JPS5833778A - Mark processing system - Google Patents

Abstract

PURPOSE:To obtain a mark reading rate with high accuracy, by correcting the effect of a card skew at carrying as well as expansion and shrinkage of paper. CONSTITUTION:A photoelectric conversion signal is stored in a picture memory 2. If a sheet 7 is normally contained in a picture memory 8, the outermost location of each direction can be informed from the projections to both X and Y directions. Thus, a television scanning region to detect coordinates of four end points of a document is set. A coordinate (ax, ay) of B at a point A is scanned toward the X between projections III and III' in Y direction. Then, the change in side of the forms is continuous between them, the document is not jammed and brokened and the position of the end points is normal; those can be confirmed. Simultaneously, the coordinate ax, ay of the end points can be informed.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for processing marks on a form read by a character reading device or the like.

In conventional mark reading devices, a timing mark is placed at the top or bottom edge of a form, data mark areas are set on the form according to this and the sheet size, and the mark is determined based on the black area (reflectance) within each data mark area. However, data tended to be missed due to the form being lifted off the base surface or skewed, making it difficult to read the data. Although it is possible to correct the positional deviation of the mark due to the skew to some extent, since the detector used a solar cell/photodiode whose position from the form surface was fixed, A complete correction could not be made for the disfigurement caused by the floatation. Also, it was difficult to prepare and adjust the transport to release the IIL.

The present invention has been made in view of the above-mentioned drawbacks, and it is possible to correct various irregularities caused by the conveyance of forms.
Father, data near the boundary value where it is difficult to determine the presence or absence of a mark can be determined by repeating rescanning while changing the slice level, and the validity of the mark can be checked not only from its area but also from its shape. The purpose is to provide a mark processing method.

The present invention will be explained in detail below using Examples.

FIG. 1 is a block diagram showing the configuration of an example of a device to which the present invention is applied. In the figure, (1) is a photoelectric conversion system, (2
) is an image memo +3 that stores images for l forms,
(3) is a controller unit that performs mark identification and overall control; (3) is an external interface unit that responds to control signals and transmits data with external equipment;

To explain the operation of FIG. 1, first, a form to be read is conveyed to the read head section, and charged and converted by the read head. It is suitable for the charging conversion element used here to be one that does not have moving optical parts and can convert the entire image on the form, such as an in-scan or surface sensor. The charging conversion signal is stored in the image memory (2), but the capacity of this memory needs to be larger than the maximum size of the form to be handled plus the amount of skew and expansion/contraction allowed. Second
As shown in the figure, the skew (5) is large and the memory size (
If the number exceeds 6), the sheet will be rejected. If the sheet (7) is not stored normally in the ms memory (8) as shown in FIG. IV, IV'
) to know the outermost position in each direction,
As a result, the television scanning area for finding the four end points of the form (m+) is set. And for example, #2 (However)
The coordinates of point A ("z +"y) are projected in the X direction using the charging conversion element.Each time television scanning is performed in the X direction between ■ and ", the change in the side of the form during that time is continuous. to ensure that the form is in good condition with no jams or tears, and that the end points are in the correct position.
At the same time, the values of the coordinates l x a y of the end points can be known.

The same applies to the other three points. At the same time, we also check whether there is any problem in recognizing the end point as an end point based on the linearity and constant slope of the end point where white appears for the first time during scanning.

It is very difficult to correct the skew of a form that is not pre-marked (- is printed in black and changed to ``Y'' by writing a mark), but according to the method of the present invention, the coordinates of the four end points are By comparing and calculating the size of the form included in the format information, it is possible to correct not only the skew but also the expansion/contraction to the capacity. First, when correcting the expansion/contraction when there is no skew, the format of the form For the form size included in the information, A to B,
Correction is performed using distances between A to C1C to D and 8 to 9.

Standard size (specified size) (Sx.

), then A, B', C, D, A, C', B, D
The correction rate (R1, R,) in the case of R1=A to B/8
x, R,==mu~D/8. A~B,C-,,
If there is a difference in D, set one of A to B, c, and n as the reference side, find the following inclination from the TV scanning results, perform coordinate conversion, and calculate the corrected coordinates in memory. .

As for the mark shape, since the mark pattern is stored in memory, you can use this to redo the preprocessing as needed until satisfactory data is obtained. It becomes easy to judge whether it is a mark or dirt not only from the appearance but also from the shape. In addition, if the image memory has a multi-level sheet buffer in order to increase the reliability of the data by changing the slice level and reading the marks using several threshold values, the standard slice level and area threshold are used for the first reading. At the same time, the average density and area of the data are determined, and the data between the marks determined to be data and the marks determined to be dirt are compared, and the relationship between each distribution and the standard slice level is determined. If it is determined that this is not the case, the slice level and area threshold are set again, and by rescanning, it is possible to correct parameters that depend on the line marking tool and the marking person, such as the voltage and the size of the mark.

The present invention is as described above, and has the following features: 1) It is possible to correct not only the expansion and contraction of paper but also the effects of card skew during conveyance, and a highly accurate mark reading rate can be obtained. 11) The shape of the data can be repeatedly corrected and the reliability of the data can be improved.

111) It is also possible to improve the accuracy by providing a multi-level sheet buffer and comparing the displayed data by changing the slice level. Alternatively, after reading the entire image at the average slice level, the average data density can be extracted, the area threshold value can be reset at the slice level, and rescanning can be performed. etc. results can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an apparatus to which the present invention is applied, and FIG. 12 (A) @ is a diagram for explaining the functions of the present invention. 1: Charging conversion system, 2: 1lii image memory, 3: Controller section, 4: Interface section. Agent Benzioshi Inoue - Male Figure 1 Figure 2

Claims (1)

[Claims] A photoelectric conversion system that converts the image on the form into an electrical signal,
an image memory that stores an image of the autumn equinox in the form after analog-to-digital conversion or quantization of the photoelectric conversion signal;
A controller section that controls the conveyance of the form, mark identification, and overall control; an image memory control section that controls data input/output to and from the image memory and address control; and a mark reading position based on the form peripheral information in the image memo 1. A memo 1J that stores all the images on the form is equipped with a function to correct the appearance information and a function to reset the reading parameters at the time of rescanning by feeding back various appearance information at the first face attack. (1) Set a TV scan area from the projection in the X and Y directions, obtain the coordinates of the end points of the form by TV scanning this area, and combine the coordinates of the obtained end points with the form size included in the form format information. A mark processing method characterized by performing comparative calculations to correct various fluctuations caused by the conveyance of forms.