JPS639266B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to a position correction circuit for correcting misalignment of read data using simple calculations without performing strict alignment when setting a document to be read in an image processing device, and obtaining desired image information. It is. When image processing is performed in units of one to several bits, a highly accurate reading device is required.
This type of reading device mechanically aligns the document using a reference line, fine adjustment mechanism, etc. when setting the document, and also uses advanced calculations to correct deviations such as rotational movement, parallel movement, etc. using the read information. The alignment accuracy was improved by performing the correction.
This means an increase in the cost of the mechanism for achieving accuracy and a delay in calculation time. In order to eliminate and improve such defects, the present invention detects the coordinates of the four corners of the document by painting the platen cover black or whiter than the white document.
Based on this value, a circuit that obtains information that corrects deviations such as rotational movement and parallel movement of the image using a simple arithmetic circuit can be used without special reference lines on the platen, fine adjustment mechanism, or document format limitations. This is what we provide. Generally, the following steps are normally performed from when image information is read until it is stored in a storage unit. Place the original in the reading position. Reads image information and detects its starting and ending edges. Detects the amount of tilt of the document and the amount of deviation in the horizontal and vertical directions. Perform tilt correction. Perform horizontal and vertical misalignment correction. In the conventional apparatus, the following means were used to implement the above-mentioned steps. A reference line is provided on the platen for precise positioning. Start and end marks are written at predetermined positions on the document and detected using software. This case has the disadvantage that it is limited to manuscripts in a certain format. Calculate the deviation of the start and end marks from the reference position described in section 2. in this case,
Since a fine adjustment mechanism must be provided and fine adjustment must be made during reading, there is a drawback that alignment requires time. , is calculated using the following formula in software, but it has the drawback that the calculation takes time. However, C _x : Amount of movement in the main scanning direction C _y : Amount of movement in the sub-scanning direction θ: Angle of rotation
Points moved by C _x and C _y The present invention eliminates and improves the drawbacks of the prior art described above. To this end, in the present invention, first, a straight line parallel to the main scanning direction is drawn on the platen cover, or the entire surface is made black, or the entire surface is painted in a color that can be distinguished from the white level of the original. , the coordinates of the four corners of the manuscript (A, B, C, D
point). Based on the coordinate values obtained as a result, check whether the misalignment of the original is as shown in Figure 1.
Alternatively, depending on the state shown in Figure 2, perform the calculations necessary to correct the deviation by using equations (1), (2), or (3), (4) below. , thereby performing a predetermined deviation correction. Here, m=tanθ=b/α−a However, a is the distance in the main scanning direction between corners C and A on the platen, b is the distance in the sub-scanning direction between corners C and A, and α is the corner The distance between C and B in the main scanning direction, X and Y are the coordinates after correction, and x and y are the coordinates before correction. As shown in FIG. 5, the present invention includes a first memory 30 for storing a read image signal, a second memory 31 for storing a corrected image signal, and a rotating coordinate calculation for calculating addresses of the first and second memories. It is composed of a circuit 8, a coordinate calculation circuit 9, a circuit 10 for detecting the coordinates of the four corners, and a timing generation circuit 7 for controlling timing signals of these circuits. The originals placed on the platen are read and sequentially stored in one memory 30. The coordinates at this time are generated by the coordinate calculation circuit 9 in response to the clock signal, and are simultaneously given to the first memory 30 as a memory address. Since the color of the platen cover is white or entirely black, which can be distinguished from the white level of the paper surface, the white signal of the paper surface can be easily separated. That is, the four corner coordinate detection circuit 10 is provided with eight registers for storing the x and y coordinates of each of the four corner points A to D of the original 5, and as will be described in detail later,
When the platen cover is completely black, for point A, the minimum y value at which a white signal appears and the x value at that time, and for point B, the maximum x value at which a white signal appears and the y value at that time. , For point C, the minimum x value at which a white signal appears and the y value at that time, and for point D, the maximum y value at which a white signal appears and the x value at that time, at the four corners A to D, respectively. are stored in these registers as the coordinates of . The coordinate values 48 of these four corners are calculated by the rotating coordinate calculation circuit 8.
and calculates a, b, α, θ, etc. in FIGS. 1 and 2. Next, by sequentially incrementing the coordinate values of the coordinate calculation circuit 9 from the initial values by the timing generation circuit 7, the image signal is read from the first memory 30, and at the same time, after the position correction calculated by the rotational coordinate calculation circuit 8, By giving the new coordinates 49 of , to the second memory 31 as a memory address, the image can be rotated,
By performing the parallel movement, the second memory 31 can store an image signal whose position has been correctly corrected. FIG. 6 is an example of a circuit that embodies the embodiment shown in FIG. 5 more specifically. Now, when scanning and reading a document placed on platen 1 as shown in Fig. 1 as shown in Fig. 3, if the platen cover is completely black,
The image signals 32 shown as #1 to #8 in FIG. 4 are stored in the first memory 3 in accordance with the respective scanning lines #1 to #8.
0 and is input to the timing generation circuit 7. 1st
In the memory 30, a value x obtained by counting the clock signal of the image by a main scanning direction coordinate counter and a sub scanning direction coordinate counter 12 which counts the number of scanning lines are stored.
The image signal 32 at that time is stored using the output y from as an address. When the platen cover is entirely black, as is clear from FIG.
The “black” level signal continues until the point is detected.
A white signal appears only when point A is detected.
The four corner coordinate detection circuit 10 detects the coordinate values (x _a ,
y _a ) in its register (not shown). Conversely, the point D (lower right corner) where the signal containing white has changed from being obtained continuously to a signal consisting only of black is detected and determined as the end of the document, and the coordinate value of that point D (x _d ,
y _d ). Also, the coordinates (x _c , y _c ) of point C, which is the lower left corner, correspond to the point where the x coordinate is the minimum among the points where the black signal changes to white signal, and the coordinates of point B ( x _b , x _b ) corresponds to the point where the x coordinate is maximum among the points where the white signal changes to the black signal. The coordinates of the four corners A to D obtained as described above are sent to the α calculation circuit 13, the a calculation circuit 14, and the b calculation circuit 15. Then, the values of α, a, and b are calculated by the following calculations. α=x _b −x _c a=(x _a −x _c ) or (x _b −x _a ) Smaller value, b=(y _d −y _c ) or (y _c −y _a ) small value. Also, the amount of parallel movement in the main scanning direction C _x (coordinate value x _c )
and the amount of parallel movement in the sub-scanning direction C _y (coordinate value y _a )
is transferred to the C _x arithmetic circuit 17 and the C _y arithmetic circuit 16 . Based on the above-mentioned values α, a, and b, the arithmetic circuit 18
23 perform the following calculations, respectively. 18...α-a 19...b/α-a=m 20...1+m ² 21...√1+ ² 22...

【formula】 23...

[Formula] When the reading into the first memory 30 is completed, the first memory 30 is set to the read mode, and the second memory 31 is set to the write mode. Then, the values of the main scanning direction coordinate counter 11 and the sub-scanning direction coordinate counter 12 are repeated from the initial value to the full count, and the original information is sequentially read out. The respective coordinate counter values x, y and the previously calculated value of m are calculated by the calculation circuits 24 to 29.
In addition, the following calculations are performed. 24...

【formula】 25...

【formula】 26...

【formula】 27...

[Formula] 28...X in equation (1) or (3) 29...Y in equation (2) or (4) Finally, outputs X and Y from arithmetic circuits 28 and 29
The information read from the x, y coordinate position of the first memory 30 is stored in the second memory 31 using as an address. Through the above operations, the second memory 31 has the following information:
The read document information is stored in a state in which the document is corrected for deviations such as inclination and parallel movement in the direction of the coordinate axes, and is placed in the correct position. In the above, for convenience of explanation, an example is shown in which the memories 30 and 31 are for two originals, but the memory is only the first memory 30 for one original, and the four corners A are used for the first reading. ~ Calculate α, a, b, m, etc. from the coordinate values of D, and at the second reading, use these values to correct the address X,
Supplying Y to the first memory 30 and storing data 2
A pass method can also be used. Also, reading is performed only once, and the read data from the first memory 30 is returned to the first memory 30 again, and X and Y obtained by the same calculation as described above are given to the first memory 30 as addresses, and the timing is generated. It is clear that a time-division method may be used in which the circuit 7 switches the read/write mode temporally. However, in the latter case, the condition that C _y ≠0 is required. Also, in the above embodiment, the platen cover was entirely black, but this was replaced with a straight line parallel to the main scanning line,
The document may be detected at a point where this straight line breaks. Furthermore, the color of the platen cover is
The color may be white or another color to the extent that the reading element can distinguish the level difference from the white portion of the document.
Furthermore, in the case of a thin document, the difference in level between black and white of the document may be reduced due to the influence of the black level of the platen cover. For this correction, a signal level determination circuit 50 may be added as shown by dotted lines in FIGS. 5 and 6. As is clear from the above description, according to the present invention, an input document can be read without the need for providing a special reference line on the platen or a fine adjustment mechanism for the platen, and without any restriction on the document format. There is an advantage that a correct image signal, which is obtained by rotating and translating the original image and correcting the deviation with respect to the reference coordinate axes, can be obtained easily and quickly.

[Brief explanation of the drawing]

1 and 2 are plan views showing a document placed on a platen at an angle with respect to the coordinate axis and shifted in the axial direction, and FIG. 3 explains the situation in which the document in FIG. 1 is scanned. 4 is a waveform diagram showing the electric signal obtained at that time, FIG. 5 is a block diagram of one embodiment of the present invention, and FIG. 6 is a detailed block diagram thereof. 7... Timing generation circuit, 8... Rotation coordinate calculation circuit, 9... Coordinate calculation circuit, 10... Four corner coordinate detection circuit, 11... Main scanning direction coordinate counter, 1
2... Sub-scanning direction coordinate counter, 13... α arithmetic circuit, 14... a arithmetic circuit, 15... b arithmetic circuit, 16... C _y arithmetic circuit, 17... C _x arithmetic circuit,
30...first memory, 31...second memory, 32
...image signal.

Claims (1)

[Claims] 1. A scanning photoelectric conversion device that reads information on a document placed on a platen and outputs a time-series electrical signal, and the difference between the output signal when the platen cover is scanned and the output signal when the document is scanned. a device that detects the coordinates of the four corners of a document based on the coordinates of the four corners; a device that calculates the tilt of the document on the platen with respect to the coordinate axes and the amount of deviation in the direction of the coordinate axes based on the coordinates of the four corners; A document position correction device comprising: a device that calculates new coordinates after correction for each point on a document; and a memory that stores information about each point on the document according to the new coordinates.