JPH11177775A - Detector for original - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow a detector to detect the position and inclination of an original with high accuracy and few midsections, even in the case of presence of read pitch unevenness by deciding the inclination of a line segment with the inclination of a specific angle or below in a subscanning direction among line segments extracted by a line segment extract means as the inclination of the original. SOLUTION: An original inclination detection means decides the inclination of a line segment, less than 45 degrees with respect to a subscanning direction from among line segment extracted by a line segment extract means. Furthermore, an original area decision means decides an original area with the inclination decided by the original tilt detection means and inclin ding the line segments extracted by the line segment extract means. A CPU 314 of the detector detects the change in the indication between coordinate data of each original edge, based on coordinate data stored in a stack memory 308 obtain points (edge change points) at which the inclination is changed and the extract line segments connecting the edge change points. Then line segments less than 45 degrees with respect to a direction X area selected from among the line segments extracted by the line segment extract means.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital image forming apparatus, and more particularly, to a digital image forming apparatus that detects a document placement position and an image area based on image data.

[0002]

2. Description of the Related Art A digital image forming apparatus such as a digital copying machine forms an image on a sheet after image processing of digital image data. Digital image data for each pixel is generated by reading an image of a document placed on a document table by a reading unit such as a CCD. The placement position and image area of the document can be detected based on the digital image data. For example, a document reading apparatus described in Japanese Patent Application Laid-Open No. 62-166652 detects the coordinates of four points that take the maximum and minimum values in each axis direction of a document, and determines the reference of the adjacent side of the side connecting the four points. The angle with respect to. Also, JP-A-7-29
The image processing apparatus described in Japanese Patent Application Laid-Open No.
The coordinates of one vertex are detected, the inclination is calculated from the coordinates of the first and second detected vertices, and the movement amount of the document is calculated from the coordinates of the first detected vertex.

[0003]

In a general digital image forming apparatus, image data of a document is obtained by moving a sensor such as a CCD element in a sub-scanning direction. However, JP
According to the apparatus described in Japanese Patent Application Laid-Open No. 2-166651 or Japanese Patent Application Laid-Open No. 7-298031, the inclination of the document is detected by detecting each vertex of the document. In this case, unevenness of the reading pitch in the sub-scanning direction is detected. Due to the influence, there is a possibility that the position of each vertex of the document is erroneously detected, and it is not possible to accurately detect the inclination of the document or the document area. Also, when detecting the inclination of a document from the detection results of the edges of a plurality of document images, if the inclination of a line segment of 45 ° or more with respect to the sub-scanning direction is detected, unevenness in the read pitch may be detected at each edge. As a result, the inclination and the original area could not be detected accurately.

It is an object of the present invention to provide a document detecting apparatus capable of detecting the position and inclination of a document with higher accuracy and less erroneous detection even if the reading pitch is uneven.

[0005]

According to the present invention, there is provided an original detecting apparatus for reading an original placed on an original platen, detecting an edge of the original on the original platen, and an edge detecting unit. A line segment extracting unit for extracting a line segment composed of the detected continuous edges; and a line segment having an inclination of 45 ° or less with respect to the sub-scanning direction among the line segments extracted by the line segment extracting unit. Document tilt detecting means for determining the tilt as the tilt of the document; document area determining means for determining a document area including the line segment extracted by the line segment extracting means, comprising a tilt determined by the document tilt detecting means; Is provided. Even when there is unevenness in the original reading pitch, the inclination of the line segment having an inclination of 45 ° or less with respect to the sub-scanning direction is determined as the original inclination, so that an appropriate original area and inclination can be detected. Preferably, the document inclination detecting means sets the inclination of the longest line segment among the line segments of 45 ° or less with respect to the sub-scanning direction as the document inclination. Thus, the longest side of a non-rectangular document can be detected, and an appropriate document area and inclination can be detected based on the longest side. Further, for example, the document inclination detecting means sets the inclination of the longest and orthogonal line segment among the line segments of 45 ° or less with respect to the sub-scanning direction as the document inclination. .

[0006]

Embodiments of the present invention will be described below with reference to the accompanying drawings. In these drawings, the same reference numerals indicate the same or similar elements. First, FIG. 1 shows an overall configuration of a digital copying machine according to an embodiment of the present invention. The copier reads a document and converts it into an image signal, a scanning system 10, an image processing unit 20 that processes an image signal sent from the scanning system 10, and directly outputs image data input from the image processing unit 20 to a printer device. A rotation memory unit 30 for controlling whether or not image data is processed and then output to a printer device; a print processing unit 40 for driving a semiconductor laser 61 based on image data output from the rotation memory unit 30; A laser optical system for guiding laser light from the semiconductor laser 61 to an exposure position on the photosensitive drum 71; an image forming system for developing a latent image by exposure, transferring the image onto recording paper, and fixing to form an image; A paper transport system for feeding and discharging, an operation panel 90 (not shown) provided on the upper surface of the copying machine main body, a document transport for transporting a document And a section 500. The image reading unit 100 includes the scanning system 10 and the image processing unit 20, and the printer unit 200 includes the print processing unit 40, the laser optical system 60, the image forming system, and the like.

The document transport section 500 is mounted on the platen glass 19 of the image reading section 100 so as to be openable and closable.
The document conveying section 500 automatically conveys the document set on the paper feed tray 510 onto the platen glass 19, and transfers the document read by the scanning system 10 to the discharge tray 5.
Discharge to 11 In the normal mode, one or a plurality of documents are set on the paper feed tray 510 with the surface to be read facing upward, and the side regulating plate 513 is adjusted to the width of the document. Then, the presence or absence of a document is detected by an empty sensor (not shown). When the conveyance operation is started, the sheet feeding rollers 501 are sequentially arranged from the lowermost document on the tray 511.
The paper is conveyed by the separating roller 502 and separated by the separating pad 503 and fed one by one. The conveyed document passes through the intermediate roller 504, and after the registration sensor 551 and the width size sensor 553 detect the document, the skew of the document is corrected by the registration roller 505. Thereafter, the document is conveyed on the platen glass 19 by the registration roller 505 and the conveyance belt 506, and immediately after the front end of the document abuts on the document scale 512,
The transport belt 506 and the registration roller 505 stop. As a result, the left end of the document comes into contact with the edge of the document scale 512, and the document is set at an accurate position on the platen glass 19. At this time, the leading edge of the next document has reached the registration roller 505, so that the transport time of the next document is reduced.

When the original is set at an accurate reading position on the platen glass 19, the scanning of the original by the scanning system 10 is performed. When the reading of the document is completed, first, the document scale 512 is pushed down below the height of the upper surface of the platen glass 19 by a solenoid (not shown). Thereafter, the document is conveyed to the left by the conveyance belt 506, the conveyance direction is changed by the reversing roller 507, and the document is discharged above the switching claw 508 onto the discharge tray 511.

In the case where the step feed mode is selected, when the document size is less than half the size from the exposure reference position to the nip position of the registration roller 505, the preceding document is stopped at the exposure reference position. At the same time, the next original is transported to the intermediate position between the exposure reference position and the registration roller 505, and the next original (third sheet) is fed in advance until the leading end of the original comes into contact with the registration roller 505. By doing so, since the original is stepped by half the distance between the exposure reference position and the registration roller, the original replacement time is short, and the original is returned within the time when the scanning system 10 returns to the home position after the exposure is completed. Can be replaced, and copy productivity is improved. In addition, the succeeding original (third sheet) is advanced and fed until its leading end contacts the registration roller 505. This advance feeding is performed during the exposure of the original, which contributes to the improvement of copy productivity.

In the case of a double-sided original, when reading of the first side is completed, the original is conveyed to the left by the conveying belt 506, the conveying direction is changed by the reversing roller 507, and then the platen glass 19 is again moved by the switching claw 508. And the original is also set to the reading position on the second side. The document whose second surface has been read is conveyed leftward by the conveyance belt 506, and is discharged onto the discharge tray 511 via the reversing roller 507, the switching claw 508, and the discharge roller 509. Further, when a mode such as the step feed mode or the duplex mode is selected, the conveyance of the conveyance belt 506 stops immediately before the leading edge of the document abuts on the right end of the document scale 512, and the document scale 512 The document is set at a position slightly away from the document.

The surface of the conveyor belt 506 on the platen glass 19 side is colored orange. As a result, the reflected light of the light of the exposure lamp 12 by the document conveying belt 506 becomes a color having a small spectral sensitivity for the line sensor 17. That is, for the line sensor 17, the transport belt 5
06 is the same as black. Therefore, the background of the document is usually white, so that even when the transport belt 506 is closed, the line sensor 17 detects the original and the transport belt 50.
6 can be identified. Further, even when the document conveying unit 500 is not closed, the reflected light of the conveying belt 506 by the exposure lamp 12 does not reach the line sensor 17, so that the document area can be identified.

The image reading section 100 includes a platen glass 19
The image of the document placed on the document is read, and image data corresponding to each pixel of the image of the document is generated. In the document reading unit 100, the exposure lamp 12 and the first mirror 1
First scanner 11 having 3a and second and third mirrors 13
The second scanner 14 having b and 13c is driven by the scan motor M2 in the directions of arrows b and b '(sub-scanning direction).
Moved to The light of the exposure lamp 12 is the platen glass 1
9, mirrors 13a, 13b,
13 c, the light is irradiated to the line sensor 17 via the lens 15. The line sensor 17 has a large number of photoelectric conversion elements (CCDs) arranged in a direction (main scanning direction) orthogonal to the paper surface of FIG. 1, reads an image at 400 DPI, and outputs image data corresponding to each pixel. Further, as described above, when the first scanner 14 moves in the directions of the arrows b and b ′, the line sensor 17 can scan the document image in the sub-scanning direction. The scanning of the image by the line sensor 17 when the scanners 11 and 14 move in the direction of arrow b is the preliminary scanning. At this time, the position of the document on the document table is detected based on the image data output from the line sensor 17. Is done. On the other hand, the scanning of the image by the line sensor 17 when the scanners 11 and 14 move in the direction of the arrow b 'is the main scanning. At this time, the original image is copied based on the image data output from the line sensor 17. Will be The image data output from the line sensor 17 is processed by the image processing unit 20 and then transmitted to the rotating memory unit 30. The rotation memory unit 30 stores the image data received from the image processing unit 20 and transmits the image data to the printing unit 200 after the rotation editing process or directly.

Next, the printing section 200 will be described. In the printing unit 200, the print processing unit 40 controls the laser optical system based on the image data received from the rotation processing memory unit 30. The laser optical system emits a laser beam modulated (on / off) by the print processing unit 40, and a laser beam emitted from the semiconductor laser 61 scans the photosensitive drum 71 on the photosensitive drum 71. Polycon mirror 62, fθ
A lens 63 and mirrors 64a, 64b, 64c are provided. Around the photosensitive drum 71 that is driven to rotate in the direction of arrow c, the charger 72, the developing device 73, the transfer charger 74, the separation charger 75, the cleaner 76, and the eraser lamp along the direction of rotation (direction of arrow c). A toner image 77 is formed by a well-known electrophotographic process, and is transferred onto a sheet. The paper is supplied from paper cassettes 81a and 81b by paper feed rollers 82a and 82b, and is supplied to a paper conveyance path 83 and a timing roller 8a.
4 to the transfer charger 74. The sheet on which the toner image has been transferred at the position of the transfer charger 74 is discharged onto a discharge tray 88 via a conveyor belt 85, a fixing device 86, and a discharge roller 87.

FIG. 2 is an overall block diagram of a control system for controlling the digital copying machine. A control unit 102 of the image reading unit 100, a control unit 300 of the rotary memory unit 30,
The control unit 202 includes a control unit 202 of the printer 200 and a control unit 520 of the document conveyance unit 500, and is connected to the overall control unit 400 via a communication line. The overall control unit 400 includes each control unit 1
02, 300, 202, and 520, and also controls the operation panel 90. FIG. 3 shows the operation panel 90. The operation panel 90 is provided with a tilt correction mode key 99 for setting a tilt correction mode and a display unit 99a for displaying that the mode is the tilt correction mode. Further, the operation panel 90 is provided with a liquid crystal touch panel 91 and a numeric keypad 9 for inputting the number of copies, such as the order of page numbers and the number of copies, and the copy magnification as in a normal copying machine.
2, a clear key 93 for returning the set value to the standard value "1", a panel reset key 94 for returning the set value set in the copying machine to the standard value, and a stop key 9 for stopping the copying operation.
5. A start key 96 for starting a copy operation is provided. Further, a copy mode setting key 97 for selecting and setting any one of a copy single-sided mode, a single-sided 2-in-1 mode, and a single-sided 4-in-1 mode as a copy mode, and selecting and setting one of a document single-sided mode and a document double-sided mode as a document mode. Original mode setting key 98, display portion 97a for displaying that the selected and set copy mode is the copy one-sided mode, display portion 97b for displaying the one-sided 2in1 mode, and display portion for displaying the one-sided 4in1 mode 97c, a display section 98a for displaying that the selected and set original mode is the original simplex mode,
A display unit 98b for displaying that the original is in the both-sided document mode is provided. The liquid crystal touch panel 91 displays the operation state of the copier such as an exposure level, a copy magnification, a recording paper size, various abnormal states of the copier such as occurrence of a jam, and other information. An input for designating an automatic selection mode for recording paper or the like can be made.

Next, rotation of an image in the rotation memory unit 30 will be described. FIG. 4 shows the rotating memory unit 30.
5 shows a timing sequence from the image reading control unit 102 to the rotating memory unit control unit 300, and FIG. 6 shows a timing sequence from the rotating memory unit controlling unit 300 to the printer. 4 shows a timing sequence to the control unit 202.
As shown in FIG. 4, the image data input / output interface sends signals VD_IR, HD_
To receive IR, SYNCK_IR, the _VIDEO 0~7 _IR. FIG. 5 shows a sequence of image data transferred from the image reading control unit 102. Here, the VD_IR signal indicates page data output, and becomes active during a low level. The HD_IR signal indicates line data output and is active during the low level. VD_IR, H
When both D_IR are active, SYNCK_IR
Valid image data _VIDEO 0~7 _IR is transferred in synchronization with the signal. Here, multi-value data of 8 bits per pixel is used.

As shown in FIG. 4, the image data input / output interface sends a signal IDR to the printer control unit 202.
EQ, VD_PR, LSYNC, HD_PR, SYNCK
_PR, VIDEO ₀ to ₇ Receive _PR. FIG. 6 shows a sequence of image data transferred to the printer control unit 202. Here, the IDREQ signal indicates a page data transfer start signal from the printer, and LLSYNC indicates
This is a one-line start reference signal from the printer, and transfers an image signal from the rotating memory unit 30 in synchronization with these signals. The VD_PR indicates a page data output active at a low level, and the HD_PR signal becomes a line data output active at a low level. VD_PR, when HD_PR are both active, valid image data _VIDEO 0~7 _PR is transferred in synchronization with SYNCK_PR signal.

FIG. 7 shows a rotary memory unit controller 300.
FIG. The image data transferred from the image reading unit 100 is stored in an input page memory 302 serving as a buffer.
And input to the document edge detection unit 304. As a result, the coordinate data of the document edge is generated by the coordinate data generator 306, and the obtained coordinate data is sequentially written to the stack memory 308. The input page memory 302 is managed by two-dimensional coordinates, and the stored image data is edited by the rotation processing unit 310 based on the document edge data in the stack memory 308 and transferred to the output page memory 312. Note that input / output of data for editing processing, command setting, and the like are performed by the CPU 314 based on a signal from the overall control unit 400. The output page memory 312 is also managed by two-dimensional coordinates, and internal image data is sequentially output at the time of printout.

FIG. 8 shows the document edge detection circuit 304. Image data VIDE sent from image reading unit 100
O _0-7 _IR is calculated based on the density of the background of the original and the original transport belt 5.
06 or a comparator 320 that compares the density with nothing on the platen glass 19. Image data V
IDEO _0-7 _IR (8-bit multi-valued data per pixel)
Is transferred in synchronization with the SYNCK_IR signal when line data is being output (HD_IR is active). By comparing each image data with reference data ref in consideration of a margin, the presence / absence of a document is reliably determined and converted into binary data. The shift register 322 at the next stage removes noise by performing processing in units of eight pixels. The output signal of the shift register 322 is input to an AND gate 324 and a NAND gate 326, and the output of both gates 324, 326 is then J-
It is input to the J and K inputs of the K flip-flop 328. The output signal of JK flip-flop 328 is
The negative logic input of the D gate 320 and another AND gate 322 are input. This JK flip-flop 3
28 is also input to the D flip-flop 334. The output signal of D flip-flop 334 is
The signal is input to a negative logic input of a D gate 320 and another AND gate 322. The HD_IR signal and the SYNCK_IR signal are input to the negative logic AND gate 336,
The output signal is supplied to the T terminal of the shift register 320, the JK flip-flop 328, and the D flip-flop 334. Due to this final stage configuration, the AND gate 3
Reference numerals 30 and 332 denote the non-existent edges (+
EDGE), the presence / absence edge of the document (−EDGE) is detected, and a one-shot pulse is output.

FIG. 9 shows the coordinate data generation circuit 306. HD_IR (CLK terminal) and V
By inputting D_IR (CLEAR terminal), an X coordinate on the sub-scanning side is generated. Similarly, SYNCK_IR
(CLK terminal) and the counter 342 to which HD_IR (CLEAR terminal) is input, the Y coordinate on the main scanning side is generated. The latch 344 temporarily latches the Y coordinate at + EDGE, subtracts 16 from it at the adder 346, and stores it in the stack memory 308 together with the X and Y coordinate data at -EDGE. The write address is sequentially updated by -EDGE in the CLK counter 348, and is initialized by VD_IR.

FIG. 10 is a block diagram of the rotation processing unit 310, and FIG. 11 is a diagram for explaining its operation. An affine transformation process is used for the image rotation process. This is expressed by the following equation using a geometric conversion method between coordinates.

(Equation 1) The affine transformation unit 350 transforms the data (image) in the xy coordinate system into the uv coordinate system according to equation (1), and performs parallel movement, enlargement, reduction, rotation, and the like of the image. . In the present embodiment, only processing of translation and rotation is performed. The affine transformation unit 350 shown in FIG. 10 specifies a rotation processing target area in the input page memory 302 by setting a rectangular area using four-point coordinates (see the left side of FIG. 11), and further edits the origin coordinates (U ₀ , V ₀ ), the origin (x, y) of the coordinates for the rotation processing and the rotation angle θ are designated (see the center of FIG. 11). Next, assign the origin (rotation coordinate) of the rotary Zumi image area relative to u-v coordinate the editing process (U _o, V _o)
(See the right side of FIG. 11). This can be expressed as follows.

(Equation 2) Also, the max coordinates (U _max ,
V _max ) is output. Since the coordinates (u, v) obtained by the affine transformation are not usually integers,
The output density value f (u, v) is converted to the density data f (X _n , Y
_It is necessary to interpolate using _n ). Density interpolation processing unit 352
Performs this interpolation. As an interpolation method, a nearest neighbor method, a linear interpolation method, a three-dimensional spline interpolation method, and the like have been proposed, but a detailed description thereof will be omitted here. The data complemented by the density interpolation processing unit 352 is sent to the output page memory 312 and stored according to the two-dimensional coordinate axes (uv coordinates). Data is output in line units according to the print timing.

FIG. 12 shows the output page memory 312, and FIG. 13 is a diagram for explaining its operation. After the data is stored in the output page memory 312, an unnecessary portion of the image can be erased. As a method, two erase area coordinates (U _erase0 , V _erase0 ),
By setting (U _erase1 , V _erase1 ), a rectangular area parallel to the coordinate axis which is a diagonal is converted into white data (see the left side of FIG. 13). Further, by setting the paper size coordinates (U _paper , V _paper ), the paper size becomes a data output area in a rectangular area having two points of origin coordinate and paper size coordinate as diagonal lines (see the right side of FIG. 13). . Here, the V axis is the main scanning direction, and the U axis is the sub scanning direction. By inputting an output enable signal to the output page memory 312, a signal VIDEO is output.

Next, document detection using the above-described system will be described. In this system, even for a document having a heading or the like or a document other than a rectangle, the inclination of the document and the document area can be determined with high accuracy and with few erroneous detections. The original placed on the platen glass is read. In FIG. 14, an image reading area indicated by a hatched portion (platen glass 19)
The area other than (white part) represents the original. In this example, the original is placed at an angle. In the image reading unit 100,
While the scanner 11 moves in the X-axis (sub-scanning) direction shown in FIG. 15, an image is detected by the CCD sensor 17 on a line-by-line basis (broken line position in FIG. 15). Here, the actual reading by the scanner 11 is not an ideal state in which the sampling pitch in the X direction is uniform as shown in FIG.
Normally, the sampling pitch in the sub-scanning direction varies due to the one pitch unevenness. This is because the sampling is processed by the timing signal (SYNCK_IR) of the hardware circuit as shown in FIGS. 8 and 9, that is, at a fixed time interval. However, since the operation of the scanner 11 in the image reading unit 100 involves vibration (pitch unevenness) in the sub-scanning direction, sampling is not performed at an ideal sampling pitch position. Therefore,
It is desirable to detect (the edge of) the document so as not to be affected by the variation in the sampling pitch.

The editing process based on the read data is performed in the rotating memory unit 30. CCD sensor 1
7 is transmitted in line units according to the sequence shown in FIG. 5 and stored in the input page memory 302 in the rotating memory unit 30. At the same time, the document edge detection unit 304 detects the document edge. 15 and 16, an edge (+ EDG) at which a white point changes from black to white
E), an edge at which the black spot changes from white to black (−EDG
E) is shown. These two sets of coordinate data are sequentially written to the stack memory 308 as a pair. In the stack memory 308, the line number X _n, + EDGE count value YW _m, is stored in the count value YB _m is set -EDGE, are picked up in the stored order is processed.

Thereafter, as shown in FIG. 17, the CPU 314 detects a change in inclination between the coordinate data of each document edge from the coordinate data stored in the stack memory 308, and detects a point at which the inclination changes (edge change). ) And extract line segments connecting these edge change points (in FIG. 16, line segments a, b, d, and c). In FIG. 18, a small circle indicates an edge change point, and a thick line indicates a detected line segment. From the coordinate data stored in the stack memory, a change in inclination between the coordinate data of each document edge is detected, and a line segment connecting the detected change points is extracted. Then, of the extracted line segments, a line segment of 45 ° or less with respect to the X direction (sub-scanning direction) (in the example of FIG. 18, the line segments 520a, 520c, and 520).
e, 520 g) are selected. By selecting a line segment of 45 ° or less with respect to the X direction (sub-scanning direction), an accurate inclination can be detected even if the reading pitch in the sub-scanning direction varies. Preferably, among the extracted line segments, the longest line segment (line segment 520a in the example of FIG. 18) that is 45 ° or less with respect to the X direction (sub-scanning direction) is selected. You. As a result, even for a non-rectangular document, an optimum document side serving as a basis for setting a document area is selected. Alternatively, preferably, among the extracted line segments, a line segment that is 45 ° or less with respect to the X direction (sub-scanning direction) and has a line segment that is orthogonal to the line segment (the line segment in the example of FIG. 18). 520a) is selected. Thereby, erroneous detection of the document area can be avoided.

Here, the reason for selecting a line segment having a predetermined angle (45 ° here) with respect to the sub-scanning direction among the extracted line segments is as follows. Since the edge of the original on one line is detected at two points, 45
Edge detection on a side having an angle exceeding ゜ (45 ° or less with respect to the main scanning direction) becomes very coarse as shown in FIGS. Therefore, the influence of the above-described variation in the sampling pitch is very likely to occur. Conversely, in sampling of a side substantially parallel to the sub-scanning direction, the amount of change in the address of the detected edge is very small even if the position slightly deviates from the ideal position. Therefore, different handling depending on whether it is below a predetermined angle,
A line segment having a predetermined angle or less with respect to the sub-scanning direction is selected.

Here, 45 ° is a threshold value for discriminating whether parallel to the main scanning direction or parallel to the sub-scanning direction, whereby two straight lines (main lines) in which the angle (inclination) of a certain straight line (side) is orthogonal. (The scanning direction and the sub-scanning direction). Since a normal document is rectangular, each side is orthogonal to each other. When the document is inclined by 45 ° with respect to the sub-scanning direction, all sides have an angle of 45 ° with respect to the sub-scanning direction. If one side of the document has an angle of 45 ° with respect to the sub-scanning direction, all sides have an angle of 45 ° with respect to the sub-scanning direction. A side having an angle exceeding 45 ° with respect to the sub-scanning direction is a side having an angle of 45 ° or less with respect to the main scanning direction. Further, a side adjacent to the side is a side having an angle of 45 ° or less with respect to the sub-scanning direction, and a side facing the side is a side having an angle exceeding 45 ° with respect to the sub-scanning direction. is there. Therefore, when attention is paid to a side substantially parallel to the sub-scanning direction, a side having an angle of 45 ° or less with respect to the sub-scanning direction may be searched for.

The inclination of the document is determined based on the line segment thus selected. Further, four coordinates (large circles in the figure) for defining a document area including all detected vertexes are determined. Further, the inclination direction and the rotation angle of the document are detected from the inclination of the longest line segment. FIG. 19 shows that X ₁ −X _min <Y ₁
The position of the document in the case of −Y _min is shown, and FIG. 20 shows the setting of the rotation angle of the document. The rotation is based on (X ₁ , Y _min ) as the origin, and the rotation angle θ is −tan ⁻¹ ｛(Y ₁ −Y _min ) /
(X _min −X ₁ )}. FIG. 21 shows that X ₁ −X _min >
FIG. 22 shows the position of the document in the case of Y ₁ −Y _min , and FIG. 22 shows the setting of the rotation angle of the document. The rotation is (X _min , Y ₁ )
And the rotation angle θ is −tan ⁻¹ ｛(Y _max −Y ₁ ) /
(X ₂ −X _min )｝. Based on the above results, each setting for the rotation processing and the parallel movement is performed in the rotation processing unit 310, and the rotation processing and the parallel movement processing are performed based on the settings. Then, the obtained image data is stored in the output page memory 312. The image forming section 200 forms an image based on the image data.

Next, the details of the image forming operation will be described with reference to FIGS.
This will be described with reference to the flowchart of FIG. The extraction of each line segment by detecting the edge change point and the calculation of the size, inclination angle, and shift amount of the document area are performed by the image reading control unit 10.
The editing process including the arbitrary angle rotation is performed by the rotation memory unit 30 under the control of 2, and the setting of parameters is performed by the memory unit control unit 300. Here, for the sake of simplicity, the control is described in the entire flow of one flow. FIG. 23 shows an overall flowchart of the present system. First, initialization is performed (step S11), and the operation panel 90
(Steps S12, see FIGS. 24 and 25), document transport by the document transport device 500 (step S13), and the document reading unit 1.
Then, an image input process (Step S14) is performed. Then, it is determined whether or not the input of the image data has been completed (step S15). If the input has been completed, the document is discharged (step S16). Next, it is determined whether or not the mode is the tilt correction mode (step S17). If the mode is not the tilt correction mode, a rotation process is executed in a through process (step S18, see FIG. 34), an image data output process is performed (step S19), and the process returns to step S12. In the determination in step S17, the platform in the inclination correction mode performs the document vertex detection process (step S20, see FIGS. 26 to 28), the area address setting process (step S21, see FIGS. 29 to 30), and the editing process. (Step S22, see FIG. 32)
An image data output process is performed (step S19, FIG. 33)
Reference), and return to step S12.

FIGS. 24 and 25 show input signal processing (FIG. 2).
3 is a flowchart showing details of step S12). First, the selection setting state of the document mode is determined by the ON edge of the document mode setting key 98 (change from the OFF level to the ON level) (step S101). The lighting state is determined (step S102). If it is turned on, the display unit 98a is turned off, the display unit 98b indicating selection of the document duplex mode is turned on, and the document duplex mode is set (step S103). The display unit 98 is determined in step S102.
If a is turned off, the display unit 98b is turned on, so that the display unit 98b is turned off, the display unit 98a indicating selection of the single-sided document mode is turned on, and the single-sided document mode is set (step S104). Next, the selection setting state of the copy mode is determined by the on-edge of the copy mode setting key 97 (step S105), and in the case of the on-edge, the lighting state of the display unit 98a indicating selection of the copy one-sided mode is determined (step S106). In the case of lighting, the display unit 97a is turned off, the display unit 97b indicating selection of the single-sided 2in1 mode is turned on, and the single-sided 2in1 mode is set (step S1).
07). If the display unit 97a is turned off in the determination of step S108, the lighting state of the display unit 97b is determined (step S108). In the case of lighting, the display unit 97b is turned off, the display unit 97c indicating selection of the single-sided 4in1 mode is turned on,
The single-sided 4-in-1 mode is set (step S109).
If the display unit 97b is turned off in the determination of step S110, the display unit 97c is turned on.
Is turned off, and a display section 97a indicating selection of the copy one-sided mode is displayed.
Is turned on to set the copy one-sided mode (step S1).
10).

Next, the selection setting state of the inclination correction mode is determined by the on-edge of the inclination correction mode setting key 99 (step S111), and in the case of the on-edge, the lighting state of the display section 99a indicating selection of the inclination correction mode is determined. (Step S112). If it is turned on, the display unit 99a is turned off, and the tilt correction mode is released (step S113). If the display unit 99a is turned off in the determination in step S112, the display unit 99a indicating selection of the tilt correction mode is turned on to set the tilt correction mode (step S114). Next, it is determined whether or not the start key 96 on the operation panel for instructing the start of copying has been pressed based on the on-edge of the start key 96 (step S119), and in the case of the on-edge, an empty sensor (not shown) of the document feeding unit 500 is determined. ) Is turned off (step S120), and if it is off, a scan start request is output (step S121). If it is determined in step S120 that the empty sensor is not off, the paper feed tray 5 of the document feeder (ADF) 500
Since the original is set in the ADF 10, the ADF
A start request is output (step S122).

FIGS. 26 to 28 show original vertex detection (FIG. 2) for detecting a vertex based on a change in the edge address of the original.
3 is a flowchart of step S20). First,
The coordinate data paired by the + EDGE coordinates and the -EDGE coordinates is decomposed (step S301). _{(X n, YW m, YB} m) → edge_first_adr (X n), edge_l
ast_adr ( _Xn ) where edge_first_adr ( _Xn ) is + EDGE of _xn line
Edge_last_adr (X _n ) is -ED of X _n line
GE coordinates.

Next, an edge-undetected line is extracted. First, variables and flags are initialized (step S30).
2). Then, the check is started from the line of x = 1. Leading edge address edge in one line of interest
_first_adr (x) and trailing edge address edge_last_adr
(x) is determined to be both ε1 or more (step S
303). Note that the value of ε1 is Y at the position of the original scale.
The direction address is appropriate. edge_first_adr (x) and ed
If both ge_last_adr (x) are greater than ε1, it is determined that an edge of the document has been detected, and the process proceeds to step S306. If any one of the edge addresses is smaller than ε1, it is determined that the line is a line for which no edge has been detected, and then a flag indicating whether or not the leading edge of the document has been detected.
Check flg_doc_top (step S304). If the leading edge of the document has not been detected (N in step S304)
O) indicates that each edge address of the line of interest is ε1
The process of steps S303 to S304 is repeated while incrementing x (step S305) until it becomes larger. If the leading edge of the document has been detected, it is determined that document vertex detection has been completed, and the process ends.

Next, the processing of the leading edge of the document will be described. First,
Flg_, which is a flag indicating whether the leading edge of the document has been detected
Check doc_top (step S306). When the leading edge of the document has not been detected (NO in step S306)
Determines whether or not each edge address of the unchecked x-1 line is larger than ε1 in the processing of the above-mentioned edge undetected line (step S30).
7) If both are greater than ε1, the value of the x-1 line is set as the start point of line segment detection of each edge address, and the following values are stored in the array of points_f and point_l for each vertex (step S).
308).

[Equation 3] point_f [0] [0] = x−1 point_f [0] [1] = edge_first_adr (x−1) (3) point_l [0] [0] = x−1 point_l [0] [1] = Edge_last_adr (x-1) If any is smaller than εl, the following values are stored in the array point_f and point_l for each vertex using the value of the x-line as the start point of line segment detection of each edge address (step S309) ).

[Equation 4] point_f [0] [1] = x point_f [0] [1] = edge_first_adr (x) (4) point_l [0] [0] = x point_l [0] [1] = edge_last_adr (x) and Then, 1 is set to the document leading edge detection flag flg_doc_top (step S310), and the process proceeds to step S311. If the leading edge of the document is detected (YE in step S306)
S) proceeds immediately to step S311.

Next, detection of a change in inclination (detection of a line segment (apex)) will be described. First, as shown in the following equation (5), leading edge addresses edge_first_adr (x−1) to edge_first in each of three consecutive lines before and after the line of interest.
_adr (x + 1) and the trailing edge address edge_last_adr (x
-1) The absolute value of the difference between edge_last_adr (x + 1) and the line of interest (x) is calculated (step S311).

Cline_f (x−1) = | edge_first_adr (x−1) −edge_first_adr (x) | cline_f (x + 1) = | edge_first_adr (x + 1) −edge_first_adr (x) | cline_l (x−1) = | edge_last_adr ( x-1) −edge_last_adr (x) | cline_l (x + 1) = | edge_last_adr (x + 1) −edge_last_adr (x) | (5) Next, as shown in the following equation (6), three consecutive lines before and after the line of interest Then, the absolute value of the difference between the inclinations of the lines is calculated (step S312).

Sub_f = | cline_f (x−1) −cline_f (x + 1) | sub_l = | cline_l (x−1) −cline_l (x + 1) | (6)

Next, the absolute value of the difference between the leading edge addresses calculated in step S311, line_f (x−
1) It is determined whether both cline_f (x + 1) are smaller than ε2 (step S313). The value of ε2 is 2
mm is appropriate. cline_f (x-1), cline_l (x
If any one of +1) is larger than ε2 (NO in step S313), it is determined that the line segment is substantially parallel to the main scanning direction, and immediately the step S for the trailing edge is performed.
Proceed to 317. Also, cline_f (x-1), cline_f (x +
1) Both are smaller than ε2 (step S313)
YES), it is determined that each leading edge address is not substantially parallel to the main scanning direction of the document because there is no rapid change in the edge address, and the process proceeds to step S314. Then, in order to determine whether or not they are on the same line segment, it is determined whether or not the absolute value sub_f of the difference between the inclinations of the leading edge addresses calculated in step S312 is smaller than ε3 (step S314). The value of ε3 is 8
Dots are appropriate. If sub_f is larger than ε3 (NO in step S314), it is determined that the inclination of the leading edge address has changed, and the process proceeds to the vertex setting process (step S315), where the number x of the target line and the number of the target line The leading edge address edge_first_adr (x) is stored in the leading edge vertex array point_f, and the leading edge vertex counter is incremented (step S31).
6). Then, the process proceeds to step S316. Also, sub_f
Is smaller than ε3 (YES in step S314)
Determines that the inclination of the trailing edge address is not a vertex of the side of the document because there is no rapid change, and the process immediately proceeds to step S317.

Similarly, in step S317, the difference cline_l of the trailing edge address calculated in step S311 is determined.
It is determined whether both (x−1) and cline_l (x + 1) are smaller than ε2. cline_l (x-1), cline_l (x +
If any of 1) is larger than ε2 (NO in step S316), it is determined that the line segment is substantially parallel to the main scanning direction, and the process immediately proceeds to step S321. Also,
If both cline_l (x-1) and cline_l (x + 1) are smaller than ε2 (YES in step S316), the rear edge addresses do not change abruptly, so that the original main scanning direction and the main scanning direction do not change. It is determined that they are not substantially parallel sides, and the process proceeds to step S318. In order to determine whether or not they are on the same line segment, the absolute value sub_l of the difference between the inclinations of the trailing edge addresses calculated in step S312 is ε3 It is determined whether it is smaller than. If sub_l is larger than ε3 (NO in step S318), it is determined that the inclination of the trailing edge address has changed, and the process proceeds to the vertex setting process (step S319). Then, the number x of the line of interest and the trailing edge address edge_las of the line of interest
t_adr (x) is stored in the array point_l for the trailing edge vertex (step S319), and the counter of the trailing edge vertex is incremented (step S320). Also, sub_
l is smaller than ε3 (YE in step S318)
In S), it is determined that the inclination of the trailing edge address is not the vertex of the side of the document because there is no rapid change, and the process immediately proceeds to step S320. In step S320, the line of interest x is incremented, and the processing from step S303 is repeated.

FIGS. 29 and 30 show an area address setting process for setting four-point coordinates for defining an area of a document from each vertex detected in the document vertex detection process (FIG. 23, step S20). (FIG. 23, Step S2
It is a flowchart of 1). First, the distance between the vertices is calculated (the length of the line segment is calculated). First, the distance between adjacent vertices is calculated (step S330). Here, since the purpose is to compare the distances between the vertices, the sum of the squares of the differences in the x and y directions is calculated as described below. The adjacent vertices are the first vertex of the leading edge (point_f [0] [0], point_f [0] [1]) and the first vertex of the trailing edge (point_l [0] [0], point_l [0]
[1]) and the last vertex of the leading edge (point_f [0]
[0], point_f [0] [1]) and the last edge of the trailing edge (point_l [0] [0], point_l [0] [1]). Vertices that match.

Length_f [line_count_f] = (point_f [point_f_count] [0] −point_f [point_f_count + 1] [0]) ² + (point_f [point_f_count] [1] −point_f [point_f_count + 1] [1]) ² length_l [line_count_l] = (Point_l [point_l_count] [0]-point_l [point_l_count + 1] [0]) ² + (point_l [point_l_count] [1]-point_l [point_l_count + 1] [1]) ² length_f [0] = (point_l [0] [0] ] −point_f [0] [0]) ² + (point_l [0] [1] −point_f [0] [1]) ² length_l [0] = (point_l [point_l_count] [0] −point_f [point_f_count] [0 ]) ² + (point_l [point_l_count] [1] −point_f [point_f_count] [1]) ² (7)

Next, the distances between the respective item points are compared, and the two vertices (x ₁ , y ₁ ) and (x ₂ , y ₂ ) having the longest distance are selected (step S331). Further, an equation of a straight line passing through the two selected vertices is calculated according to the following equation (step S332).

Y = (y ₂ −y ₁ ) / (x ₂ −x ₁ ) * X− (x ₁ y ₂ −x ₂ y ₁ ) / (x ₂ −x ₁ ) (8) Is (y ₂ −y ₁ ) / (x ₂ −x ₁ ), and the y intercept is − (x ₁ y ₂ −x ₂ y ₁ ) / (x ₂ −x ₁ ). Line [point_count] [0] and line [point_coun
t] [1].

Next, grouping is performed based on the obtained linear equation (slope and y-intercept) of each line segment based on whether or not the line segment is 45 ° or less with respect to the sub-scanning direction (step S).
332). Here, the slope (line [point_count] [0]) is
The case of “1” is a 45 ° line segment, and the sub-scanning direction corresponds to the Y direction, so that a line segment having a slope of “1” or more (line [point
By extracting _count] [0]> 1), it can be determined that the line segment is 45 ° or less with respect to the sub-scanning direction. next,
The longest line segment is extracted from the line segments of 45 ° or less with respect to the sub-scanning direction (step S33).
3). Next, in order to enhance the effectiveness of the line segment, it is checked whether or not a line segment substantially orthogonal to the line segment exists (step S334). It is deleted from the group of line segments of 45 ° or less with respect to the scanning direction (step S325), and step S333.
And repeat the process. If there is a line segment that is substantially perpendicular to the line segment (YES in step S334), the gradient line [point_count] [0] of the line segment is set as the document gradient doc_cline (step S336).

As described above, the longest and orthogonal line segment is extracted from the line segments of 45 ° or less with respect to the sub-scanning direction, and as will be described later, the extracted line segment is extracted. The inclination of the document is determined based on the minute inclination. For example, to explain an example of a non-rectangular document shown in FIG. 31, a group of line segments of 45 ° or less with respect to the sub-scanning direction is sides A and C. Here, the longest line segment among the line segments of 45 ° or less with respect to the sub-scanning direction is the side C. However, in the case of this original, it is not preferable to determine the inclination of the original based on the side C, as is apparent from the drawing. There are no orthogonal line segments on this side. Therefore, if the side C is excluded from the selection targets, the side A is considered next. Side A has orthogonal line segments B and D. Therefore, the side A is selected. As described above, in the case of a non-rectangular document as shown in the drawing, a correct side can be selected by examining whether an orthogonal line segment exists.

Next, in order to determine the area of the document, all the straight lines orthogonal to the straight line calculated in step S336 and passing through the vertices detected by the document vertex detection process are calculated (step S337). . The equation of the orthogonal line is:
Are substituted for the coordinates of each term, and b, which is the y-intercept, is obtained.

Y = −1 / doc_cline * X−b (9) and y of the straight line passing through each vertex calculated as described above
Among b the intercept, and calculates the maximum value Vb _max and the minimum value Vb _min (step S338). Step S336
Then, all the straight lines parallel to the straight line calculated in and passing through the vertices detected by the document vertex detection process are calculated (step S339). The equation of the parallel straight line is as follows. The coordinates of each vertex are substituted into X and Y of this equation, and b which is a y-intercept is obtained.

Equation 10] Y = doc_cline * X-b ( 10) Of the b is the y intercept of a straight line parallel through each vertex calculated in the step S339, and calculates the maximum value Hb _max and the minimum value Hb _min (Step S340).

By the above processing, the following four straight line equations surrounding the document area are calculated. The orthogonal straight lines are as follows.

Y = −1 / doc_cline * X−Vb _max (11) Y = −1 / doc_cline * X−Vb _min (12) The parallel straight lines are as follows.

Y = doc_cline * X−Hb _max (13) Y = doc_cline * X−Hb _min (14) Then, each of the above-mentioned parallel straight line and the orthogonal straight line whose y-intercept is the maximum value and the minimum value is By calculating the intersection, the coordinates of four points as the document area are determined (step S341). Next, from the coordinates of the determined four points, X
The minimum and maximum coordinates and the minimum and maximum Y coordinates are extracted, and (X _min , Y ₁ ), (X
_max , Y ₂ ), (X ₁ , Y _min ), and (X ₂ , Y _max ) (Steps S341 to S345). Thus, the document area is detected.

FIG. 32 shows the editing process (FIG. 23, step S
It is a flowchart of 22). Here, the actual processing contents of CPU data and command setting for the rotation processing unit 310 described with reference to FIGS. First, the coordinates of the four points extracted in the area address setting process (FIG. 23, step S21) are set as conversion area coordinates (step S401). Next, by comparing the distance between two specific coordinates, the inclination direction of the document is determined, and the rotation coordinate conversion and the rotation angle are set. Here, the setting is performed according to the rules shown in FIGS.

If X ₁ −X _min > Y ₁ −Y _min (YES in step S402), the origin of the rotational coordinates (x, y) is
(X _min , Y ₁ ) and the rotation angle θ are tan ⁻¹ ((Y _max −Y
₁ ) / (X ₂ −X _min )) (rotation direction is counterclockwise) (steps S403 and S404). X ₁ −X _min
If ≦ Y ₁ −Y _min (NO in step S402), the origin of the rotation coordinates (x, y) is (X ₁ , Y _min ) and the rotation angle θ
Is −tan ⁻¹ ((Y ₁ −Y _min ) / (X _min −X ₁ )) (the rotation direction is clockwise) (steps S 405 and S 4).
06). By following the above conditions, the position of the side serving as the reference for correction and the position of the reference coordinates are integrated at a small correction angle (45 ° or less). Then, edit origin coordinates (u ₀ ,
v ₀ ) is set (step S407), and when the preparation is completed (step S408), a rotation process is executed based on the above settings (step S409).

FIG. 33 is a flowchart of the image data output (FIG. 23, step S19). First, the paper size coordinates (U _paper , V _paper ) are set (step S501). Next, it is determined whether there is an area that needs to be erased (step S502). If there is an area that needs to be _erased , two points (U _erase0 , V _erase0 ) and (U _erase1 ) of the erase area coordinates are _determined. , V _erase1 ) are set (step S503). Thereafter, it is determined whether the erase setting has been completed (step S504), and if not, the process returns to step S503. When the erase setting is completed and the preparation is completed (YES in step S505), the output is enabled and the data transfer to the printer is permitted (step S506).

FIG. 34 is a flowchart of the through processing (FIG. 23, step S18) executed when the mode is not the tilt correction mode. First, the coordinates of the four points of the maximum size are
The coordinates are set as conversion area coordinates (step S601). Then, the rotation coordinates are set to the original image origin (step S60).
2), the rotation angle is 0 ° (step S603), and the editing origin coordinates (u ₀ , v ₀ ) are (α, β) (step S604).
Set as When ready (Step S60
5) A rotation process is executed based on the above settings (step S606).

[0047]

According to the present invention, when the original is placed at an angle, even if the reading pitch varies in the sub-scanning direction in reading the original, an appropriate original area and the inclination of the original are accurate and the number of erroneous detections is small. Detected, and according to the result, the arbitrary angle rotation correction of the document image is performed. Further, even if the shape of the original is not rectangular, an appropriate original area and the inclination of the original are detected. Thus, a desired copy can be easily obtained simply by placing the original on the platen without regard to the position on the platen, the inclination of the original, and the shape of the original.

[Brief description of the drawings]

FIG. 1 is a front view schematically showing the internal configuration of a copying machine.

FIG. 2 is a block diagram showing entire blocks of a control system.

FIG. 3 is a plan view showing a configuration of an operation panel.

FIG. 4 is a block diagram showing a configuration of an image input / output interface of a rotary memory unit control unit.

FIG. 5 is a diagram showing a sequence of image data transferred from the image reading control unit.

FIG. 6 is a diagram showing a sequence of image data to be transferred to an image forming control unit.

FIG. 7 is a block diagram of a rotary memory unit control unit.

FIG. 8 is a circuit diagram of a document edge detection circuit.

FIG. 9 is a circuit diagram of a coordinate data generation circuit.

FIG. 10 is a block diagram of a rotation processing unit.

FIG. 11 is a diagram illustrating the operation of a rotation processing unit.

FIG. 12 is a block diagram of an output page memory unit.

FIG. 13 is a diagram illustrating an operation of an output page memory unit.

FIG. 14 is a view showing a document set on a platen glass.

FIG. 15 is a view for explaining how a document is read.

FIG. 16 is a view for explaining a state in which a document is read when the sampling pitch for reading the document varies.

FIG. 17 is a diagram illustrating a change point and a line segment of an edge.

FIG. 18 is a view for explaining document area setting.

FIG. 19 is a diagram illustrating the direction of inclination of a document.

FIG. 20 is a diagram illustrating setting of a rotation angle.

FIG. 21 is a diagram illustrating the direction of inclination of a document.

FIG. 22 is a diagram illustrating setting of a rotation angle.

FIG. 23 is a flowchart showing the overall flow.

FIG. 24 is a flowchart of a part of an input process.

FIG. 25 is a flowchart of a part of the input process.

FIG. 26 is a flowchart of a part of document vertex detection.

FIG. 27 is a flowchart of a part of document vertex detection.

FIG. 28 is a flowchart of a part of document vertex detection.

FIG. 29 is a flowchart of a part of an area address setting process;

FIG. 30 is a flowchart of a part of an area address setting process.

FIG. 31 is a view for explaining an example of reading the inclination of a document.

FIG. 32 is a flowchart of an editing process.

FIG. 33 is a flowchart of image data output.

FIG. 34 is a flowchart of a through process.

[Explanation of symbols]

Reference Signs List 10 scanning system, 20 image signal processing unit, 30 rotation memory unit unit, 40 print processing unit, 60 laser optical system, 70 image forming system, 500 document transport unit,
302 input page memory, 304 original edge detection unit, 308 stack memory, 310 rotation processing unit,
312 output page memory, 314 CPU.

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Sou Hirota 2-3-13 Azuchicho, Chuo-ku, Osaka-shi, Osaka Inside Osaka International Building Minolta Co., Ltd. (72) Daisetsu Toyama Azuchi-cho, Chuo-ku, Osaka-shi, Osaka 2-3-13 Osaka International Building Minolta Co., Ltd.

Claims (3)

[Claims] An edge detecting unit that reads an original placed on a platen and detects an edge of the original on the platen; and extracts a line segment composed of continuous edges detected by the edge detecting unit. A line segment extraction unit, a document inclination detection unit that determines the inclination of a line segment having an inclination of 45 ° or less with respect to the sub-scanning direction among the line segments extracted by the line segment extraction unit as a document inclination. And a document area determining means for determining a document area having a tilt determined by the document tilt detecting means and including a line segment extracted by the line segment extracting means. 2. The original document inclination detecting device according to claim 1, wherein the inclination of the longest line segment among the line segments at 45 ° or less with respect to the sub-scanning direction is set as the original document inclination. Document detection device described. 3. The document inclination detecting means according to claim 1, wherein the inclination of the longest and orthogonal line segment among the line segments at 45 ° or less with respect to the sub-scanning direction is defined as the inclination of the document. 2. The document detecting apparatus according to claim 1, wherein: