JPH03242779A - Character information detecting system - Google Patents

Abstract

PURPOSE:To efficiently and strictly find out character size or a character interval by a simple constitution by computing the vertical/horizontal white length of a print based upon reading information optically read out and corrected by prescribed correction and finding out the character size or the character interval. CONSTITUTION:The character information detecting system is provided with an inclination correcting means 3 for detecting and correcting the reading inclination of reading information developed by a developing means 2 and developing the corrected information again, a computing means 4 for computing the vertical and horizontal white length of a print based upon the developed reading information and an output means for outputting the computed result of the means 4. Thus, the print is optically read out, binarized and developed as reading information, the reading inclination of the reading information is detected to correct the developed reading information and then the vertical and horizontal white length of the print is computed based upon the reading information. Consequently, the character size and the character interval can be detected and the detected result can be outputted and informed to an operator.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a character information detection method for detecting information on the size and spacing of characters on printed matter.

[Conventional technology]

In recent years, in the field of information processing interconnection, information processing devices that create printing information, such as office documents, etc., are coded and transmitted, and information processing devices that receive this information and print it according to the code. The number of cases where the models are different is increasing. Moreover, the number of cases in which information processing apparatuses not only from the same vendor (information processing apparatus supply organization) but from different vendors are connected is increasing. In the case of such interconnection between different types of devices, it is tested whether the code instructions created by the information processing device on the information sending side match the print results of the information processing device in the information receiving procedure. Therefore, a conformity test needs to be conducted. In this compatibility test, if the recipient's printing device does not have the character size or character spacing specified by the sender, it is permitted to print with an alternative character size or character spacing called a fallback.

[Problem to be solved by the invention]

By the way, when using an optical reading device for such a compliance test, the area in which characters are written is generally strictly defined on the sheet in advance, so the character dimensions and character spacing cannot be adjusted. could only be used fixedly for each sheet. Additionally, it is necessary to register a large amount of information for each person who fills in the information, which increases the system scale.

For this reason, conventionally, when the character size and character spacing are unknown on the receiving device, such as in printed matter printed with alternative character size and character spacing, which is called a fallback in the field of heterogeneous interconnection, the receiving device , the only way to judge the character size and spacing is by visually checking the character size and spacing with a ruler.
This method has the drawbacks of poor work efficiency and inability to measure accurately.

The present invention has been made to improve these conventional drawbacks.The purpose of the present invention is to improve the character size and character spacing by human beings when the character size and character spacing of obtained printed matter are unknown in the field of interconnection between different types of devices. The object of the present invention is to provide a character information detection method that can efficiently, accurately, and easily measure without requiring additional work.

[Means to solve the problem]

The character information detection method of the present invention includes an optical reading means for optically reading and binarizing a printed matter, receiving read information read by the optical reading means,
A developing means that virtually expands the read information to the same dimensions as when reading; a tilt correction means that detects and corrects the tilt of the read information expanded by the expanding means when reading the information; and a tilt correcting means that expands the expanded information again. It has a calculation means for calculating the vertical and horizontal white lengths of the printed matter by manipulating the read information, and an output means for outputting the calculation result by the calculation means.

[For production]

After optically reading the printed matter and converting it into binary information and developing it as read information, the tilt at the time of reading is detected and the developed read information is corrected, and this read information is manipulated to calculate the horizontal and vertical white length of the printed matter. do.

This allows the character size and character spacing to be detected, and the results are output and notified to the operator.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

FIG. 1 is a block diagram of one embodiment of the present invention.

This embodiment describes an optical reading means l that has a function of optically reading a printed matter printed on any flat medium such as paper and binarizing it so that digital processing is possible, and the optical reading According to the shading on the medium by means 1, white is “
0" and black are "1". A developing means 2 receives the binarized read information, temporarily stores the read information, and virtually expands it into the same dimensions as when reading, and an optical reading means 1. A tilt correction means 3 detects the tilt at the time of reading, corrects the binarized information developed by the developing means 2, and develops it again; calculation means 4 for calculating length;
It is comprised of a display means 5 for displaying the results of the calculation in order to transfer the results of the calculation means 4 and notify the operator.

Next, the operation of this embodiment will be explained. First, the optical reading means 1 optically reads a printed matter printed on any planar medium such as paper, and converts it into a binary value so that it can be digitally processed. This binarization process is performed using facsimile machines, existing image scanners, etc.
For each scanning line, white is binarized to "OII" and black is binarized to "1" during a certain minute interval.The density of this read information is determined by the interval of this minute interval, and is 200 dpi.
(unit expressing the number of dots per inch) or 240dp
densities such as i, 300dpi, 400dpi, etc. are currently -
It is used for boat fishing. Even higher-density devices are technically possible, but they require precision in the equipment and increase the amount of information, making them expensive, so they are not common except for special uses. In this embodiment, since the print density of ordinary office documents is within the above-mentioned printing density range, the print density is 200 dpi to 400 dpi.
is sufficient. Furthermore, if higher precision is required, a higher density image scanner may be used. Theoretically, it is desirable that the density be at least twice the density at the time of printing.

The read information thus binarized by the optical reading means 1 is transferred to the expansion means 2 and expanded on the main memory of the information processing device or on an external storage device such as a magnetic disk. In the developing means 2, the read information is managed as a collection of dots expressed as "O" for white and "1" for black.The amount of information is 1 scan if read at 200 dpi on a normal A4 paper size. This is about 1728 dots per line.In the developing means 2, the vertical and horizontal coordinates,
It is only necessary to determine whether the value of a dot at a certain coordinate position is "0" or "1", that is, whether it is white or black.

The correction means 3 corrects the read information of characters so that they are correctly developed in the flat and vertical directions when the optical reading means 1 reads a medium such as paper diagonally and binarizes it. One way to correct this is to add a black rectangular mark called a start mark or timing mark to the edge of the paper outside the character printing area, as is currently done with standard mark sheets. The size of the start mark or timing mark only needs to be uniquely determined within the device. Further, the start mark or the timing mark may be printed at the same time when the printer prints on the medium, or may be printed on the print medium in advance.

The correction means 3 identifies line deviations on the scanning line of the start mark or timing mark written at both ends of the same line on the medium, and calculates the medium length from the density information taken in by the expansion means 2. Alternatively, the deviation angle during reading is determined from the medium length taken in advance. Further, the correcting means 3 rotates the read information using this deviation angle by mathematical-dimensional transformation and develops the corrected value again. Another method for this correction is a statistical method performed by the calculation means 4. That is, among the cases where white continues at regular intervals for each scanning line, the one with the longest length of white and which appears periodically is determined and used as a temporary character interval. The deviation that occurs for each scanning line of this temporary character spacing is determined, and the deviation angle at the time of reading is determined. Thereafter, the read information may be rotated by a mathematical-order transformation as described above. However, with this method, when detecting the longest continuous white length, there is a risk that the value of the line spacing portion (hereinafter, the character spacing in the line spacing direction will be specifically referred to as the line spacing) may be mistakenly used.

This is because the deviation angle during reading is so large that the - scanning line may scan two or more lines. In this case, in order to avoid using too long a continuous white line as the character spacing, it is sufficient to limit the character spacing by giving a value below a certain value as the maximum value in advance. This maximum value can be given an empirical value through experimentation. Also, the worst condition is when the deviation angle is 45 degrees and the line spacing is the same as the left and right character spacing, but in the field of information processing interconnection, it is generally rare for the line spacing and character spacing to be the same. Furthermore, it is impossible for the deviation angle to be 45 degrees. The deviation angle is in the range of several degrees at most. This method has the advantage that it does not require a start mark or timing mark and can be applied to more general print media. To explain the operation of the calculation means 4 in more detail, the calculation means 4 operates on the developed reading information to calculate character size and character spacing using a statistical method. In other words, among the rows of black and white dot information in the read information, where white dots continue at regular intervals on each scanning line, the one with the longest white length and appearing periodically is determined and the character spacing is determined. shall be. When reading print media with the optical reading means 1, character spacing,
Black dots (hereinafter referred to as black dots) that occur at points within the line spacing that should be white are ignored because they appear as an exception during the statistical processing. It is not possible to determine in general what percentage of the white dot array should be garbage. If you allow the proportion of black in the white dots to a high value (that is, reduce the sensitivity), you will not be able to determine the character spacing, and if you keep it low (that is, if you increase the sensitivity), it will be difficult to determine the spacing for print media with many bumps. In this case, the character spacing cannot be determined.

A reasonable value can be determined empirically by experiment. Therefore, the proportion of black dots in the row of white dots (
Although it is desirable to be able to externally specify how much sensitivity (sensitivity) is allowed, it is not necessarily necessary to specify it externally, as a fixed value determined through experiments is sufficient for practical use.

The following method is available as a process for determining the length of white dots that appear periodically, that is, the character spacing. That is, first, it is assumed that the series of white dots is the smallest value, that is, one dot, and the period of appearance is assumed to be the maximum length of one scanning line.

(For A4 paper size, the scanning line is assumed to be 1728 dots.) Based on this assumption, repeat the assumed value of the period so that it becomes smaller in order of 1/2 of the scanning line length, then 1/3.1/4, and then the final Specifically, 1/(scanning line length) of the scanning line length, or 1
Repeat until you have dots. During this repeated process, the smallest value among the periods of white dots that appear within the -scanning line is the desired period candidate. The candidate here is that when the character spacing is larger than the character size, the result may appear as if there are characters between them even though the white is continuous. It is also possible that there is a lot of white in the - character, and the white in this character happens to appear at the same period. Furthermore, in areas where no characters are printed because they are between lines, black dots do not appear in the scanning lines other than dust.

In this case, the minimum value of the period is one dot, and it can be easily determined that only white is continuous. The first two cases are determined using statistical methods. That is, the obtained cycle candidates are totaled for each scanning line, and the value of the candidate with the highest number of identical cycle values is adopted as the true cycle. The period thus obtained is the sum of the character size and the character spacing. Next, in each scanning line, it is verified whether this periodicity can be preserved even if the maximum number of white dots is continued within the - scanning line while satisfying the previous period. This maximum value becomes the character spacing. The font size is
Just subtract the obtained character spacing from the previous period.

Furthermore, regarding the line spacing, this can be done again in a direction rotated by 90 degrees.

Instead of the above method, a simpler method as shown below may be used. That is, a scanning line consisting of only continuous white dots is searched from the developed read information. However, since dust may be included, it is not accurate to say that only white dots are continuous, but a certain amount of black is allowed as the sensitivity value described above. The maximum number of consecutive scanning lines like this is determined. This continuous value is the line spacing. Also, determine the maximum number of consecutive lines that are not adopted as line spacing. This continuous value is the vertical character size. However, there is a risk of making a mistake if you examine only one line, so after verifying all the scan lines and listing several line spacing candidates, take the statistical values of the line spacing that was suggested as a candidate.
The value that appears most frequently is used as the line spacing. Statistical processing is also applied to the character size. Regarding the character spacing, it is sufficient to perform this process again in a direction rotated by 90 degrees.

The display means 5 receives the character dimensions and character spacing (line spacing) thus obtained from the calculation means 4, and displays them on a CRT display or the like or prints them on paper so that the operator can recognize them.

〔Effect of the invention〕

As explained above, the present invention calculates the vertical and horizontal white lengths of printed matter by operating the read information that has been optically read and has undergone predetermined corrections, and calculates character dimensions and character spacing. Therefore, even if the character size and character spacing of the obtained printed matter are unknown, the character size and character spacing can be determined efficiently, precisely, and with a simple structure without human work.

[Brief explanation of drawings]

FIG. 1 is a block diagram of one embodiment of the present invention. In FIG. 1, 1...optical reading means, 2...
- Expansion means, 3... Correction means, 4... Calculation means, 5
...Display means.

Claims (1)

[Claims] Optical reading means for optically reading and binarizing a printed matter; Expanding means for receiving the read information read by the optical reading means and virtually expanding it to the same dimensions as when reading; and the expanding means. a tilt correction means for detecting and correcting the tilt of the read information developed by the reader, and developing the read information again; and an operation for calculating the vertical and horizontal white lengths of the printed matter by operating the developed read information. 1. A character information detection method comprising: a means for detecting character information; and an output means for outputting a result of calculation by the calculating means.