by expanding to -D1) and adding the first exposure time t1 to this.

As described above, according to the equation (4), since the proportional coefficient obtained from the histogram in the range where the input / output characteristics of the line sensor 30 and the analog processing circuit are substantially linear is used, the optimum exposure time is highly accurate. It is calculated.

As described above, also in the exposure measurement, the memory capacity for storing the histogram can be reduced by creating the histogram by using only the pixel data having the value smaller than the third boundary value BE.

Also, the histogram H in the exposure measurement
The range of the pixel data of 6 and H7 (abscissa) is not always the same as the range in the calculation of the color correction parameter. Therefore, usually, the first, second and third boundary values are different from each other in the exposure measurement and the color correction parameter calculation.

[0082]

As described above, according to the present invention, the memory capacity can be saved by reducing the amount of histogram data stored in the memory.

[Brief description of drawings]

FIG. 1 is a block diagram showing an image reading apparatus according to an embodiment of the present invention.

FIG. 2 is a perspective view showing a document transfer mechanism, a light source, and a line sensor when a transparent document is used as a document to be read.

FIG. 3 is a diagram showing an arrangement of a light source, a line sensor, and the like when a reflective original is used as a read original.

FIG. 4 is a flowchart showing an image reading routine.

FIG. 5 is a histogram showing a distribution of pixel values obtained by a line sensor.

6 is a diagram showing a histogram obtained by performing offset subtraction on the histogram of FIG.

FIG. 7 shows a coefficient (L2-L1) / in the histogram of FIG.
It is a figure which shows the histogram obtained by multiplying by (D-d).

FIG. 8 is a diagram showing a histogram of normalized data.

FIG. 9 is a diagram showing a histogram corresponding to gamma-corrected pixel data.

FIG. 10 is a diagram showing a configuration example of a lookup table.

FIG. 11 is a flowchart of a subroutine for performing a rough scan.

FIG. 12 is a diagram showing an example of a memory map of a histogram.

FIG. 13 is a flowchart of a subroutine for creating a histogram.

FIG. 14 is a flowchart of a subroutine for obtaining a maximum effective value.

FIG. 15 is a flowchart of a subroutine for obtaining a minimum effective value.

FIG. 16 is a flowchart showing a first half of a subroutine for performing exposure measurement.

FIG. 17 is a flowchart showing a latter half of a subroutine for performing exposure measurement.

FIG. 18 is a diagram showing a histogram obtained in exposure measurement.

[Explanation of symbols]

M Read original

─────────────────────────────────────────────────── ─── Continuation of the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H04N 1/40-1/409 H04N 1/46 H04N 1/60

Claims (6)

(57) [Claims] 1. A pixel data reading unit for reading pixel data forming one image, and a distribution of pixel values of each pixel data only for pixel data within a predetermined range from the pixel data read by the pixel data reading unit. Histogram creating means for creating a histogram, and the pixel data reading means provided to detect the image
The optical sensor, and the maximum pixel value in the distribution range of the histogram
The optical sensor based on a maximum effective value that is smaller by a predetermined amount
An exposure measuring means for determining the optimal exposure time for, than the minimum pixel value in the distribution range of the histogram
Based on the minimum effective value that is larger by a predetermined amount and the maximum effective value,
The pixel data output from the optical sensor.
Color for which the color correction parameters for performing color correction are calculated
A predetermined range of the histogram used in the exposure measurement means, which comprises a correction parameter calculation means
And a hist used in the color correction parameter calculation means.
An image reading device , wherein a predetermined range of grams is a range different from each other . 2. The color compensating means for generating the histogram
When calculating the color correction parameters by the positive parameter calculation means
To the maximum chromatic from the maximum value of the possible range of the pixel values
And pixel data of up to effective value is smaller than the first boundary value, the
Create a histogram showing the distribution of the value of each pixel data in the pixel data from the second boundary value that is larger than the minimum effective value and is smaller than the first boundary value to the minimum value of the range that the pixel value can take. The image reading apparatus according to claim 1, wherein: 3. The histogram creating means sets the exposure.
When measuring the exposure with the measuring means,
The image reading apparatus according to claim 1, wherein a histogram showing a distribution of values of each pixel data is created only for pixel data smaller than the third boundary value determined based on the output value. . 4. The image reading apparatus according to claim 1, further comprising a data storage unit that stores, in a memory, data relating to the histogram created by the histogram creation unit. 5. The histogram creating means sets the pixel
From the maximum value in the range of values that is less than the maximum effective value
First and pixel data to the boundary value have the minimum effective value
A second boundary value that is larger than the first boundary value and pixel data from the minimum value of the possible range of the pixel value to a minimum boundary value, and creates a histogram showing a distribution of pixel values of each pixel data, said data storage means, and data relating to the histogram of the pixel data of up to the minimum value from the first data and the second boundary value relating histogram of pixel data to a boundary value from the maximum value, the continuous of the memory The image reading apparatus according to claim 4, wherein the image reading apparatus stores the image at the specified address. 6. The maximum effective value is in the histogram.
Where the sum of the frequencies from the maximum pixel value is less than a predetermined threshold value.
It is the maximum pixel value when it is above, and the minimum effective value
Is the degree from the minimum pixel value in the histogram
It is the minimum pixel value when the sum of numbers exceeds a predetermined threshold.
The image reading apparatus according to claim 1, characterized in that.