JPS6244739B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for reading a document in a copying machine, facsimile machine, printer, etc. and converting the document into an electrical signal.

In general, conventional copying machines, facsimile machines, etc. use a one-dimensional image sensor as a photoelectric conversion element to read a document and convert it into an electrical signal. By using the dimensional image sensor, the image is read while simultaneously being moved in the sub-scanning direction by moving the original or moving an optical system such as a mirror.

However, when using a one-dimensional image sensor,
Since the light amount accumulation time can only be taken for a time equal to the scanning time of one line, there is a risk that the light amount will be insufficient during high-speed reading. Therefore, it is possible to increase the amount of light by using a halogen lamp as the light source, but
It is not preferable to use a halogen lamp as a light source because the compatibility with the spectral sensitivity characteristics of the image sensor or the resolution characteristics deteriorate particularly with respect to infrared light.

Also, since the sub-scanning is a continuous movement, there is a risk that the resolution will deteriorate. Therefore, it is possible to use a step motor, but considering its response characteristics, if a one-dimensional image sensor is used, the resolution will be degraded. Therefore, resolution deterioration cannot be prevented.

In order to solve this problem, the reading line in the main scanning direction (hereinafter simply referred to as "line")
It is conceivable to read the document using a two-dimensional image sensor having multiple rows of images on one chip.
In other words, if you use a two-dimensional image sensor,
There are the following advantages compared to using a one-dimensional image sensor.

(1) The light intensity accumulation time can be multiple times the scanning time of one line, thereby increasing the intensity of the light source,
Directly, it becomes possible to lower the image plane illuminance.

(2) In the case of a one-dimensional image sensor, in order to increase light utilization efficiency, it is necessary to illuminate only the area of the document to be read with one line of the image sensor, and if the light source is a fluorescent lamp, an aperture should be provided on a part of the fluorescent screen. This required measures such as adding a cylindrical lens. However, when a two-dimensional image sensor is used, a relatively wide area of the document is illuminated, so there is no burden on the light source.

(3) Since the image is stopped during photoelectric exchange, there is less deterioration in resolution.

In this way, by reading a document using a two-dimensional image sensor, it is possible to eliminate the problems that occur when a one-dimensional image sensor is used, especially during high-speed reading.

However, due to manufacturing problems, two-dimensional image sensors currently have a maximum number of lines of about 500 lines, and there is no element that can read the entire surface of a document at once. Therefore, when reading the entire surface of a document using an existing two-dimensional image sensor, the surface of the document must be divided into several parts in the sub-scanning direction and the reading must be performed multiple times.

Here, if the number of lines required to read the entire document is L and the number of effective lines read by the two-dimensional image sensor at one time is n, then the number of readings N is N=L/n.

In this way, when using a two-dimensional image sensor, reading is required N times, and if the image moves continuously at this time, resolution becomes impossible. For this purpose, it is necessary to move the image intermittently, but intermittent movement may impair the continuity of the reproduced image, resulting in duplication or omission of images, for example.

This invention has been made in view of the above-mentioned points.
A two-dimensional image sensor having multiple rows of reading lines in the main scanning direction sequentially reads a predetermined range that partially overlaps in the sub-scanning direction of the document at each conveyance step of the document and converts it into an image signal as an electrical signal.
The image signals of the overlapping lines of the image signals output at each conveyance step are cut to correct the joints in the sub-scanning direction of the image, enabling high-speed reading with good resolution and ensuring continuity of the reproduced image. The present invention provides a method for reading a document without causing any damage to the document.

Embodiments of the present invention will be described below with reference to the accompanying drawings.

FIGS. 1 and 2 are schematic perspective views showing the configurations of different parts of a document reading device embodying the present invention, and FIG. 3 is a schematic perspective view showing a two-dimensional image sensor. In FIG. 1, reference numeral 1 denotes a document, and 2 denotes a conveyor belt that intermittently conveys the document 1 in the direction of arrow A in order to perform width scanning. The conveyor belt 2 rotates intermittently in the direction of arrow B under the control of the step rotation of feed rollers 3 and 3' driven by a step motor (not shown), and the amount of movement per step is shown in FIG. The width is smaller than the width in the sub-scanning direction of a single imaging range P by the image sensor 6 shown in FIG.

Reference numeral 4 in FIG. 2 denotes a slit illumination light source consisting of a fluorescent lamp 4a and a reflecting mirror 4b, which slit-illuminates an area wider than a single imaging range P by the image sensor 6. Reference numeral 5 denotes a lens that reduces and forms a document image formed by reflected light from the document 1 on the image sensor 6 .

The image sensor 6 has a plurality of reading lines 6b
It is a two-dimensional type element having a photoelectric conversion surface 6a consisting of (Fig. 3), which simultaneously reads the screen of the original 1 within one imaging range P and converts it into an electric signal, and converts the image signal a to each line. Output each time sequentially. 7
is a signal processing circuit which inputs the image signal a output from the image sensor 6 and outputs the effective image signal b, and the configuration and operation of this circuit will be described in detail later.

Next, a method of reading a document in the document reading device configured as described above will be explained. A document 1 is intermittently moved by a conveyor belt 2 on a document table (not shown) (contact glass). Each time the document 1 stops, the image sensor 6 accumulates light on the document screen located within the imaging range P according to the amount of light reflected from each part, generates a charge corresponding to the amount of light, and transfers the charge from the controller described later. The signals are transferred to a shift register using pulses, and sequentially read out using readout pulses to output image signal a. This image signal a is converted into an effective image signal b which is binarized via the signal processing circuit 7.
is output as Then, the image sensor 6 performs the above-mentioned N times of reading, thereby reading the original 1.
can read the entire screen.

By the way, since the amount of movement of the conveyor belt 2 for each step is smaller than the width of the imaging range P for one time in the sub-scanning direction, the image of the document 1 that comes to the position of the imaging range P for each transport step is One part overlaps in sequence. Therefore, among the signals of each line constituting the image signal a output from the image sensor 6, the signals of one part of the lines will be outputted redundantly, and if the image signal a is used as a reproduced image signal, Image duplication occurs in the reproduced image. Therefore, the signal processing circuit 7 performs a correction to cut out the signals of the overlapping lines.

FIG. 4 is a block circuit diagram showing an example of this signal processing circuit 7. As shown in FIG. In the figure, 8 is the image sensor 6
This is a controller that controls the signal processing circuit 7 and each part of the signal processing circuit 7. 9 is image sensor 6
This is a binarization circuit that converts the image signal a into a binary signal. Reference numeral 10 denotes the last line of the line group that constitutes the effective playback screen when reading the document in the pre-conveyance step (this is referred to as the "effective final reading line").
11 is a matching circuit that sequentially compares the image signal in the memory 10 with the image signal of each line of the image signal a at the time of reading the document in the current transport step; 12 is a matching circuit for each line in the matching circuit 11; 13 is a register that stores the counted number of the match counter 12 as required; 14 is a comparator that compares the contents of the match counter 12 and the register 13. Reference numeral 15 denotes a gate circuit that allows or cuts off the output of the binarization circuit 9 based on a signal from the controller 8.

A method for cutting overlapping lines using the signal processing circuit 7 configured as described above will be explained.

First, a method for outputting the image signal a from the image sensor 6 will be explained. If the number of lines of the image sensor 6 is M as shown in FIG. 3, the maximum image range P at one time is M from the first line to the Mth line.
Read in book line. When the screen of the document 1 is photoelectrically converted from the first line to the Mth line, the controller 8 outputs an image sensor control signal d.
A transfer pulse is outputted to transfer the signal to the shift register corresponding to each line in the image sensor 6, and then a readout pulse is outputted and the signal is sequentially transmitted line by line from the 1st line to the Mth line. Read out. In this way, the image signal a obtained by reading the document screen within the imaging range P at one time is output. In this embodiment, the line in the area R shown by the broken line in FIG.
It is called a line.

Next, a method of cutting the signals of overlapping lines to match the seams of images will be explained. First, when the original 1 is conveyed and the screen within the imaging range P is read, the image signal a from the image sensor 6 (this is referred to as the "previous image signal") is input to the signal processing circuit 7. , is converted into a binarized image signal a' by a binarization circuit 9. Then, the controller 8 issues a write command c to the memory 10 at the timing when the image signal a' of the last effective read line (Nth line) is input, and thereby the memory 10 stores the contents of the image signal of the Nth line. do.

Furthermore, when the original 1 is conveyed and the screen within the imaging range P is read in the next step, and the image signal a (this is referred to as the "current image signal") is input, the controller 8 is activated by the matching circuit. When the current binarized image signal a' is input to 11, a read command is issued to the memory 10 in synchronization with the readout of each line, and thereby the memory 10 reads the stored content e, that is, the first image signal of the previous image signal. Match the contents of N line circuit 1
Output to 1.

The matching circuit 11 sequentially compares the contents of the current image signal for each line starting from the first line with the contents of the Nth line of the previous image signal for each bit, and determines that the contents of both lines match, that is. If both are at high level "H" or low level "L", an output f is outputted to the coincidence counter 12 for each bit. The coincidence counter 12 outputs f for each line.
The counted number (the number of matched bits in one line) g is sent to register 13 and comparator 1.
4 respectively.

The register 13 is cleared by the clear signal i from the controller 8 immediately before the first line of the current image signal is input. Therefore, first, the count number _g1 when comparing the first line is input to the register 13, and this becomes the storage content k of the register 13.

The comparator 14 compares the count g output from the coincidence counter 12 and the memory content k of the register 13 line by line, and when g>k, the determination result l
is output to the controller 8. Thereby, the controller 8 outputs a signal j and replaces the stored content k of the register 13 with the current count number g. In other words, if the count number g ₂ when comparing the second line is larger than the count number g ₁ when comparing the first line, then the memory content k of the register 13
is replaced by g ₂ from g ₁ . Similarly, if the count number _g3 of the third line is larger than the count number _g2 , the stored content k is replaced from _g2 to _g3 .

Then, when the stored content k of the register 13 becomes larger than the count g of the coincidence counter 12, the comparator 14 outputs the determination result, and the controller 8 then selects the line at that time as the first valid image signal of the current image signal. It is determined to be a reading line. That is,
The line next to the line with the highest degree of correlation (degree of coincidence) between the image signal of each line of the current image signal and the image signal of the Nth line of the previous image signal is determined to be the current first effective read line. The controller 8 then stores this first valid read line.

In this way, when the first effective reading line of the current image signal is determined and stored, the controller 8 either stops reading each line after the first effective reading line, or stops reading the last line (the Mth line). ), in order to read the same original screen again, a read pulse is output to the image sensor 6 to read out sequentially from the first line, and a gate control signal m is output at the timing of reading the first valid read line this time. to open the gate circuit 15. Therefore, the gate circuit 15 then outputs the image signals up to the Nth line, that is, the effective image signal b obtained by cutting out the image signals of the lines overlapping with the previous image signal.

In this way, the signal processing circuit 7 corrects the seams in the sub-scanning direction of the images by reading out the same document screen twice.

FIG. 5 is a block circuit diagram showing an example in which this signal processing circuit is simplified to correct the seam in the sub-scanning direction of the image by one readout. Parts corresponding to those in FIG. 4 are designated by the same reference numerals. The explanation of that part will be omitted.

In this signal processing circuit 7', a coincidence counter 16 counts the number of bits that are matched for each line in the coincidence circuit 11 in the same way as the coincidence counter 12 in the previous embodiment, and the count value is equal to the number of bits in one line of the image sensor 6. When the number of bits is approximately the same as the number of bits (determined in consideration of reading error), the determination result n is output, and the controller 8
outputs a gate control signal m which opens the gate circuit 15.

Therefore, in this signal processing circuit 7', the contents of the Nth line of the previous image signal and the contents of each line of the current image signal are sequentially compared, and when the contents substantially match, the coincidence counter 16 is activated. outputs the discrimination result n, and the controller 8 outputs the gate control signal m.
Output. As a result, the gate circuit 15 outputs image signals from the next line to the Nth line of the current image signal that substantially matches the contents of the Nth line of the previous image signal, that is, the image signals that overlap with the previous image signal. An effective image signal b obtained by cutting the line image signal is output.

Further, in this embodiment, the coincidence counter 16 discriminates the line of the current image signal that substantially matches the Nth line of the previous image signal, but similarly to the coincidence counter 12 of the signal processing circuit 7, Of course, the coincidence counter 16 may simply count the output f of the coincidence circuit 11 and output the count value, and the controller 8 may make the determination based on the count value.

In this way, the signal processing circuit 7' corrects the seam in the sub-scanning direction of the image by reading out the original screen only once.

In each of the above embodiments, image signals from bits outside the effective area R at both ends of each line can be easily removed by controlling the gate circuit 15 in synchronization with the readout timing by the controller 8. Removal is not necessary.

By repeating the above-described operations N times, the entire screen of the original 1 can be read.

As described above with respect to the embodiments, according to the present invention, a predetermined range that partially overlaps in the sub-scanning direction of the document is read at each conveyance step of the document by a two-dimensional image sensor, and the image signals are overlapped. The entire screen of the document is read by cutting out the image signal of the line that is being scanned and correcting the seams in the sub-scanning direction of the image, which enables high-speed reading with high resolution.Moreover, the intermittent feeding of the document allows for reproduction. Image continuity is not impaired.

[Brief explanation of the drawing]

1 and 2 are schematic perspective views showing the configuration of different parts of a document reading device embodying the present invention, FIG. 3 is a schematic perspective view showing a two-dimensional image sensor, and FIG. 4 is an embodiment of the present invention. FIG. 5 is a block circuit diagram showing an example of the signal processing circuit in this example. FIG. 5 is a block circuit diagram showing another example of the signal processing circuit. 1... Document, 2... Conveyor belt, 4... Light source for slit illumination, 5... Lens, 6... Two-dimensional image sensor, 7, 7'... Signal processing circuit, 8...
Controller, 10... Memory, 11... Match circuit, 12, 16... Match counter, 13... Register, 14... Comparator, 15... Gate circuit.

Claims (1)

[Claims] 1 While conveying the document intermittently in the sub-scanning direction,
A two-dimensional image sensor having a plurality of rows of reading lines in the main scanning direction sequentially reads a predetermined range that partially overlaps in the sub-scanning direction of the document at each conveyance step of the document and converts it into an image signal as an electrical signal; The image signal of the last valid reading line when reading the original in the previous transport step is stored, and the image signal when reading the original in the current transport step is sequentially compared with the stored image signal for each reading line, and the contents of both image signals are determined. When substantially coincident, the image signal of the next reading line is output as a valid image signal, and the entire screen of the document is read while correcting the seam in the sub-scanning direction of the image.