JPH06111060A - Optical character reader - Google Patents

Abstract

PURPOSE:To read a character recorded on a form using various kinds of colors such as a general document on the market or a slip not assuming processing by an optical character reader, etc. CONSTITUTION:This reader is comprised of a scanning part 10 which detects a color image based on the form on which the character, etc., is recorded, a sheet buffer 20 to store a detected color image, a labelling part 30 which detects and stores a label image consisting of a set of connected elements at each color from among stored color images, a label identification part 40 which identifies whether or not a detected label image of each color is the label image in correspondence with the character, a character string extraction part 50 which extracts a character string based on the arrangement of the label image identified as the one in correspondence with the character, and a character recognition part 60 which performs character recognition on prescribed label images stored in the labelling part 30.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical character reader for reading characters recorded on a sheet of paper having a plurality of colors.

[0002]

2. Description of the Related Art Generally, an optical character reader for reading characters written on a form or the like processes a black and white (binary) image in which black is valid (character data) and white is invalid (background). The target of. Therefore, in the scanning unit that optically scans a form or the like to detect an image, binarization corresponding to white / black is performed based on a certain threshold value.

For this reason, the color used for printing a character entry frame or a character that does not need to be read on a form to be processed by the optical character reading device has a color in the scanning part of the optical character reading device. In the threshold theory), a color determined to be white, that is, a color called a dropout color is used. This dropout color is determined according to the wavelength of the light source used in the scanning section.

Therefore, in the case of a form for an optical character reading device, unnecessary color printing of the background or the like is dropped out, and only the written characters (character data) to be read are detected for character recognition processing. Be served.

By the way, generally, a large number of slips that are commercially available and forms that are not premised on the processing by the optical character reader are used. Therefore, it is required that the characters written on these slips and forms be read by the optical character reading device.

[0006]

However, it is not premised on the processing of commercially available slips and optical character readers (dropout color is not used).
Since forms use various colors, even if they are read by an optical character reading device, the characters to be read may drop out due to the binarization process,
Both the background and the characters are determined to be black. For this reason, it is impossible to read the characters recorded on a form other than the form for the optical character reading device.

The present invention has been made in view of the above points, and is a sheet using various colors such as a general slip that is commercially available and a form that is not premised on the processing by the optical character reading device. It is an object of the present invention to provide an optical character reading device capable of reading characters recorded on a computer.

[0008]

According to the present invention, there is provided scanning means for optically scanning a sheet on which characters and the like are recorded to detect a color image, and storing the color image detected by the scanning means. A storage unit, a label image composed of a set of connected pixels for each color is detected from the color image stored in the storage unit, and a labeling unit for storing the label image for each color; With respect to each of the label images of the respective colors, a label identification means for identifying whether or not the label image corresponds to a character, and a character string is extracted based on an array of label images identified as corresponding to the character by the label identification means. Stored in the labeling means based on the information about the character string extracting means and the character string extracted by the character string extracting means. And configure and a character recognition means for performing character recognition for the constant of the label image.

[0009]

According to such a configuration, a label image (character pattern) corresponding to a character is detected for each color in a color image detected from a paper in which a plurality of colors are mixed, and the arrangement (character string) is detected. Since the characters are extracted, it is possible to recognize each of the characters of a plurality of colors recorded on one sheet.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the arrangement of the optical character reading apparatus according to the embodiment. As shown in FIG. 1, the optical character reader according to the present invention includes a scanning unit 10, a sheet buffer 20, a labeling unit 30, a label identification unit 40, a character string extraction unit 50, and a character recognition unit 60. .

The scanning unit 10 optically scans a sheet or the like to be processed as a color image and outputs a color image signal to the sheet buffer 20. The sheet buffer 20 stores the color image obtained by the optical scanning in the scanning unit 10 for one sheet, for example.
The labeling unit 30 labels the color image stored in the sheet buffer 20 for each color according to the connectivity of pixels and stores the image (hereinafter referred to as a label image). The label identifying unit 40 identifies whether the label image detected by the labeling unit 30 is for a character, for example, based on its size.
At this time, the label identifying unit 40 identifies the label image of each color. The character string extracting unit 50 includes the label identifying unit 4
With respect to the label image identified as a character by 0, the character string is extracted based on the color information and the arrangement.
The character recognition unit 60 performs character recognition on the label image stored in the labeling unit 30 according to the information extracted by the character extraction unit 50. Next, the operation of this embodiment will be described. First, the paper to be processed is scanned by the scanning unit 10 and detected as a color image. The detected color image is stored in the sheet buffer 2
Stored in 0.

The labeling section 30 includes a sheet buffer 20.
When a color image for one sheet is stored in, the connectivity of images is checked for each color in the color image, labeling is performed for each connected set, and the image for each label (label image) is stored for each color. To do.

When the labeling for each color image is performed, the label identifying section 40 identifies whether the label image of each color corresponds to a character. This identification method is, for example, when the size of the labeling image is smaller than a preset threshold value, it is identified as a character, and when it is larger, it is identified as a non-character such as a picture, a photograph, and a background color. .

The character string extraction unit 50 extracts and generates a character string for the label image identified as a character based on its color and position information. For the character string, the character string extraction unit 50 checks the Euclidean distance between the label images, extracts one character area, and also identifies the direction of the character arrangement (vertical writing, horizontal writing, etc.). FIG. 2 is a diagram for explaining the character string extraction.

As shown in FIG. 2, attention is paid to one label image, and the distance between the label image and the label images located around the label image is obtained. As a result, the label image with the shorter distance is arranged in the same character string.

When the character string is extracted by the character string extracting unit 50, the character recognizing unit 60 determines, based on the information about the extracted character string, one character for the label image of the character stored in the labeling unit 30. The recognition process is performed for each.

That is, the character recognizing section 60 sequentially performs character recognition on the label images of the color from which the character string is extracted, in accordance with the arrangement direction of the label images. As a result, the character recognition result is obtained in the order of the arrangement of the characters recorded on the paper, and therefore the arrangement of the characters recorded on the paper to be processed is not particularly limited. Further, for example, even when the red characters are written vertically and the blue characters are written horizontally, the characters of each color can be recognized in the correct order.

In this way, a character string is extracted for each color of the characters recorded on the paper, and the character recognition is performed according to the arrangement, so that a plurality of colors may be mixed on the paper. Also, a correct character recognition result can be obtained for characters of each color.

[0019]

As described above, according to the present invention, a character image (label image) is extracted for each color from a detected color image, and character recognition is performed according to the extracted color image. It is possible to read characters recorded on papers of various colors, such as slips and forms that are not premised on processing by the optical character reading device.

[Brief description 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 diagram for explaining the processing of a character string extraction unit 50 in this embodiment.

[Explanation of symbols]

10 ... Scan unit, 20 ... Sheet buffer, 30 ... Labeling unit, 40 ... Label identifying unit, 50 ... Character string extracting unit, 60
… Character recognition part.

Claims (1)

[Claims] 1. A scanning unit for optically scanning a sheet on which characters and the like are recorded to detect a color image, a storage unit for storing the color image detected by the scanning unit, and a storage unit for storing the color image. From the stored color image, for each color, while detecting the label image consisting of a set of connected pixels, labeling means for storing the label image for each color, for each color label image detected by the labeling means, A label identifying means for identifying whether or not the label image corresponds to a character, a character string extracting means for extracting a character string based on an array of label images identified as corresponding to the character by the label identifying means, Based on the information regarding the character string extracted by the character string extracting means, the predetermined label image stored in the labeling means is Optical character reader, wherein the character recognition means for performing character recognition, by comprising a.