JPH04356870A - Image reader - Google Patents

Abstract

PURPOSE:To execute high accuracy image reading without being affected by reflectance in the other area of an original. CONSTITUTION:An area including a read picture element is irradiated with illumination light from a light source 2, and the reflected light is passed through a lens array 3 so as to obtain an output containing a shading influence at a CCD sensor 5a. On the other hand, the reflected light from a white reference plane 6b is outputted by a CCD sensor 5b. Since the output of the above- mentioned CCD sensor 5a is normalized for each picture element, an image output having no shading influence is obtained.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reading device, and more particularly to an image reading device capable of capturing image information with high precision.

0002

2. Description of the Related Art Photoelectric conversion elements such as CCD sensors are used to read input original images in copying machines, facsimile machines, image file devices, and the like. These photoelectric conversion elements have a single line configuration and are equipped with enough cells to read 1 mm with a resolution of several lines to several tens of lines. The image is read by electrical scanning in the line direction (main scanning) and by moving the entire photoelectric conversion element in the direction perpendicular to this (sub scanning). Examples of such image reading devices include:
For example, it is described in Japanese Unexamined Patent Publication No. 62-235872.

[0003] In the above-mentioned image reading device, when reading a document, the actual brightness (reflectance) of a certain point is dragged by the surrounding brightness and is read as a factor that changes the reading level. There is a phenomenon called. This is because the illumination for reading the document is applied to an area having a certain area including the point to be read, and the diffused reflected light from the surrounding document surface becomes a secondary light source. The light from this secondary light source gets mixed into the optical path from the normal reading point to the sensor, and is sensed by the sensor as a level change. Therefore, a small black area within an entirely white area is read brighter than it actually is. For example, black text on a white background is affected by the white background, making the black lighter and the edges duller, making it difficult to read as a clear black text. Also, an area of a certain size goes from white to black (from black to white)
Even during transition, the edges are not sharp and are blurred, resulting in a read image that lacks fine image information.

0004

[Purpose] The present invention has been made in view of the above-mentioned circumstances, and the influence of reflected light from other areas of the document that gets mixed into the optical path from the reading pixel to the photoelectric conversion element, which causes level fluctuations in the read image. The purpose of this invention is to provide an image reading device that can obtain the output of a photoelectric conversion element by blocking the reflected light of only the pixel, and can input an image without damaging the image information of the document. It is something.

[0005]

[Structure] In order to achieve the above objects, the present invention provides (1)
In an image reading device that uses a photoelectric conversion element to obtain an electric signal according to the reflectance of a document as an image signal, a pixel of the document to be read is almost located close to blocking the reflected light of the pixel, and the reflectance on the side of the photoelectric conversion element is The white color is determined based on the photoelectric conversion output of the pixel obtained through the shielding means having a white reference surface set to a certain constant value and the white reference surface (6(b)) provided on the shielding means (6). A normalization means (18) is provided for normalizing the photoelectric conversion output of the pixel obtained without using a reference plane, or (2) a photoelectric conversion element converts an electric signal according to the reflectance of the document into an image signal. In the image reading device obtained as a first photoelectric conversion element (5(
storage means (34) for storing information obtained by dividing the line image information obtained by reading the white reference surface by b)) by the line image information obtained by reading the white reference surface by the second photoelectric conversion element (5(a)); A white reference surface (6(b)) is provided almost close to a pixel of the document to be read to block the reflected light of the pixel and has a constant reflectance on the photoelectric conversion element side;
normalizing the output of the second photoelectric conversion element of the pixel obtained without passing through the white reference plane, with reference to the output of the first photoelectric conversion element of the pixel obtained via the white reference plane; The present invention is characterized by comprising a normalization means (18) and a multiplication means (19) for multiplying the output of the normalization means by the output of the storage means.

That is, the above purpose is to read the standard brightness level at each point of the document to be read by shielding only the normally read pixels with a white film that serves as the standard, thereby increasing the sensor output. This is achieved by normalizing the normal read output level of the pixel. Furthermore, when considering a state where information of a certain pixel is being read, the reference level of brightness at that reading position changes due to the influence of the secondary light source due to the brightness of the surrounding document. Therefore, by shielding only the pixels to be read with a reference white film and reading them under the influence of the secondary light source, the reference brightness at that location can be obtained as a sensor output. By using this output level to normalize the normal reading output level of that pixel, it is possible to obtain pixel information in which the influence of the secondary light source is canceled out. Here, the above-mentioned "influence of the secondary light source" also applies to the "influence of shading", and therefore, according to this means, the influence of the secondary light source and the influence of shading can be corrected at the same time. Hereinafter, the present invention will be explained based on examples.

FIGS. 1(a) and 1(b) are configuration diagrams for explaining an embodiment of an image reading device according to the present invention.
1 is a document (the bottom surface is the document surface), 2 is a light source, 3 is a lens array, 4 is a carriage, 5a and 5b are CCDs (Charg
e Coupled Device (charge coupled device) sensor, and 6 is a shielding means (membrane). CCD sensor 5a,
5b has two photosensitive parts (pixel rows). Further, the shielding means 6 is a transparent film for the CCD sensor 5a, and is a film having a standard white surface for the CCD sensor 5b.

As shown in Figure (b), the film 6 has a linear white reference surface 6b in the main scanning direction, and the other portions are transparent as shown in 6a. In Figure (a), the horizontal direction represents sub-scanning, and the direction perpendicular to the paper surface represents main scanning, and carriage 4 moves together with light source 2, lens array 3, CCD sensor 5, and film 6. The illumination light from the light source 2 illuminates a region including the read pixels, and the reflected light from the pixels passes through the lens array 3 and forms an image on the CCD sensor 5a. Thereby, the sensor output of the read pixel including the influence of the secondary light source and the influence of shading can be obtained as the output of the CCD sensor 5a. In addition, the illumination light from the light source 2 is diffusely reflected by the white reference surface 6b, and the reflected light is passed through the lens array 3 to the CCD.
By forming an image on the sensor 5b, a reference white level at this reading position including the influence of the secondary light source and the influence of shading can be obtained as the output of the CCD sensor 5b.

[0009] Here, in the main scanning direction, only each pixel is not shielded by the reference white film, but in the sub-scanning direction, the width is small (approximately 63 μm at 400 dpi), and the surrounding pixels are It is possible to obtain an output in which the effects of secondary light sources from almost all originals are superimposed. Therefore, by normalizing the output of the CCD sensor 5a for each pixel using this output, an image output can be obtained while eliminating the influence of the secondary light source and the influence of shading.

This will be explained using a reading characteristic diagram as shown in FIG. 5, in which the horizontal axis represents the reflectance of pixels of the document to be read, and the vertical axis represents the output value of the sensor. The reading characteristics are as follows: Influence of secondary light source: ∫r(x,y)・g(x-x0,y-y0)dxdy where r(x,y): original image g(x,y): secondary light source Weighting functions x0, y0 representing the degree of influence as: The position of the reading pixel and, in addition, the amount of light at that point: I(x0, y0) varies as A, B, C. This means that the reading characteristics change at each point on the document. Here, the output of reading the white surface (6b in Figure 1) with the reference reflectance r0 is y01, y02,
If it is y03, it can be estimated that the reading characteristics at each pixel position are A, B, and C. Therefore, when the sensor output of the document pixel is ym, the reflectance of the document pixel is rm1, rm2, r
It can be concluded that m3. This is nothing but normalizing the sensor output ym with the sensor outputs y01 to y03 obtained by reading the reference white color at that position.

FIG. 2 is a diagram showing an embodiment of a signal processing circuit used in an image reading apparatus according to the present invention. In the figure, 7 is a CCD sensor, 8 and 20 are sample and hold circuits (hereinafter referred to as S&H), 9, 21 are S&H, 10, 24 are S
&H and amplifier circuit, 11 and 25 are switch circuits, 12
, 26 are A/D conversion circuits, 13 and 27 are switch circuits,
14 and 28 are averaging circuits, 15 and 29 are memories for storing one line of dark output, 16 and 30 are multiplication circuits, 17 and 31 are subtraction circuits, 18 is a division circuit (normalization means), 19 is a multiplication circuit, 32 is a FIFO memory, 33 is a division circuit, and 34 is a memory (storage means) for storing the sensor sensitivity ratio. The feature of this signal processing is that in addition to correcting the effects of the secondary light source and shading mentioned above, it also sequentially corrects the dark output that changes moment by moment during document reading for each pixel, and the resolution of the two pixel rows of the sensor. The objective is to eliminate correction errors caused by the ratio.

First, a dark output over one line is generated and stored. The sensor 7 is a CCD sensor 5a shown in FIG.
, 5b, and their outputs are led to S&Hs 8 and 20, respectively, and the CCD transfer clock component is removed. This S
The output waveforms of &H8 and 20 are as shown in FIGS. 6(a) and 6(b), respectively, assuming that the amount of exposure to the sensor 7 is zero. In other words, the output DS(E
), the output DS(S) of the dummy pixel whose photosensitive part is shielded from light and the output of the image area part are all dark output Dk1.
, Dk2, and the output level varies for each pixel. These outputs are led to S&H9 and S&H21 respectively, sampled and held at the value of DS(E), and upper limit reference values VrH1 and Vr of A/D conversion circuits 12 and 26 are
Let it be H2.

It is also led to the S&H and amplifier circuits 10 and 24 to sample and hold and amplify the average value of DS(S), and then set the switch circuits 11 and 25 to the opposite switching position from that shown in the figure, and then to the A/D conversion circuit. The lower limit reference values VrL1 and VrL2 are 12 and 26, respectively. In other words, A
/D conversion circuits 12, 26 references VrH1, Vr
L1, VrH2, VrL2 and signal input (=S&H8, 2
The relationship with the output (0 output) is as shown in FIG. This performs A/D conversion, and the output is sent to the switch circuit 13.
, 27 are the switching positions opposite to those shown, and one line is stored in the dark output memories 15 and 29. This completes the capture of one line of dark output patterns based on the average value of the pixels of the light-shielded photosensitive area. This dark output pattern is a value indicating the dark output level of each pixel based on the average output of pixels having a light-shielded photosensitive area, and is data that depends only on the sensor 7. .

Next, the sensitivity ratios (over one line) of pixels corresponding to the same position in the main scanning of the two pixel columns 5a and 5b are captured. Therefore, the switch circuits 11 and 25
, 13 and 27 are the illustrated switching positions. The carriage 4 in FIG. 1 has a pixel row 5a all white on the reference surface 6b.
A reference plane (not shown) having the same reflectance as , is set at a reading position, and the output of the sensor 7 is obtained. The outputs of the two pixel columns pass through S&H 8 and 20, and as shown in FIGS.
The output is a superimposition of 2.Dk2 and the optical outputs Ds1 and Ds2. Here, α1 and α2 are coefficients indicating fluctuations when the dark output is taken in.

[0015] For the output of S&H8, 20, S&H9,
21 samples and holds the value of DS(E), and the output is used as the upper limit reference Vr of the A/D converter circuits 12 and 26.
H1 and VrH2. In addition, the output of the S&H 20 is guided to the peak hold circuit 22, held at the peak value, and amplified by the amplifier circuit 23, and this output is sent to the switch circuits 11, 25.
are used as the lower limit reference values VrL1 and VrL2 of the A/D conversion circuits 12 and 26 through. Therefore, the references VrH1 and VrH2 to the A/D conversion circuits 12 and 26
, VrL1, VrL2 and the input signal as shown in FIG.

Here, Ds at the same pixel position in the main scanning direction
1 and Ds2 are outputs from almost the same reading position,
Since these are the read outputs of objects with the same reflectance, the ratio Ds2/Ds1 is the sensitivity ratio of the sensors 5a and 5b. Therefore, in order to obtain Ds1 and Ds2, the output after A/D conversion is led to the averaging circuits 14 and 28 and the subtraction circuits 17 and 31 via the switch circuits 13 and 27, and the dark outputs α1・Dk1, α2・Dk2 Remove. The averaging circuits 14 and 28 have a function of adding data in the DS(S) interval and obtaining an average value.

For this output, dark output memories 15, 29
Multiplying circuits 16 and 3 output the output read over one line.
By multiplying by 0, a dark output value which is multiplied by α1 and α2 can be obtained for each pixel. This dark output value is subtracted from the original A/D converted value by subtraction circuits 17 and 31, and Ds1 and Ds2 can be obtained as the outputs, respectively. These output values are led to the division circuit 33 and Ds
2/Ds1 is determined and stored in the sensitivity ratio memory 34 for one line. This completes the acquisition of the sensitivity ratios of the pixel rows 5a and 5b of the sensor 7.

Next, an image is input. This is performed by performing signal processing for each line while moving the carriage 4 in the sub-scanning direction in FIG. 1, and the steps up to the subtraction circuits 17 and 31 are the same as the above-mentioned sensor sensitivity ratio acquisition. That is, A
/D conversion circuits 12 and 26 two references VrH1
, VrL1 and VrH2, and the relationship between VrL2 and the input signal is as shown in FIGS. are superimposed), but the output of the subtraction circuit 31 provides a light output value Dm2 corresponding to the reference white color at the position where the pixel row 5b is reading. Here, α1' and α2' in FIGS. 8(a) and 8(b) are coefficients indicating fluctuations when the dark output is taken in.

The output of the subtraction circuit 31 is stored in the FIFO memory 32.
In order to correct the difference in the position of the pixel rows 5a and 5b in the sub-scanning direction, the division circuit 18 calculates Dm1/Dm2 with the output Dm1 of the subtraction circuit 18 by delaying the corresponding line. This output is led to the multiplier circuit 19 and the sensitivity ratio memory 3
4 is sequentially multiplied by the output read over one line. As a result, (Dm1/Dm2)・(Ds2/Ds1)
can be obtained as an image reading output. In other words, this output is affected by the influence of the secondary light source, the influence of shading,
This is an output that is faithful to the original information in which dark output and the influence of the sensitivity ratio of the pixel rows 5a and 5b have all been removed.

FIG. 3 is a diagram showing another embodiment of the signal processing circuit used in the image reading device according to the present invention.
5 and 36 are read-only dark output memories, and other components that are the same as those in FIG. 2 are given the same numbers. Figure 2
The difference from the embodiment is that read-only dark output memories 35 and 36 are used instead of the dark output memories 15 and 30. Therefore, there is no operation to capture the dark output, and elements necessary for this are omitted. Storage in the dark output memories 35 and 36 is performed, for example, at the time of manufacturing or maintenance using the same concept as in the embodiment shown in FIG. This embodiment can be applied when the dark output pattern of the sensor changes little over time.

FIG. 4 shows still another embodiment of the signal processing circuit used in the image reading apparatus according to the present invention. In the figure, reference numeral 37 is a read-only sensitivity ratio memory, and other components are the same as in FIG. 3. are given the same number. The difference from the embodiment shown in FIG. 3 is that a read-only sensitivity ratio memory 37 is used instead of the sensitivity ratio memory 34. Therefore, there is no operation to capture the sensitivity ratio, and elements necessary for this are omitted. Storage in the sensitivity ratio memory 37 is as follows:
For example, during manufacturing and maintenance, the same concept as in the embodiment shown in FIG. 2 is used. This embodiment can be applied when the dark output pattern of the sensor and the sensitivity of photoelectric conversion change over time are small.

[0022]

[Effects] As is clear from the above description, the present invention has the following effects. (1) In the image reading device according to claim 1, the reference brightness level at each point of the document to be read is read by shielding only pixels that are normally read with a white film that serves as the reference. Since the normal reading output level of the pixels can be normalized, highly accurate image reading can be performed without being affected by the reflectance of other areas of the document. (2) In the image reading device according to claim 2, the reference brightness level at each point of the document to be read is read by shielding only the normally read pixels with a white film serving as the reference. The normal reading output level of the pixel in the second sensor is normalized, and the normalization error due to the photoelectric conversion sensitivity of the first and second sensors is calculated in advance by the sensitivity ratio. Since it can be corrected by memorizing this, it is possible to read an image with high accuracy without being affected by the reflectance of other areas of the document.

[Brief explanation of the drawing]

FIG. 1 is a configuration diagram for explaining an embodiment of an image reading device according to the present invention.

FIG. 2 is a diagram showing an embodiment of a signal processing circuit used in an image reading device according to the present invention.

FIG. 3 is a diagram showing another embodiment of the signal processing circuit used in the image reading device according to the present invention.

FIG. 4 is a diagram showing still another embodiment of the signal processing circuit used in the image reading device according to the present invention.

FIG. 5 is a diagram showing the relationship between the reflectance of a pixel of a document to be read and the output value of a sensor.

FIG. 6 is a diagram showing the output waveform of the sample hold circuit (during dark output reading).

FIG. 7 is a diagram showing the output waveform of the sample and hold circuit (when reading the sensitivity ratio).

FIG. 8 is a diagram showing the output waveform of the sample hold circuit (at the time of image reading).

[Explanation of symbols]

1... Original (bottom side is the original surface), 2... Light source, 3... Lens array, 4... Carriage, 5a, 5b... CCD (Charg
e Coupled Device; charge coupled device) sensor, 6...shielding means (membrane)

Claims (2)

[Claims] 1. An image reading device that obtains an electric signal according to the reflectance of a document as an image signal using a photoelectric conversion element, wherein the photoelectric conversion element is located substantially close to a pixel of the document to be read and blocks reflected light from the pixel; A shielding means having a white reference surface whose side reflectance is set to a certain constant value, and a photoelectric conversion output of the pixel obtained through the white reference surface provided on the shielding means as a reference, 1. An image reading device comprising: normalizing means for normalizing a photoelectric conversion output of the pixel obtained without any processing. 2. In an image reading device that uses a photoelectric conversion element to obtain an electric signal corresponding to the reflectance of a document as an image signal, line image information obtained by reading a white reference surface by the first photoelectric conversion element is converted to a second photoelectric conversion element. A storage means for storing information obtained by dividing the white reference surface by the line image information read by the element, and a storage means that is almost close to a pixel of the document to be read and blocks reflected light from the pixel, and maintains a certain reflectance on the photoelectric conversion element side. A white reference plane is provided, and with the output of the first photoelectric conversion element of the pixel obtained through the white reference plane as a reference, the output of the second photoelectric conversion element of the pixel obtained without going through the white reference plane is An image reading device comprising: normalization means for normalizing the output of the photoelectric conversion element; and multiplication means for multiplying the output of the normalization means by the output of the storage means.