JPH0362282A - Optical character reading device - Google Patents

Abstract

PURPOSE:To improve the read processing efficiency with a business form including marks and characters by performing the optical scan via a character image detecting image sensor with density higher than a mark detecting image sensor after setting a fixed carrying speed of the form. CONSTITUTION:The viewfield of a character image detecting image sensor 11 is set at 1/2 compared with a mark detecting image sensor 12 with the same 1-line scanning frequency, e.g., k5MHz set between both sensors 11 and 12. Thus the images can be detected for identification of both marks and characters with the carrying speed of a business form kept at a fixed level. In other words, it is not required to reduce the form carrying speed for reading the character images. As a result, the form read processing time can be shortened as the occupation rate of characters is increased in a business form where marks and characters are mixed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to an optical character reading device that reads marks and characters from a form to be subjected to recognition processing.

(Conventional technology) Generally marked using an optical character reader.

When performing a reading process on a form containing a mixture of characters, etc., as shown in the block diagram of FIG. ing. The image detected by the image sensor l is amplified by an amplifier 3.

Thereafter, the character images among the detected images are transferred to the character recognition section 4 and subjected to character recognition processing. Also,
Among the detected images, the image of the mark is transferred to the mark recognition processing section 5 and subjected to mark recognition processing to determine whether the mark is present or not.

Normally, mark image recognition processing only involves determining whether a mark is written or not, so it can be performed even with relatively coarse scanning. On the other hand, character image recognition processing requires high-density and accurate scanning because it recognizes various characters.

For this reason, when scanning a form using the image sensor 1, the transport speed of the form is lowered for areas where characters are written than when scanning marked areas, thereby performing high-density scanning. There is. That is,
The document conveyance speed is controlled as shown in FIG.

Note that if the document conveyance speed when reading characters that requires high-density scanning is v11, and the document conveyance speed when reading marks that can be recognized even in coarse scanning is v2, then V
The relationship 2Vl-V2 holds between l and v2.

(Problem to be solved by the invention [1]) In this way, when marks and characters are mixed, as in book M2 shown in the figure, the document conveyance speed prioritizes character reading that requires high-density scanning. This results in a delay of Vt.

For this reason, when a form is filled out with a mixture of marks and characters,
When a large amount of reading is performed, the conveyance speed of the portion where characters are read becomes slow, so the processing speed becomes much slower than when only marks are read.

The present invention has been made in view of the above-mentioned points, and provides an optical character (means for solving the problem) that can efficiently perform reading processing on a form containing a mixture of marks and characters. is the first method for reading the marks written on the form.
and a second image sensor that scans at a higher density than the first image sensor for reading characters, and a second image sensor that scans characters at a higher density than the first image sensor. The document conveying speed is configured to be constant by performing recognition processing and performing character recognition processing based on the image of the character detected by the second image sensor.

(Function) In the optical character reading device configured in this way, image sensors are installed in the mark reading and character reading groups, and the character reading is performed at a constant form conveyance speed matched to the mark reading. A reading image sensor performs high-density scanning to detect character images used in character recognition processing.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an optical character reading device according to the same embodiment. In the same figure, the image sensor l
l is for detecting character images, and I
It is assumed that the user has a field of view of K bits (128 bits). The image sensor 12 detects the image of the mark, and has a field of view of 2 (256wn) corresponding to the size of the form to be processed. Each of the image sensors 11.12 is connected to an amplifier 13.14 for amplifying and quantizing the signal detected by the sensor 11.12.

Further, the amplifier 13 is connected to a character recognition section 15 that performs character recognition processing based on the quantized signal quantized by the amplifier 13. Each F1 reading control information used when performing character reading processing is set in the character recognition unit 15 in advance before performing the document reading processing.

Further, the amplifier 14 is connected to a mark recognition processing section 1B that performs mark recognition processing based on the quantized signal quantized by the amplifier 14.

In this embodiment, it is assumed that reading processing is performed on the form 20 shown in the figure. The form 20 is filled with a mixture of characters and marks.

Further, it is assumed that the character entry position is set corresponding to the position to be read by the image sensor 11. Here, it is assumed that an area for writing characters is provided in the left half of the first and second lines of the form 20.

Next, the operation of this embodiment will be explained.

First, in a form conveyance mechanism (not shown), a form to be processed is conveyed on a form conveyance surface at a constant conveyance speed. Lines to be read by the image sensors 11 and 12 are set on the form conveyance surface on which the form is conveyed.

That is, the image sensor 11 targets the line through which the left half of the form passes on the form conveyance surface, and the image sensor 12 targets the line through which the entire document conveyed on the form conveyance surface passes. .

In the same embodiment, an image sensor 11 for reading characters
has a field of view of IK bits, and the one line scanning frequency is 5MlI2. The image sensor 12 for reading marks has a field of view of 2 bits, and the 1 line scanning frequency is the same as that of the image sensor 11.
It shall be MIIZ.

FIG. 2 shows a timing chart of scanning of each image sensor 11, 12. As shown in FIG. 2, if the field of view of the image sensor 11 is IK bits, the field of view of the image sensor 12 is 2 bits, and the scanning frequency is the same (5M 11z), then the image sensor 1
2, the image sensor 11 performs two scans during one scan (assumed to be 1.0 ms). Therefore, the third
As shown in the figure, when the form is conveyed at a constant speed (v2), the image sensor U can detect an image with twice the density as the image sensor 12.

That is, while the image sensor 11 scans one line (1,0m5), the document conveyance speed v2 is 0.25a.
If the form is moved (transported) by m, the image sensor 11 will move the form by 0.125 mm (0.
, 5ms). Therefore, when reading a form with a mixture of marks and characters, the mark part is read at the form transport speed v2, and the text part is read by lowering the form transport speed to Vl. Scanning is performed by each image sensor 11.12 with the same density as in the case of scanning.

Image sensor! The signal detected by l is amplified by amplifier 13 and further quantized. The signal quantized by the amplifier 13 is output to the character 2 recognition section 15.

Character recognition section 15 stores the quantized signal from amplifier 13 in memory. The character recognition unit 15 cuts out a character image for each character from the form image stored in the memory, and performs character recognition processing on this character image. Note that the form image detected by the image sensor 11 is an image of an area passing through the line to be read by the image sensor 11, and therefore includes not only an image of characters but also an image of a mark. The character recognition unit 15 cuts out a character image from a form image including a mark image based on information indicating a character writing position included in various preset reading control information. The character image to be processed here is sufficient to perform the same character recognition process as the image obtained by high-density scanning, which has conventionally been performed by reducing the transport speed of the form. Therefore, character recognition processing can be performed and recognition results can be obtained in the same manner as in the past.

Further, the signal detected by the image sensor 12 is transmitted to the amplifier 1
4 and further quantized.

The signal quantized by the amplification 114 is output to the mark recognition processing section 1B. The mark recognition processing unit 16 quantizes the signal from the amplifier 14 and stores the quantized signal in memory. The mark recognition processing unit 1B detects whether a mark is written at a predetermined position on the form (H/None) from the form image stored in the memory.

Since the image used in this mark recognition process is the same as that used in the conventional method, the same recognition result as in the conventional method can be obtained by the mark recognition process.

In this way, the image sensor 11 for character image detection
By setting the field of view to half that of the image sensor 12 for mark detection, and setting the 1-line scanning frequency of each image sensor 11.12 to the same (for example, 5 MIIZ), the conveyance speed of the form is kept constant. It is possible to detect images used for recognition processing of marks and characters. That is, there is no need to reduce the document conveyance speed in order to read character images. Follow;
mark.

The higher the proportion of characters in a mixed form, the more difficult it is to read the form compared to the conventional form reading method.
Processing time can be greatly reduced. Further, by keeping the conveyance speed of the form constant, it becomes possible to easily control the conveyance amount in the form conveyance mechanism section.

In the above embodiment, the field of view of the image sensor 11 for detecting character images is set to be half that of the image sensor 12 for detecting marks, and by making the one-line scanning frequency of each image sensor 11, 12 equal, the field of view is kept constant. The image sensor 11 performs high-density scanning at a form conveyance speed of effect can be obtained. For example, if the field of view of the image sensor 11 is two-thirds that of the image sensor 12, by setting the ratio of the one-line scanning frequency to 4:3 (image sensor 1: image sensor 12), the conveyance speed of the form is constant. If so, while the image sensor 12 scans one line,
The image sensor 11 can perform two-line scanning.

Further, in the above embodiment, the character entry area of the form is set at the left half position, but it can be set arbitrarily depending on the field of view, setting position, etc. of the image sensor 11 for character image detection. It is.

[Effects of the Invention] As described above, according to the present invention, an image sensor for character image detection and an image sensor for mark detection are provided,
By keeping the form conveyance speed constant, the optical scanning by the image sensor for character image detection is performed at a higher density than by the image sensor for mark detection. It is possible to efficiently read documents containing a mixture of marks and characters.

[Brief explanation of drawings]

Fig. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention, Fig. 2 is a timing chart of scanning of each image sensor in the embodiment, and Fig. 3 is a form chart in the embodiment. FIG. 4 is a diagram showing the configuration of a conventional optical character reading device, and FIG. 5 is a diagram showing the document transportation speed in the optical character reading device shown in FIG. 4. 11... Image sensor (second image sensor),
12..., image sensor (first image sensor), 13, 14... amplifier, 15... character recognition section (character recognition processing means), lt... mark recognition processing section (mark recognition processing means)

Claims (2)

[Claims] (1) Detect the form image by optically scanning the form that is the target of recognition processing on which a mixture of marks, characters, etc. have been written, while transporting it, and based on this form image, the marks, characters, etc. An optical character reading device for reading a document, comprising: a first image sensor for reading marks written on the form; and a higher-density image sensor than the first image sensor for reading characters written on the form. a second image sensor that performs scanning; a mark recognition processing means that performs mark recognition processing based on the image of the mark detected by the first image sensor; and a mark recognition processing means that performs mark recognition processing based on the image of the mark detected by the first image sensor; An optical character reading device comprising: character recognition processing means for performing character recognition processing based on an image of the form, and optically scanning the form at a constant conveyance speed. (2) The number of bits per line scanning of the second image sensor is smaller than the number of bits scanning one line of the first image sensor, and the one line scanning frequency of each of the first and second image sensors is set. 2. The optical character reading device according to claim 1, wherein the second image sensor scans at a higher density than the first image sensor by setting according to the conveyance speed of the form.