JPH07234910A - Mark sheet reader - Google Patents

Abstract

PURPOSE:To precisely correct the coordinate of a read binary mark when a mark sheet is fed obliquely at a certain angle. CONSTITUTION:A linear all mark 14 set in parallel with the lines of the binary marks 11, 12 and continued with length extending over the entire length of the lines is provided on the mark sheet 10. Also, this reader is equipped with an all mark detecting part 22 which detects an all mark by discriminating the continuous state of a dot from data read by an image sensor 15, an oblique feed detecting part 23 which finds the inclination ratio of a detected all mark, and an inclination correcting part 24 which corrects the coordinate of the binary mark based on the inclination ratio of the all mark.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mark sheet reading device, and more particularly to a mark sheet, which is filled or left blank without filling, arranged in a matrix with timing marks for each line. The present invention relates to a mark sheet reading apparatus for reading a mark from a mark sheet provided as described above, and more particularly to improvement of a means for correcting the coordinates of read data when the mark sheet is skewed (conveyed at an angle).

[0002]

2. Description of the Related Art As a skew correction means for a mark sheet in a conventional mark sheet reading apparatus of this type, a method of physically correcting the inclination of the mark sheet itself by a transport mechanism is generally used. However, such a physical skew feeding correction means has a drawback that it causes an increase in the size of the entire mark sheet reading apparatus and is expensive because the conveyance mechanism section becomes complicated.

On the other hand, Japanese Laid-Open Patent Publication No. 3-15970 discloses a means for correcting skew by image processing after reading. FIG. 9 shows the processing method. In the figure, the conveyance direction of the mark sheet 50 is the x-axis direction,
If the direction perpendicular to this is the y-axis direction, the element is 1
The line-type image sensors 51 arranged in rows are installed parallel to the x-axis. On the mark sheet 50, filled-on marks 52 and off-marks 53 which are left blank without filling are arranged in a matrix, and filled timing marks 54 indicating each line of the matrix. Is provided on one side edge of the mark sheet 50. Furthermore, on the mark sheet 50,
In front of the first line of the matrix, the two left and right skew feed detection marks 55 and 56 are filled with a constant interval W.
It is provided after. These skew detection marks 55
56 is located on one line parallel to the matrix line.

Now, in FIG. 9, it is assumed that the mark sheet 51 is skewed with respect to the image sensor 51 at an angle α. Image sensor 51 is mark sheet 50
After reading all the marks above, 2 from the data
One skew detection mark 55/56 is detected, the difference in the y-axis direction between these skew detection marks 55/56 is obtained, and this difference is converted into the number of lines to be the skew line number T. The triangular ratio (T / W) between T and the distance W between the two skew detection marks 55 and 56 is regarded as the skew inclination α of the mark sheet 50. Then, in correcting the coordinates of each on mark 52 and each off mark 53, the on mark 52 is corrected.
Alternatively, the line before the line to which the off mark 53 belongs is referred to. The reference line is determined as follows.

If the number of skew lines is T, and the inclination of the mark sheet 51 is rising to the right, the image sensor 51
In the pixel data of one line read in step 1, the pixel located at the distance n from the left end to the right, that is, the ON mark 52 or the OFF mark 53 corresponding to this pixel is (T / W) × n lines from the line. See the previous line. Further, when the inclination of the mark sheet 51 is rising to the left, in the one-line pixel data read by the image sensor 51, a pixel located at a distance n from the right end to the left direction from the line is (T / W) × Reference the line n lines before. Here, assuming that the number of lines up to the line to be referred to is s, this conventional method uses a triangular ratio (s / n) of s and n.
It is assumed that n) is equal to the triangular ratio of T and W (T / W) obtained from the two skew detection marks 55 and 56 that are separated from each other.

[0006]

However, while the distance W between the two skew detection marks 55 and 56 is predetermined, the skew line number T is 2 from the read data of the image sensor 51. The two skew detection marks 55 and 56 are detected and obtained from the difference in the y-axis direction, and it can be said that the triangular ratio (T / W) accurately represents the actual inclination of the mark sheet 51. Absent. Further, the relationship of s / n = T / W is established only when the skew inclination α of the mark sheet 51 is small, and is not satisfied when the skew α is large. Since the coordinates of 52 and the off mark 53 are erroneously detected, the skew feeding cannot be accurately corrected.

An object of the present invention is to provide a mark sheet reading apparatus capable of accurately correcting the coordinates of a read binarized mark when the mark sheet is skewed at a certain angle.

[0008]

According to the present invention, marks which are binarized by being filled or left blank without being filled are arranged in a matrix together with timing marks for each line. A mark sheet reading apparatus comprising: a mark sheet; a conveyance mechanism for conveying the mark sheet; an image sensor for reading a mark on the mark sheet; and a mark detection processing section for detecting and discriminating a binarized mark from the read data. The feature is that it has the following configuration. That is, on the mark sheet, linear all marks that are parallel to the lines of the matrix and are continuous over the entire length of the lines are provided. The mark detection processing unit includes an all-mark detection unit that detects the all-mark by determining the continuous state of dots from the data read by the image sensor, and an inclination that determines the inclination rate of the detected all-mark. A line detection unit and a skew correction unit that corrects the coordinates of the binarized mark based on the inclination rate of the all mark are provided.

When two-dimensional coordinates are assumed for the read data of the image sensor, the inclination rate of the all mark is
It can be represented by the trigonometric ratio of a right-angled triangle having the all mark as the hypotenuse, that is, the ratio between the substantial length D of the all mark and the x-axis component length L. Of these two lengths, the x-axis component length L of the all mark is the difference between the coordinate X ₁ on the x axis at one end of the all mark and the coordinate X ₂ on the x axis at the other end (X ₂ −
X ₁ ), that is, the number of pixels between the coordinates of these two points on the x-axis. There are two methods for obtaining the actual length D of the all mark.

In the first method, in the case where the image sensor is a line type, if the scanning direction of the pixel is the x-axis direction and the transport direction by the transport mechanism is the y-axis direction, the x-axis component of the previously obtained all mark is obtained. From the length L, the detection time difference T in the y-axis direction at both ends of the all mark, and the transport speed V of the transport mechanism,
The actual length D of the all mark is calculated by the following mathematical expression (1).

D = √ {L ² + (V · T) ² } (1)

The second method is to store the read data of the image sensor as two-dimensional data in the x-axis direction and the y-axis direction in the memory, and then, at the two-dimensional data, coordinates (X ₁ , Y ₁₎ of both ends of the all mark. ) And (X ₂ , Y ₂ ), the actual length D of the all mark is calculated by the following mathematical expression (2).
Ask for.

D = √ {(X ₂ −X ₁ ) ² + (Y ₂ −
Y ₁ ) ² } (2)

[0014]

Now, the case of correcting the coordinates (X _n , Y _n ) of one filled on-mark on the n-th line will be described. The x-axis component between the timing mark of the line and the on-mark will be described. Where L _{n is} the distance and the real distance along the line is D _n , these distances and the real length D of the all mark and the x-axis component length L have the relationship of the following mathematical expression (3).

D _n : L _n = D: L (3)

The distance L _n of the x-axis component can be obtained from the number of pixels of the x-axis component between the timing mark and the on mark. On the other hand, the actual distance D _n along the line can be obtained from the following formula (4).

D _n = L _n · (D / L) (4)

Then, from the ratio (D _n / P _M ) of the obtained distance D _n and the predetermined mark pitch P _M (the interval between marks on the same line), the on-mark is n lines. It is possible to determine what number mark the eye has. The number of the line is determined by detecting the timing mark.

[0019]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 shows a mark sheet reading apparatus according to the first embodiment of the present invention together with a mark sheet. As in the conventional case, the mark sheet 10 has a mark that is binarized by filling it or leaving it blank without filling it, that is, a large number of filled on marks 11 and a large number of blank off marks 12. , Are arranged in a matrix together with the timing marks 13 for each line.
The mark pitch P _M between the marks of each line is constant, and the line pitch P _L between the lines is
Is also constant. Although the timing mark 13 is provided on the left side edge of the mark sheet 10 in FIG. 1, it may be provided on the right side edge or on both the left and right side edges.

Further, the mark sheet 10 is provided with a single linear all mark 14 in front of the above-mentioned matrix-shaped mark arrangement region. The all mark 14 is continuously painted from the left end of the mark sheet 10 in parallel with the matrix line and over the entire length of the line.

The mark sheet reading apparatus shown in FIG. 1 conveys the line type image sensor 15 which is sufficiently longer than the width of the mark sheet 10 on the left and right, and conveys the mark sheet 10 in a direction orthogonal to the image sensor 15 by driving a motor 16. The reading control unit 19 that controls the reading operation of the mechanism 17 and the image sensor 15 and stores the read data in the memory 18, and the motor 16 that controls the carrying speed of the carrying mechanism 17 to be constant. A timer 20 for measuring the time required to detect the mark 14 and a mark detection processing unit 21 for detecting and determining the mark from the data stored in the memory 18 are provided.

The mark detection processing section 21 detects the all mark 14 by judging the continuous state of the dots from the data stored in the memory 18, and the inclination rate of the detected all mark 14. And a skew feeding correction unit 24 that corrects the coordinates of the on mark 11 based on the tilt rate of the all mark 14.
And a discrimination processing unit 25 that performs a predetermined discrimination process or the like on the corrected mark.

Assuming two-dimensional coordinates in which the pixel scanning direction of the image sensor 15 is the x-axis direction and the mark sheet transport direction by the transport mechanism 17 is the y-axis direction, the mark sheet 10 is shown in FIG. 2 with respect to the x-axis. To the right or as shown in Figure 3
As shown in (3), if all the marks 14 are inclined at an angle α to the left, the all mark 14 is also inclined by the same angle α with respect to the x-axis. Here, assuming a right-angled triangle having the all mark 14 as the hypotenuse, the length of the hypotenuse (substantial length of the all mark 14) is D, and the length of the horizontal side parallel to the x axis (x of the all mark 14). The length of the axis component is L, and the length of the vertical side parallel to the y-axis (y of the all mark 14
If the axial component length) is H, and L and H are given,
The angle α can be calculated from the triangular ratio of L and H, and D
Can be obtained from L and H by the following equation (5).

D = √ (L ² + H ² ) (5)

Therefore, the mark sheet reading apparatus of FIG.
First, the all mark 14 is detected from the data read by the image sensor 15 and stored in the memory 18, L is obtained from the x-axis coordinates of both ends thereof, and the detection time difference T in the y-axis direction at both ends thereof and the conveyance mechanism 17 are detected. H is obtained from the transport speed V, and the processing will be described with reference to the flowchart of FIG. Here, a variable representing the number of dots of the read data of the image sensor 15 is n, and its maximum value is n _MAX ,
A variable indicating the x-axis coordinate of a certain dot is x _n , a variable indicating the y-axis coordinate is y _n , a variable indicating the value of the timer 20 shown in FIG. 1 is t, and a dot detection flag is FL. The timer 20 is constantly counting up.

In FIG. 4, first, in step S1, x _n and y _n are cleared as initial settings, and n = 0,
After setting FL = 0 and clearing the timer 20 in step S2 to set t = 0, the presence or absence of dots is determined in step S3. If there is a dot, the process proceeds to step S4, the x-axis coordinate of the dot is substituted for x _n , and
The product (V · t) of the transport speed V of the transport mechanism 17 and the value t of the timer 20 is substituted for the variable y _n of the y-axis coordinate, and n =
n + 1 is incremented to n and FL =
After setting the dot detection flag to 1 and turning on the dot detection flag, the process returns to step S3 to move the coordinates for detecting the dots of the all mark 14 and repeat the same operation.

When there is no dot in step S3, it is determined in step S5 whether the flag FL is 0. If FL = 0, the process returns to step S2. With FL is substituting n-1 to n _MAX proceeds to not 0 step S6, since by substituting the current value t of the timer 20 in the detection time difference T in step S7, in step S8 the angle α of the following Formula (6)
Ask by.

Α = tan ⁻¹ {(V · T / (x _MAX −
x ₀ )} (6) where x ₀ is the x-axis coordinate value of the dot found first in steps S3 and S4, and x _MAX is the x-axis coordinate value of the dot found last. is there.

Next, from step S9 to step S14, the continuous state of the dots is confirmed to detect that it is the all mark 14. First, in step S9, n =
After being set to 0, n is incremented in step S10 and the process proceeds to step S11. In step S11, the x-axis coordinate value and the y-axis coordinate value of the n-th dot are x _n and y _n , and the x-axis coordinate value and the y-axis coordinate value of the first dot are x.
_{Given 0} and y ₀ , it is determined whether or not the following expression (7) is satisfied.

Tan ^-1 {(y _n -y ₀ ) / (x _n-
x ₀ )} = α (7)

If this is satisfied, the process proceeds to step S12, and the n-th dot is considered to be a part of the all mark 14, and if not satisfied, the process proceeds to step S13, and it is not considered to be a part of the all mark 14. . When you have to be part of all marks 14, it is determined whether n = n _MAX in the next step S14, returns to n _MAX Otherwise step S10, the same processing is repeated by incrementing the n. When n = n _MAX in this step S14,
Since the all mark 14 is finally detected, the process proceeds to step S16, where (x _MAX −x ₀ ) is substituted as the x-axis component length L of the all mark 14, and
The substantial length D of the all mark 14 is obtained by the above mathematical expression (1).

Thus, the mark sheet reading apparatus of FIG. 1 thus has the substantial length D and the x-axis component length L of the all mark 14.
After calculating and, the timing mark 13 and the on mark 11 are detected and the coordinates of the on mark 11 are obtained. The process will be described with reference to the flowchart of FIG. Note that this flow chart exemplifies a case where the mark sheet 10 is skewed to the upper right as shown in FIG.

In FIG. 5, after the read data for all the marks are stored in the memory 18 in step S21,
In the next step S22, all marks 1 that have been detected
The dot of 4 is excluded, and the coordinates of the dot (corresponding to the timing mark 13 of the first line) having the smallest coordinate in the x-axis direction are searched from the memory 18, and this coordinate is (x _T , y _T ).
And Furthermore, in the next step S23, y is more than y _T.
A dot having the next smallest coordinate in the axial direction (corresponding to the on mark 11) is searched from the memory 18 and the coordinate is (x _s , y _s ).
And Then, in step S24, the distance D _n is calculated from the following equation (8).

D _n = (x _s −x _T ) · (D / L) (8)

Next, in step S25, this distance D _n
The position of the on-mark 11 is calculated from the ratio (D _n / P _M ) of the mark pitch P _M (the interval between marks on the same line) to a predetermined mark pitch P _M. The dot associated with the on-mark 11 for which this calculation has been completed is deleted from the memory 18 in the next step S26.

In the next step S27, it is judged whether or not there is still a dot having a coordinate smaller than y _{T in the} y-axis direction in the memory 18, and if there is, a return is made to step S23 and the next on-mark 11 on the same line. Similarly, the position of is calculated. If there is not, it means that the processing of the first line has been completed, so the process proceeds to step S28, and it is determined whether or not there is a dot having the next smallest coordinate in the x-axis direction as a rank. If so, the process proceeds to step S29, and the dot having the smallest x-axis coordinate (corresponding to the timing mark 13 of the second line in this case) among them is searched from the memory 18 and the coordinate is set to (x
_{(T 1} , y _T ), the process returns to step S23, and the position of the on-mark 11 of the second line is calculated in the same manner as above. By repeating such an operation, the position of the on-mark 11 of each line is calculated one by one. Step S
If there is no dot having the next smaller coordinate in the x-axis direction at 28, it means that the timing mark 13 has disappeared and the process is terminated.

Next, FIG. 6 shows a mark sheet reading apparatus according to the second embodiment of the present invention. This mark sheet reader
A matrix-type image sensor 35 that captures a range that fills the entire area of one of the mark sheets 10 as an imaging range 30 at a time, and the reading operation of the image sensor 35 is controlled to read the two-dimensional data of the imaging range 30. And a mark detection processing unit 41 for detecting and discriminating marks from the two-dimensional data stored in the memory 38. FIG. 7 schematically shows the storage state of the memory 38. In addition, in FIG. 6, the conveyance mechanism for conveying the mark sheet 10 is omitted.

The mark detection processing section 41 includes an all-mark detection section 42 that detects the all-mark 14 from the data stored in the memory 38, and the detected all-mark 14.
A skew detection unit 43 for obtaining the inclination rate of the on-mark 11, a skew correction section 44 for correcting the coordinates of the on-mark 11 based on the inclination rate of the all mark 14, and a predetermined determination process for the corrected mark. It is divided into a determination processing unit 45 to perform.

The mark sheet reading device shown in FIG.
First of all 2 from the two-dimensional data stored in 8
4 is detected, and the all mark 14 is detected from the x-axis coordinates of both ends.
The x-axis component length L of is calculated, and the substantial length D of the all mark 14 is calculated from the x-axis coordinates and the y-axis coordinates of both ends thereof. The process will be described with reference to the flowchart of FIG. Here, a variable that represents the value of the X-axis coordinate of a certain point is X, a variable that also represents the value of the Y-axis coordinate is Y, a variable that represents the number of dots of data is n, and its maximum value is n _MAX , and a certain dot. Let x _{n be} the variable that represents the x-axis coordinate, and y _n be the variable that also represents the y-axis coordinate.

In FIG. 8, first, in step S31, Y =
After 0, the Y-axis coordinate value Y is incremented by setting Y = Y + 1 in step S32, and X in step S33.
= 0, then X = X + 1 is set to X in step S34.
The axis coordinate value X is incremented. And step S
At 35, it is determined whether or not there is a dot at the point (X, Y) indicated by the X-axis coordinate value and the Y-axis coordinate value after the increment, and if there is a dot, the process proceeds to step S36, where x _n is X-axis coordinate value X of y to y _n then Y-axis coordinate value Y
And n is incremented by setting n = n + 1, the process returns to step S34, and the detected coordinates are moved in the X-axis direction to perform steps S34, S35, and S3.
The process of 6 is repeated.

When there is no dot in step S35, the process proceeds to step S37, and it is determined whether the X-axis coordinate value X is smaller than the maximum value X _MAX . If it is smaller, step S3
Returning to step 4, the processes of steps S34, S35, S36 and S37 are repeated. If the X-axis coordinate value X reaches the maximum value X _MAX in step S37, the process proceeds to step S38, it is determined whether the Y-axis coordinate value Y is smaller than the maximum value Y _MAX , and if it is smaller, the process returns to step S32 to step. The processing from 32 to step S38 is repeated.

In step S38, the Y-axis coordinate value Y is the maximum value Y.
_{When MAX} is reached, it means that the detection of all dots related to the all mark 14 has been completed in both the X-axis direction and the Y-axis direction, so the routine proceeds to step S39, where n _MAX = n−
Then, the angle α is obtained by the following mathematical expression (9) in step S40.

Α = tan ⁻¹ {(y _MAX −y ₀ ) / (x
_MAX- x ₀ )} (9) where x ₀ is the x-axis coordinate value of the dot first discovered in steps S35 and S36, and x _MAX is the x-axis coordinate of the last discovered dot. It is a value.

In the next steps S41 to S46, the continuous state of the dots is confirmed to detect that it is the all mark 14, and the processing is the same as the processing from steps S9 to S14 shown in FIG. Therefore, the description is omitted.

When the all mark 14 is finally detected by the processing from step S41 to step S46,
Proceeds to step S47, the conjunction is substituted as x-axis component length L of all marks 14 (x _MAX -x _0), obtaining the real length D of all the mark 14 by the equation (2).

Thereafter, the mark sheet reading apparatus of the second embodiment also detects the timing mark 13 and the on mark 11 by the processing as shown in FIG. 5 similarly to the first embodiment, and detects the coordinates of the on mark 11. Ask for.

[0048]

As described above in detail, according to the present invention, the mark sheet is provided with linear all marks which are parallel to the lines of the binarized marks arranged in a matrix form and which are continuous over the entire length of the lines. Then, this all mark is detected and its inclination rate, that is, the substantial length D and x of the all mark is detected.
Since the ratio with the axial component length L is obtained and the coordinates of the binarized mark are corrected based on this ratio, the correction can be performed accurately.

[Brief description of drawings]

FIG. 1 is a block diagram showing a mark sheet reading apparatus according to a first embodiment of the present invention together with a mark sheet.

FIG. 2 is a plan view showing a mark sheet that is obliquely rising to the right.

FIG. 3 is a plan view showing a mark sheet that is skewed to the upper left as opposed to FIG.

FIG. 4 is a flowchart showing a process of detecting an all mark and calculating a substantial length D and an x-axis component length L by the mark sheet reading apparatus of FIG.

5 is a flowchart showing a coordinate correction process performed after the process shown in FIG.

FIG. 6 is a block diagram showing a mark sheet reading apparatus according to a second embodiment of the present invention together with a mark sheet.

7 is a schematic diagram showing a storage state of the memory shown in FIG.

8 is a flowchart showing a process of detecting an all mark and calculating a substantial length D and an x-axis component length L by the mark sheet reading apparatus in FIG.

FIG. 9 is a schematic diagram for explaining a skew feeding correction method by a conventional mark sheet reading device.

[Explanation of symbols]

 10 mark sheet 11 on mark 12 off mark 13 timing mark 14 all mark 15/35 image sensor 17 transport mechanism 18/38 memory 21/41 mark detection processing unit 22/42 all mark detection unit 23/43 skew feeding detection unit 24/44 Skew correction unit 25/45 Discrimination processing unit

[Procedure amendment]

[Submission date] December 20, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] 0040

[Correction method] Change

[Correction content]

The mark sheet reading device shown in FIG.
First of all 2 from the two-dimensional data stored in 8
4 is detected, and the all mark 14 is detected from the x-axis coordinates of both ends.
The x-axis component length L of is calculated, and the substantial length D of the all mark 14 is calculated from the x-axis coordinates and the y-axis coordinates of both ends thereof. The process will be described with reference to the flowchart of FIG. Here, a variable that represents the value of the x- axis coordinate of a certain point is x , a variable that also represents the value of the y- axis coordinate is Y, a variable that represents the number of dots of data is n, and its maximum value is n _MAX , a certain dot. a variable representing the x-axis coordinate x _n, a variable also represents the y-axis coordinates and y _n of.

[Procedure Amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0041

[Correction method] Change

[Correction content]

In FIG. 8, first, in step S31, Y =
After 0, the y- axis coordinate value Y is incremented by setting Y = Y + 1 in step S32, and X in step S33.
= 0, X is set to X = X + 1 in step S34, and x is set.
The axis coordinate value X is incremented. And step S
At 35, it is determined whether or not there is a dot at the point (X, Y) designated by the x- axis coordinate value and the y- axis coordinate value after the increment. If there is a dot, the process proceeds to step S36, where x _n is of the x-axis coordinate values X, y-axis coordinate value at that time to _{y n} Y
And increment n by setting n = n + 1, and then returns to step S34 to move the detected coordinates in the x- axis direction to steps S34, S35, S3.
The process of 6 is repeated.

[Procedure 3]

[Document name to be amended] Statement

[Correction target item name] 0042

[Correction method] Change

[Correction content]

When there is no dot in step S35, the process proceeds to step S37, and it is determined whether or not the x- axis coordinate value X is smaller than the maximum value X _MAX.
Returning to step 34, steps S34, S35, S36, S37
The process of is repeated. If the x- axis coordinate value X reaches the maximum value X _MAX in step S37, the process proceeds to step S38,
It is determined whether or not the y- axis coordinate value Y is smaller than the maximum value Y _MAX . If it is smaller, the process returns to step S32 and the processes from step 32 to step S38 are repeated.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0043

[Correction method] Change

[Correction content]

In step S38, the y- axis coordinate value Y is the maximum value Y.
_When reaching _MAX, it means that detection of all the dots relating to the all mark 14 has been completed in both the x- axis direction and the y- axis direction, and therefore the process proceeds to step S39 and n _MAX =
After setting n−1, the angle α is calculated by the following formula (9) in step S40.

Claims (3)

[Claims] 1. A mark sheet provided with binarized marks arranged in a matrix together with timing marks for each line by filling or leaving blank without filling, and the mark sheet is conveyed. A mark sheet reading apparatus comprising: a conveyance mechanism; an image sensor for reading a mark on a mark sheet; and a mark detection processing unit for detecting and discriminating a binarized mark from read data of the mark sheet. A linear all-mark continuous in parallel with the entire length of the line is provided, and the mark detection processing unit determines the continuous state of dots from the data read by the image sensor, All mark detection means for detecting all marks and the detected all mark Mark sheet reader that the skew detector means for determining the oblique ratio, it has and a skew correction means for correcting, based coordinates of the binarized mark ramp rate of all marks, characterized. 2. When the scanning direction of the pixels of the image sensor is the x-axis direction and the conveying direction of the mark sheet by the conveying mechanism is the y-axis direction, the skew feeding detecting means detects the x of all marks detected by the all-mark detecting means. From the means for obtaining the axial component length L, the means for detecting the detection time difference T in the y-axis direction at both ends of the all mark, the detection time difference T, the transport speed V of the transport mechanism, and the x-axis component length L. A skew correction means for correcting the coordinates of the binarized mark on the basis of the ratio between the real length D of the all mark and the x-axis component length L. The mark sheet reading device according to claim 1, wherein: 3. A memory for storing read data of the image sensor as two-dimensional data, wherein the skew feeding detecting means detects all marks from the coordinates of the two-dimensional data stored in the memory in the x-axis and y-axis directions. The x-axis component length L and the substantial length D are obtained, and the skew feeding correcting means corrects the coordinates of the binarized mark based on the ratio between the x-axis component length L and the substantial length D of the all mark. The mark sheet reading device according to claim 1.