JPH03134784A - Optical reader - Google Patents

Abstract

PURPOSE:To reduce the man-hour of input in a data processing system by optically reading a business form and converting its picture data to format information in a code form to input it to a higher-order processor. CONSTITUTION:The data processing system is provided with an optical reader 1, a business form 2, a higher-order processor 3, a display device 4, and an interface cable 30 between the higher-order processor 3 and the optical reader 1. A form 2 is sucked to the optical reader 1 by the read command from the higher-order processor 3, and picture data obtained from the form 2 is recognized by the reader 1 and is converted to format information. Converted format information in a code form is sent to the higher-order processor 3 through the interface cable 30. Thus, much man-hour is not required for input of the data processing system.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an input device in a data processing system, and particularly to an input device for optically reading a form (slip, paper) and converting the image data into format information in code format. The present invention relates to an optical reading device as an input device that can be converted and input to a host processing device.

[Prior Art] Conventional optical reading devices of this type read and code characters, special symbols, marks, etc. written at predetermined positions on paper, or input image data as is. there were. These include so-called OCR, MCR, barcode readers, and image scanners.

Further, as a graphic input device that facilitates the input of handwritten line drawings such as ruled lines, there is, for example, Japanese Patent Laid-Open No. 138988/1988. The configuration of this graphic input device recognizes handwritten line drawings by providing non-fixed knowledge regarding ruled lines and the like drawn freehand.

[Problems to be Solved by the Invention] In order to create and output a form using a data processing system, it is necessary to input format information such as formats and characters to the data processing system as coded data.

In the above conventional technology, characters, symbols to be read,
Marks and the like are set at predetermined positions, and their format information is given from the outside by some means, and the image data of the target area is read according to that information.

There are problems in that generating this format information requires manual labor and a large amount of man-hours, and that a dedicated utility program is required.

An object of the present invention is to provide an optical reading device that recognizes format information and inputs it to a host processing device.

[Means for Solving the Problems] In order to achieve the above object, the optical reading device of the present invention has the following features:
A means for reading image data on a form, a recognition processing means for recognizing formats and characters from the read image data and converting it into code-format data, and transmitting the code-format data as format information to a host processing device. have the means.

[Operation] The format and characters drawn on the form are read as image data, and the recognition processing means recognizes the format and characters from this image data, converts it into code format data, and transfers it to the host processing device as format information. Therefore, the format information of the form, including the format and characters, can be input to the higher-level processing device using the optical reading device as the input device.

[Example] An example of the present invention will be described below in detail with reference to the drawings.

FIG. 1 shows an example of a system configuration using a reading device according to the present invention, in which 1 is an optical reading device according to the present invention, 2 is a form, 3 is a host processing device represented by a workstation or office processor, etc. 4 is a display device, and 30 is an interface cable between the host device 3 and the optical reading device 1.

In response to a reading command from the higher-level processing device 3, the form 2 is sucked into the optical reading device 1, and the device 1ii1 reads the form 2.
Recognizes the image data obtained from and converts it into format information. The converted code format information is sent to the higher-level processing device 3 via the interface cable 3°.

FIG. 2 shows a configuration block diagram of FIG. 1, in which 5 is an interface adapter, 6 is an interface control section in the optical reading device 1, 7 is an image processing section, 8 is an image reading section,
Reference numeral 9 indicates a reading mechanism section, and 10 indicates a file on the upper processing device side.

FIG. 3 shows a functional block diagram of the optical reading device 1. As shown in FIG. The image processing operation will be explained below using FIG.

A form reading command from the host processing device 1 is transmitted to the interface control unit 6 via the interface cable 3o. When the interface control section 6 receives the command, it outputs an interrupt for receiving the command to the processing control section 11 via the signal group 101.

Upon receiving this interrupt, the processing control unit 11
A document reading instruction is output to the reading control unit 17 via 02. As a result, the reading control section 17 moves the reading mechanism section 9 to start sucking in the form 2.

An optical signal 103 input from the form 2 at the same time as the start of the suction operation of the form 2 is converted into an electrical signal 104 by the photoelectric conversion section 19 . The electrical signal 104 is an analog signal having a level corresponding to the shading on the form 2, and the binarization circuit 18 converts the white level into a logic "0" and the black level into a logic u 1 u bit string signal 105 at a predetermined threshold value. Output as . This bit string signal 105 is converted into serial-to-parallel data by the reading unit 17 and then stored in the image memory 20 in the image memory unit 12 via the signal line 102.

Here, as shown in form 2' in FIG. In contrast, a command to rotate the image by 90° is output.

In response to the same command, the image processing mechanism 13 reads out the image data stored in the image memory 20, performs a 90° rotation process, and then stores it again in the image memory 20.

Next, the processing control section 11 outputs a recognition processing command to the recognition processing control section 14. The recognition processing control unit 14 includes a sequence control unit 15 that controls a processing sequence, a recognition program storage area 25 that stores a recognition algorithm, and a high-speed calculation processing unit that performs bit processing of image data, dictionary matching processing, calculation processing, etc. at high speed. It consists of a section 24.

Upon receiving the recognition processing command, the recognition processing unit 14 sequentially reads out the image data of the form stored in the image memory 20 and performs (1) noise removal from the image data (2) image tilt detection and correction thereof according to the recognition program. (3) Extract line segments, arcs, and characters 4) Extract feature quantities (5) Perform processing such as comparison with the standard time trace amount table stored in the recognition dictionary 21 to recognize ruled lines, fields, characters, etc. The result is stored in the image memory section 1 as format information.
It is stored in the data area 23 in 2.

Hereinafter, specific examples of recognition processing for ruled line frames, fields, line segments, characters, etc. will be explained using FIGS. 4, 5, and 6. FIG. 4 is a flowchart of recognition processing, FIG. 5 is a diagram explaining the operation in FIG. 4, and FIG. 6 is a diagram explaining conditions for determining frames, line segments, and characters.

In FIG. 4, in step 201, image data to be processed is transferred from the image memory 20 to the recognition processing section 14. Next, this image data is projected in the vertical and horizontal directions, and black connected contour extraction processing is performed to obtain the outermost contour of the black mass (FIG. 4 202, FIG. 5 (A)). For the outermost contour obtained here, it is determined whether it is a table based on the contour size (
Figure 4 203). The conditions for determining whether it is a table are shown in Figure 6 (A
), where the width of the outline is a and the height is b, a≧Limit value A and b≧Limit value B. If it is determined to be a table, the image within the area determined to be a table is subjected to projection processing, a white contour is obtained, and a white connected contour extraction process is performed (204 in Figure 4, (B) in Figure 5).
. Next, the black outline is determined again for the image within the white area determined in step 204 (205 in Figure 4, 205 in Figure 5 (C
)). The black outline obtained at this point can be determined to be a printed character or a line segment within the ruled line frame. In order to determine either of these, it is determined whether it is a line segment or a character in steps 206 and 210 in FIG. 21
1).

Here, the line segment judgment condition and the character judgment condition are set to the sixth
Shown in Figures (B), (C) and (D). That is, line segments or characters are determined based on the black connected outline size and the ratio of vertical length to horizontal length. FIG. 6(B) shows a horizontal solid line, where the length (horizontal direction) is X, the line width (vertical direction) is y, and the conditions for determining the horizontal solid line are: X≧Limit value X and Limit value Y1≦y≦ Limit value Y2 and x/y≧limit value Z. Figure 6 (C) shows the vertical solid line, and the length (vertical direction)
is y, and the line width (horizontal direction) is X, and the conditions for determining a vertical solid line are: limit value X1≦X≦limit value X2, y≧limit value Y, and y/x≧limit value Z. FIG. 6(D) shows a character, where the horizontal direction is X and the vertical direction is y, and the conditional determination of the character is as follows: limit value x1≦X≦limit value X2 and limit value Y1≦y≦limit value Y2.

Finally, the existence of black line segments is determined in four directions around the white area obtained in step 205 in FIG. 4, and one peripheral line segment extraction process is performed (208 in FIG. 4, (D) in FIG. 5). ). Based on the results, a ruled line frame recognition process is performed (209).

On the other hand, if it is determined in step 203 of FIG. 4 that the table is not a table, line segment determination (212) or character determination (214) is performed.
According to the results, line segment recognition (213) and character recognition 1 (2
15) is performed.

By repeating the above process, ruled line frames, fields, line segments, and characters can be recognized.

The format of the encoded data is a code that represents a character, line segment, or frame; characters are positional information based on the number of dots from a reference point (for example, the top left of a form) and the character code; and line segments are positional information and character code of the start and end points. The thickness of the frame (represented by the number of dots) and the frame can be expressed by the position information of the corner (intersection).

Next, the processing control section 11 performs an operation of comparing the characters among the format information outputted by the recognition processing section 14 with the word dictionary 22 stored in the image memory section 12, and attempts to correct unreading and misreading. A plurality of words used in forms are stored in the word dictionary 22, and recognition accuracy can be improved by comparing the words. In other words, J1%! The resulting character string is compared as a word with words in a word dictionary, and if there is a matching word, it is recognized as correct. If not, select a similar word, or display the word so that the operator can see that it is not available, and allow the operator to input.

When the above operations are completed, the processing control unit 11 instructs the interface control unit 6 to transfer the format information stored in the data area 23, and the interface control unit 6
from data area 2 via interface cable 3o.
The format information in 3 is transferred to the higher-level processing device 3.

Furthermore, when the higher-level processing device 3 outputs a command to download the recognition program, the processing control unit 11 outputs a command to the interface control unit 6 to receive data corresponding to the content of the recognition program from the higher-level processing device 3. , the data received by the interface control unit 6 is transferred to the data area 2.
Store in 3.

After all data has been stored, the processing control section 11 sequentially reads out the received data stored in the data area 23 and sequentially writes them into the recognition program storage area 25 in the recognition processing section 14. This makes it possible to use the so-called optical character reading function (OCR) by simply changing the recognition program content.
) and a format information recognition function.

According to this embodiment, the following effects can be obtained.

(1) The format definition information of forms and vouchers, which is the most complicated in data processing systems, can be obtained automatically and quickly by simply reading the forms and vouchers.
This will significantly reduce the number of inputs into the data processing system.

(2) The same reading device can function as both a conventional optical character reading device and a format recognition device.

(3) There is no need for a format information creation utility, which was essential in conventional optical character reading devices.

[Effects of the Invention] According to the present invention, formats and characters on a form can be recognized by an optical reader serving as an input device of a data processing system, and inputted as encoded format information to a host processing device.

[Brief explanation of the drawing]

FIG. 1 is a system configuration diagram for explaining one embodiment of the present invention, FIG. 2 is a block diagram of the configuration of FIG. 1, FIG. 3 is a functional block diagram showing one embodiment of the present invention, and FIG. 4 5 is a flowchart of the recognition process, FIG. 5 is a diagram for explaining the operation in FIG. 4, and FIG. 6 is a diagram for explaining the conditions for determining a frame and a line segment of five characters. DESCRIPTION OF SYMBOLS 1... Optical character reading device, 2... Form, 3... Upper processing device, 11... Processing control section, 12... Image memory section, 13... Image processing mechanism, 14... - Recognition processing unit, 17... Reading control unit, 18... Binarization circuit, 1
9... Photoelectric conversion unit, 21... Recognition dictionary, 25...
Recognition program storage area, (A (8) (D)

Claims (1)

[Claims] 1. An optical reading device connected to a host processing device, comprising:
means for reading image data on a form; recognition processing means for recognizing formats and characters from the read image data and converting it into code-format data; and transmitting the code-format data as format information to the above-mentioned host processing device. An optical reading device characterized by having a means for reading. 2. An optical reading device connected to a host processing device,
A means for reading image data on a form, an image memory for storing the read image data, a recognition dictionary storing formats and character feature values, and a recognition program, and according to the recognition program, the recognition dictionary is read. Recognition processing means for referencing and recognizing formats and characters from the image data in the image memory and converting it into code format data, and means for transmitting the code format data as format information to the higher-level processing device. An optical reading device featuring: 3. The optical reading device according to claim 2, wherein the recognition program downloaded from the host processing device is stored in the recognition processing means. 4. The optical system according to claim 1, 2 or 3, further comprising a word dictionary in which a plurality of words are registered, and the character recognized by the recognition processing means is compared with the words in the word dictionary. reading device.