JPH06103412A - Optical character reader - Google Patents

Abstract

PURPOSE:To provide an optical character reader capable of preventing the deterioration of a character recognition rate even if floating is caused in the threshold determining area of a business form. CONSTITUTION:The threshold determining area Pa of definite width L where no character, etc., is recorded is provided on the tip end of the business form P to be the processed object of this optical character reader 10, and a picture area where the character, etc., is recorded is provided while being joined to the threshold determining area Pa. When a bend is caused in the threshold determining area Pa of the business form P, the light quantity of reflected light to be inputted to a photoelectric converter 111 is changed compared with the case that no bend is caused, and an output value is changed according to it, as well and difference from a stored value is generated. When the difference becomes a prescribed value or above, a threshold determining part 14 replaces the output value with the stored value, and determines a threshold on the basis of the replaced value. Thus, even if the floating is caused in the threshold determining area Pa of the business form P, the deterioration of the character recognition rate is prevented.

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, and more particularly, to improvement of binarization processing.

[0002]

2. Description of the Related Art An optical character reading apparatus reads an image having a relatively simple density distribution such as a character or a ruled line recorded on a form, so that the character is read by performing a binarization process. I have to. In other words, the scanner unit irradiates the form with light from a light source, and the reflected light is converted into a photoelectric conversion element (C
(CD) to obtain a form image (multi-valued image) by converting it into an electric signal according to the light amount, obtain a binary image from the form image by binarization, and cut out characters from the binary image. Character recognition processing is performed to read the character.

By the way, threshold value processing is generally used for the binarization processing. However, the white level, which is the level of the white background of the form, changes depending on the paper quality and color of the form, so the threshold value is set in advance. It is difficult to decide.

Therefore, a threshold value determination area P having a constant width L, which is a white level with no characters or the like recorded on the leading edge of the form, is formed.
a (see FIG. 8 described later) is provided, and the threshold value is determined based on the reflected light actually obtained from the threshold value determination area Pa.

Conventional examples of the threshold value determination are shown in FIGS.
Will be described with reference to. FIG. 7 is a side view of the main part of the scanner unit,
FIG. 8 is a plan view of the main part, and FIG. 9 is an output characteristic diagram of the photoelectric conversion element.

As shown in FIGS. 7 and 8, when the form P is conveyed by a conveying mechanism (not shown) and reaches the reading position R, the light source 1 moves from the light source 1 in the threshold value determining area Pa provided at the leading end of the form P. Light is reflected, and the reflected light is incident on a plurality of photoelectric conversion elements arranged in an array via a lens system. The photoelectric conversion element converts the reflected light into an electric signal corresponding to the amount of light and outputs the electric signal to the threshold value determination unit.

The threshold value determining section extracts the maximum value among the output values corresponding to the threshold value determining area Pa of each photoelectric conversion element. Then, the output curve 2 as shown by the solid line in FIG. 9 is obtained. In the figure, the vertical axis represents the maximum output value of each element (CCD), and the horizontal axis represents the element direction (channel direction). As shown in the figure, the distortion (shading) in which the output characteristics decrease on both sides is mainly due to the light emitting characteristics of the light source 1 and the condensing characteristics of the lens system.

Then, the threshold value determining unit determines, for example, 1/2 of each value of the output curve 2 as a threshold value. The determined threshold value 3 is shown by a broken line in FIG.

[0009]

However, as shown in FIG.
(Imaginary line), as shown in FIG. 8 (hatched area),
If the threshold determination area Pa has a lift P'due to bending or warping, that portion approaches the light source 1 side, so the amount of light reflected on the surface of the form P increases, and the maximum output value is imagined in FIG. As shown by the line, a curved curve 2'is formed.
As a result, the threshold 3'determined based on the curve 2'also jumps up as shown by the imaginary line in the figure.

Therefore, in such a case, the output value of the photoelectric conversion element may be smaller than the threshold value 3'when the portion other than the threshold value determining area Pa of the form P is not raised. However, there is a problem in that the portion that should originally become white (1) by the binarization process becomes black (0), and the character recognition rate decreases.

Therefore, the present invention has been made in view of the above circumstances, and provides an optical character reading device capable of preventing a decrease in the character recognition rate even when the threshold value determination area of a document is raised. The purpose is to do.

[0012]

In order to achieve the above object, the optical character reading device of the present invention is provided with a threshold value determining area in which no characters are recorded and an image area in which characters are recorded. In an optical character reading device for reading characters recorded in the image area by optical scanning of the formed form, a light source for irradiating the form with light and photoelectric conversion for converting reflected light from the form into an electric signal. An element, a storage means for storing an output value of the photoelectric conversion element in a state where no bending occurs in the threshold value determining area, and an output value of the photoelectric conversion element based on reflected light from the image area and the storage means When the difference between the output value and the stored value is equal to or more than a predetermined value, the output value is replaced with the stored value and the threshold value is determined based on the replaced value. It is characterized in that it has a threshold determination section.

[0013]

According to the optical character reader having the above structure, when the threshold value determining region of the form is curved, the amount of the reflected light input to the photoelectric conversion element is higher than that when the curve is not curved. It changes, and the output value also changes accordingly, causing a difference from the stored value. When the difference exceeds a predetermined value, the threshold value determining unit replaces the output value with the stored value, and determines the threshold value based on the replaced value. As a result, even if the threshold value determination area of the form rises, it is possible to prevent the character recognition rate from decreasing.

[0014]

Embodiments of the present invention will now be described in detail with reference to the drawings.

FIG. 1 is a schematic block diagram showing an embodiment of the optical character reading device of the present invention.

The optical character reading apparatus 10 of the present embodiment shown in FIG. 1 is an A / D converter for a scanner unit 11 for detecting a form image from a form P and a form image (multivalued image) detected by the scanner unit 11. A sheet buffer 13 that is stored via the converter 12, a threshold value determination unit 14 that determines a threshold value, and a binarization processing unit 15 that performs binarization processing on the form image stored in the sheet buffer 13. A shading correction unit 16 that performs shading correction on the output (binary image) of the binarization processing unit 15, a recognition unit 17 that performs character cutting and character recognition processing on the binary image corrected by the shading correction unit 16, This device 10
And a control unit 18 that controls the respective units.

The form P to be processed by the present apparatus 10 is provided with a threshold value determining area Pa (see FIG. 8) of a constant width L in which characters and the like are not recorded, and the threshold value is determined. It is assumed that an image area in which characters and the like following the area Pa are recorded is provided.

Next, the details of each of the above parts will be described.

As shown in FIG. 1, the scanner unit 11 includes a light source 110 for irradiating the form P with light, and a photoelectric converter 111 for converting the reflected light from the form P into an electric signal corresponding to the amount of light. The lens system 112 collects the reflected light from the form P and guides it to the photoelectric converter 111.

As the light source 110, for example, a tubular fluorescent lamp that can illuminate the entire width of the form P with a predetermined illuminance is applied.

The photoelectric converter 111 has a plurality of photoelectric conversion elements arranged in an array along the width direction of the form P, and a CCD linear image sensor is applied, for example.

The threshold value determining section 14 has a maximum value extracting section 140, a maximum value memory 141, a reference value memory (storage means) 142, and a judging section 143, as shown in the schematic configuration diagram of FIG. , And a correction unit 144.

Under the control of the control unit 18, the maximum value extraction unit 140 takes in the data corresponding to the threshold value determination region Pa in the form image stored in the sheet buffer 13 and stores it in the threshold value determination region Pa. Is extracted for each element (channel) of the photoelectric converter 111, and the extracted maximum value is stored in the maximum value memory 141.

The reference value memory 142 stores the reference curve 20 and the uncorrected area 21, as shown in FIG. The reference curve 20 is obtained, for example, by actually reading a reading target on which characters and the like are not recorded when the device 10 is initialized (adjustment work). Therefore, the reference curve 20 reflects the light emission characteristics of the light source 110 and the light collection characteristics of the lens system 112. The obtained reference curve 20 is stored in the reference value memory 142 under the control of the control unit 18. Uncorrected area 21
Is added to the upper and lower sides of the reference curve 20 by the control unit 18, and the same reference value memory 1 is added.
42 is stored.

The determination unit 143 determines that the maximum value (output curve 2) stored in the maximum value memory 141 is the uncorrected area 21.
It is determined whether or not it is inside, and the determination result is output to the correction unit 144 (see FIG. 4 described later).

The correction unit 144 corrects the output curve 2 stored in the maximum value memory 141 based on the determination result from the determination unit 143. That is, regarding the portion where the output curve 2'is out of the uncorrected area 21,
The portion 2'is replaced with the value of the reference curve 20, and, for example, 1/2 of each value of the corrected output curve is determined as the threshold value.

Under the control of the control unit 18, the binarization processing unit 15 makes the form image stored in the sheet buffer 13 white / black based on the threshold value determined by the threshold value determination unit 14. Is converted into a binarized image corresponding to, and sent to the shading correction unit 16.

The shading correction unit 16 includes the light source 1
The output distortion (shading) of the photoelectric converter 111 caused by the light emitting characteristics of the lens 10, the condensing characteristics of the lens system 112, and the like is corrected.

Next, the operation of this embodiment having the above structure will be described with reference to FIGS. FIG. 4 is a diagram showing the maximum CCD output value before correction, FIG. 5 is an enlarged view of the portion A in FIG. 4, and FIG. 6 is a diagram showing the maximum CCD output value after correction and the decision threshold value. It is assumed that the reference value memory 142 stores the reference value 20 and the uncorrected area 21, as shown in FIG.

First, when the apparatus 10 is activated, the form P
However, the reading position R where the light is emitted from the light source 110
The upper part is transported by a transport mechanism (not shown). The light reflected by the form P is condensed by the lens system 112 and guided to the photoelectric converter 111. Next, each element of the photoelectric converter 111 changes the reflected light from the form P into an electric signal according to the amount of light. The electric signal output from the photoelectric converter 111 is converted into digital data by the A / D converter 12 and stored in the sheet buffer 13 as a form image.

The maximum value extraction unit 140 of the threshold value determination unit 14
Under the control of the control unit 18, takes in the data corresponding to the threshold value determining area Pa in the form image stored in the sheet buffer 13, and sets the maximum value in the threshold value determining area Pa to the photoelectric converter 111. For each element, and the determined maximum value is stored in the maximum value memory 141 as the output curves 2 and 2 '.

The judging section 143 judges whether all the output curves 2 stored in the maximum value memory 141 are within the uncorrected area 21, and outputs the judgment result to the correcting section 144.

Based on the determination result from the determination unit 143, the correction unit 144 determines that the output value curve 2'stored in the maximum value memory 141 has an uncorrected area 21 as shown by the phantom lines in FIGS. For the portion deviating from, the value of the portion 2'is replaced with the value of the reference curve 20. That is, as shown in the enlarged view of FIG. 6, the output curve 2 remains as it is in the uncorrected area E1, and the value is corrected to the value of the reference curve 20 in the corrected area E2. Then, for example, 1/2 of each value of the corrected output curve 22 is determined as the threshold value 3 and the determination information is output to the binarization processing unit 15.

Under the control of the control unit 18, the binarization processing unit 15 makes the form image stored in the sheet buffer 13 white / black based on the threshold value determined by the threshold value determination unit 14. It is converted into a corresponding binarized image and sent to the shading correction unit 16. Shading correction unit 16
Performs shading correction on the output (binary image) of the binarization processing unit 15. The recognition unit 17 performs character cutting and character recognition processing on the binary image corrected by the shading correction unit 16.

According to the above-described embodiment, even if the threshold value determination area Pa of the form P rises, the character recognition rate can be prevented from lowering. In addition, the threshold value determination area Pa
Has a constant width L, it is possible to prevent the occurrence of an error in determining the threshold value even if dirt, dust or the like adheres.

The present invention is not limited to the above embodiment,
Various modifications can be made without changing the gist of the invention.

[0037]

According to the present invention described in detail above, even if the threshold value determining region of the form is lifted, the output value of the photoelectric conversion element is stored in a state where the threshold value determining region is not bent. Since the value is replaced and the threshold value is determined based on the replaced value, it is possible to provide an optical character reading device capable of preventing a decrease in the character recognition rate.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram showing an embodiment of an optical character reading device of the present invention.

FIG. 2 is a schematic configuration diagram of a threshold value determining unit according to the present embodiment.

FIG. 3 is a diagram for explaining an operation of a threshold value determining unit of the present embodiment.

FIG. 4 is a diagram showing a CCD maximum output value before correction.

5 is an enlarged view of part A in FIG.

FIG. 6 is a diagram showing a corrected CCD maximum output value and a determination threshold value.

FIG. 7 is a side view of a main part of a scanner unit.

8 is a plan view of the main part of the scanner unit shown in FIG.

FIG. 9 is an output characteristic diagram of a photoelectric conversion element.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Optical character reader 110 Light source 111 Photoelectric converter 14 Threshold value determination unit 142 Reference value memory (storage unit) 143 Judgment unit 144 Correction unit P Form Pa Threshold value determination region

Claims (1)

[Claims] 1. An optical system for optically reading a character recorded in the image area by optically scanning a form in which a threshold value determination area in which the character or the like is not recorded and an image area in which the character or the like is recorded are provided. In the character reading device, a light source that irradiates the form with light, a photoelectric conversion element that converts reflected light from the form into an electric signal, and the photoelectric conversion element in a state where the threshold value determining region is not bent. Of the output value of the photoelectric conversion element based on the reflected light from the image area and the storage value of the storage means, the difference between the output value and the storage value is a predetermined value or more. And a threshold value deciding unit for deciding a threshold value based on the replaced value, when the output value is changed to the stored value.