JPH01144181A - Optical character reader - Google Patents

Abstract

PURPOSE:To simplify a slip preparing process by using the slip having a character frame recorded with a non-dropout color such as a black color as a reading object. CONSTITUTION:A character frame detecting part 22 prepares projected data (histogram) corresponding to the character frame from picture pattern data stored into a sheet buffer 1, and the position of the character frame is detected based on the projected data. A character detection segmenting part 23 decides a character segmenting position based on the position of the character frame detected by the character frame detecting part 22, and a processing to segment the picture pattern data at every character from a sheet buffer 21 is carried out. Since the slip having the character frame recorded with the non-dropout color such as the black color can be used as the reading object, the process to print the character frame with the non-dropout color in the preparing process of the slip can be omitted. Thus, the slip preparing process can be simplified.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Industrial Application) The present invention relates to an optical character reading device that performs character recognition processing for each character recorded on a form.

(Prior Art) Conventionally, in an optical character reading device, a document to be read is optically scanned, and image pattern data obtained by photoelectric conversion processing, for example, for one sheet (one document) is stored in a sheet buffer. is stored in The sheet buffer is configured so that detection and cutting processing is performed for each character, and character recognition processing is performed for each character.

In the detection cutting process, as shown in FIG. 7, a predetermined line area is set within the sheet buffer, and projection data (histogram) 10 corresponding to each character pattern P within this line area is created. The line area is set in advance based on line position information M of format control information given from, for example, a host computer at the start of a reading operation. Roughly based on the created projection data 10, a cutting position 11 that can be cut out for each character is detected. This cutting position 11 is located at a character frame 13 printed on the form in advance.
This position is approximately equivalent to the position of . Then, depending on the detected cutout position 11, the character pattern data for each character is determined from the sheet buffer? The character is extracted from the character HH processing section and transferred to the character HH processing section.

By the way, the character frame 13 as described above is printed on the form in advance and is set in order to set the recording position range of each character to be read. However, if the image pattern data of the character frame 13 is stored in the sheet buffer together with the character pattern, the data becomes inconvenient in the character detection and cutting process as described above.

For this reason, the character frame 13 is printed in a so-called dropout color so as to be unreadable.

However, printing the character frame 13 in a dropout color poses a major constraint when creating a form, and has the drawback of complicating the process of creating a form and increasing the cost of the form. Further, for example, in a copy of a form, the character frame 13 is black, so it cannot be used as a reading target.

For this reason, there is a drawback that the types of forms that can be used as objects for reading are limited.

(Problems to be Solved by the Invention) In conventional devices, character frames provided in advance on a form must be printed in dropout color due to character detection and cutting processing. For this reason, there are problems in that the process of creating a form is complicated, the cost of the form is increased, and the types of forms that can be used as objects for reading are limited.

An object of the present invention is to simplify the process of creating a form and expand the types of forms that can be read by making it possible to use forms whose character frames are recorded in a non-dropout color such as black as objects to be read. The object of the present invention is to provide an optical character reading device that can read characters.

[Structure of the Invention] (Means and Effects for Solving Problems) The present invention creates a character frame and each image pattern data corresponding to the characters recorded in the character frame, and creates an image corresponding to the character frame. This is an optical character reading device that uses pattern data to detect the position of a character frame as a character cutting position. For this reason, this device includes a character frame detection means that creates OA shadow data corresponding to a character frame from image pattern data, and detects a character frame for one character based on this projection data; The apparatus is provided with a detection cutting means for detecting and cutting out image pattern data for each character as a detection cutting position.

With this configuration, the character frame printed in non-dropout color is actively used as the character cutting position, so the printing process using dropout color is omitted, simplifying the form creation process and copying. It is possible to make it possible to use forms such as documents as objects to be read.

(Example) The present invention will be described in detail below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of the optical character reading device of the same embodiment. As shown in FIG. 1, this device includes a scanning section 20,
It includes a sheet buffer 21, a character frame detection section 22, a character detection cutting section 23, and a recognition section 24. The scanning unit 20 optically scans the form to be read, and converts it photoelectrically into 15
This is a circuit that outputs image pattern data. The sheet buffer 21 is a buffer 7 memory that stores image pattern data for one sheet output from the scanning section 20.

The character frame detection unit 22 is a circuit that creates tJJ shadow data (histogram) corresponding to a character frame from the drawing pattern data stored in the sheet buffer 21, and detects the position of the character frame based on this projection data.

The character detection and cutting section 23 determines a character cutting position based on the position of the character frame detected by the character frame detection section 22,
This circuit performs a process of cutting out image pattern data for each character from the sheet buffer 21.

The recognition unit 24 recognizes the 1 extracted by the character detection and extraction unit 23.
This is a circuit that processes character OFF for image pattern data for each character.

Next, the operation of this embodiment will be explained. First, when a document to be read is scanned by the scanning unit 20, one sheet of image pattern data corresponding to characters recorded on the document is stored in the sheet buffer 21. Here, as shown in FIG. 2, the form to be read includes a character frame 30 printed in a non-dropout color such as black, and the character frame 30 printed in a non-dropout color such as black.
Characters 31 written (or printed) character by character in 0 are recorded. The character frame 30 is composed of a matrix of ruled lines, and is a frame in which a group of characters is divided into characters by vertical and horizontal ruled lines.

The sheet buffer 21 stores image pattern data corresponding to the recorded contents of one sheet of forms as shown in FIG. Is the character frame detection unit 22 a sheet buffer? A process of detecting a character frame corresponding to one character from the image pattern data stored in 21 is performed. The character frame detection unit 22 is
As shown in FIG. 4, projection data 40a corresponding to ruled lines 30a and 30b corresponding to the top and bottom of the character frame, for example.
, 40b are created, and each projection data 40a, 4 is created.
Each ruled line 30a based on 0b.

Edges A and B on the character side (inside) of 30b are detected.

When creating each of the projection data 40a and 40b, the sheet buffer 21 is scanned based on the line position information of format control information given in advance from a host computer etc. at the start of reading, and the ruled line 30a is created.
, 30b will be detected. Furthermore, Edge A
, B, as shown in FIG. 6, the maximum value L1 of the created projection data 40a (40b) is obtained, and the maximum value L1 is multiplied by a predetermined coefficient (for example, 0.9) to obtain L2. is required.

Then, points a and b that cross the projection data L2 are found, and one of these points a and b inside the character frame (on the character side) is selected as an edge.

Similarly, as shown in FIG. 5, projection data 4 corresponding to the ruled lines 30c and 30d corresponding to the left and right sides of the character frame are
0c and 40d, and each projection data 40C9
40d, each ruled line 30c, character side of 30d (
Detect edge C9D (inside). As a result, as shown in FIG. The character frame position data for the minutes is created and output to the character detection and cutting section 23. The character detection cutting unit 23 uses the character frame position data corresponding to the character frame range created by the character frame detection unit 22 as character cutting position data, and uses the character frame position data corresponding to the character frame range created by the character frame detection unit 22 to
Image pattern data for one character is cut out from the recognition unit 2.
Output to 4.

That is, each edge A to D shown in FIG. 3 becomes a cutting position, and image pattern data for one character within that frame is cut out.

The recognition unit 24 executes character recognition processing for each character on the image pattern data extracted by the character detection and extraction unit 23, and outputs the recognition result to, for example, a host computer.

In this way, by using the character frame printed in advance in a non-dropout color such as black on the form to be read as the cutting position for each character, the characters recorded on the form can be recognized character by character. can be processed. Therefore, since there is no need to print the character frame of a form in dropout color, it is possible to simplify the process of creating a form and reduce costs. Furthermore, when a copy of a form with a character frame printed in a non-dropout color is copied using a copying machine, the character frame of the copied form will be black, so it is possible to read the copied form.

Furthermore, in the conventional method, since the space between characters (that is, the white data portion) is used for the cutting position of characters, it is necessary to provide a certain amount of space between characters. In contrast, in the method of the present invention, the ruled lines of the character frame serve as the cutting positions, so there is no need for spaces between characters as in the conventional method, and each character can be recorded close to each other. It is. Therefore, it is possible to increase the recording density of characters on a form compared to the conventional method, and it is possible to improve the distance between recorded information on a single form.

In the above embodiment, the sheet buffer 21
Although the case where one sheet's worth of image pattern data is stored has been described, it goes without saying that the present invention can also be applied to a case where one line's worth of image pattern data is stored using a line buffer.

[Effects of the Invention] As detailed above, according to the present invention, a form in which the character frame is recorded in a non-dropout color such as black can be used as a reading target. Therefore, the step of printing a character frame in a non-dropout color can be omitted in the form creation process, thereby simplifying the creation process. In addition, it is possible to read documents with black character frames, such as copies of conventional documents, so it is possible to expand the types of documents that can be read. Furthermore, since the character frame is used as a cutting position, the space between characters when recording characters on a form can be significantly saved.

As a result, it is possible to increase the recording density of characters for a form with the same space as in the past, and it is possible to improve the recorded information ω.

[Brief Description of the Drawings] Fig. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention, and Fig. 2 is a conceptual diagram of image pattern data for explaining the operation of the embodiment. , FIGS. 3 to 6 are conceptual diagrams for explaining the operation of the same embodiment, and FIG. 7 is a conceptual diagram for explaining the conventional system. 20... Scanning section, 21... Sheet buffer, 22.
...Character frame detection section, 23...Character detection cutting section, 24
...Recognition part. Applicant's Representative Patent Attorney Takehiko Suzue Figure 4 Figure 6 Figure 5 Figure 7

Claims (1)

[Claims] A scanning means for outputting image pattern data obtained by scanning and photoelectrically converting a document to be read in which character frames are divided into character frames and characters are recorded within the character frames, and output from the scanning means. a buffer memory means for storing image pattern data corresponding to the character frame and character to be displayed; and a character frame for one character is detected based on projection data of the character frame created from the image pattern data stored in the buffer memory means. a character frame detecting means for detecting, a detecting and cutting means for detecting and cutting out image pattern data for each character from the buffer memory means with the character frame detected by the character frame detecting means as a character detection and cutting position; 1. An optical character reading device comprising: recognition means for performing character recognition processing for each character detected and cut out.