JPH11312216A - Character recognition device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To correctly identify a character by automatically setting the optical condition of illumination to various kinds of character recognizing objects. SOLUTION: In order to optimize character recognition at an image processing part 5 by deciding the normal/defective conditions of an optical condition by light sources 2 and 2, based on the mixing degree of contrast and noise, a control signal for setting the optical condition of illumination to an inspecting object matter 1 by the light sources 2 and 2 is outputted from the image processing part 5 to a motor driving control part 6 for varying the light incident angle and a light quantity control part 7.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention comprises an image processing unit and a television camera for picking up an image of a target part and transmitting an image signal to the image processing unit in order to automatically identify the serial number of an industrial product. It relates to a character recognition device.

[0002]

2. Description of the Related Art In recent years, liquid crystal panels, semiconductor wafers, PDs
In an inspection process of an industrial product such as a P substrate, accurate classification of a product and classification of a defective type are important issues. There is a strong demand for automation of these classification tasks, and an image input unit including an optical system and a television camera, and an image processing unit including a computer that performs image signal processing are widely used. For example,
As an example, a character recognition device using a television camera for imaging a required part in order to automatically identify a product number or the like written on a product can be given as an example.

FIG. 11 is a configuration diagram showing a configuration of a television camera portion in a conventional character recognition device, and FIG.
FIG. 11B is a front view, and FIG. 11B is a side view, in which light La emitted from the light source 101 is incident obliquely from above onto an object 100 having a product number stamped or drawn on the surface. In this configuration, the television camera 102 captures an image using the reflected light Lb from the object 100, outputs the image signal to an image processing unit (not shown), and performs character identification processing in the image processing unit.

[0004]

However, in the above-mentioned conventional character recognition device, the state of film formation, polishing, etc. differs depending on the type of production or each process due to the production of many kinds in small quantities or the complexity of the production process. The state of the character portion changes significantly, and a signal input to the image processing portion for performing character identification is likely to be disturbed due to contrast fluctuation or mixing of a noise component. It is strongly desired to be able to flexibly follow fluctuations.

However, in the configuration of the light source provided for the television camera in the character recognition device as shown in FIG. 11, the light incident angle of the light source 101 with respect to the object 100 is fixed, It is necessary to set the optical conditions of the light source 101 according to the situation of the character portion of the object 100 that differs depending on the difference.

As described above, in the related art, in order to accurately identify characters of an object, it is necessary to perform an operation such as setting an angle of a light source for each type or process. If the setting operation is inaccurate, the identification process will be seriously hindered. Therefore, the light source setting operation needs to be done with great care, which is a complicated operation.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned conventional problems by providing a simple and reliable automatic setting of optical conditions, and a character recognition apparatus capable of accurate character recognition in various character recognition targets. To provide.

[0008]

In order to achieve the above object, the present invention provides an optical system in which a light emitting angle and a light amount of a light source with respect to an imaging target portion of a television camera are changed by remote control from an image processing unit. The conditions can be automatically changed. With this configuration, complicated setting work is not required, and accurate and accurate character recognition for various character recognition targets using a television camera is performed with ease.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS According to the first aspect of the present invention, an image processing unit and an image of a target part are taken and an image signal is sent to the image processing unit in order to automatically identify a serial number and the like. In a character recognition device equipped with a television camera, the optical condition is automatically changed by enabling the setting of the irradiation angle and the light amount of the light source with respect to the imaging target portion by remote control from the image processing unit. With this configuration, it is possible to appropriately illuminate the imaging target site for imaging by the television camera so that character identification by the image processing unit is accurately performed without performing complicated setting work.

The invention according to claim 2 is characterized in that the light emission angle of the light source is made variable by a driving device.

According to a third aspect of the present invention, a plurality of light sources are provided at different light emission angles, and the light amounts of the respective light sources are changed independently of each other. By selecting the light source and setting the amount of light, a device for driving the emission angle of the light source can be eliminated.

According to a fourth aspect of the present invention, in the image processing section, optimal optical conditions are set based on contrast judgment information and identification evaluation information.

According to a fifth aspect of the present invention, in the character recognition in the image processing unit, if the image cannot be identified by the imaging under the set optical conditions, the setting is changed to another optical condition to perform the imaging and recognition. Is performed again, and with this configuration, the optical conditions that always match the character recognition ability of the image processing unit are set.

According to a sixth aspect of the present invention, in the image processing section, several kinds of optical conditions are prepared in order based on the optical conditions for a recognition target having a high frequency of character identification execution, and the set optical conditions are set. In the case where it is not possible to identify by the imaging based on the condition, it is possible to perform the imaging and recognition again by changing the setting according to the above order to other optical conditions, and this configuration allows the character in the image processing unit The setting of the optical condition that matches the recognition ability is performed practically and efficiently based on the use results.

Hereinafter, a preferred embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a perspective view for explaining the entire character recognition device according to the first embodiment of the present invention, FIG. 2A is a front view of a light source portion of FIG. 1, and FIG. FIG. 1 is a side view of a light source portion, where 1 is an inspection object having characters written on its surface, 2, 2.
Is a pair of light sources installed above the inspection target 1 in this example at symmetrical positions, and 3 and 3 are installed on each of the light sources 2 and 2 to change the light emission angle of the light sources 2 and 2. An angle changing drive motor 4 is installed at a position above the light sources 2 and 2 to receive a reflected light from the inspection object 1 and to take a picture of the irradiation surface of the inspection object 1 and a television camera 5. An image processing unit that performs image processing and character identification in response to an image pickup signal from 4; and a motor drive control unit 6 that drives an angle changing drive motor 3, 3 in response to a motor control signal output from the image processing unit 5. , 7 are light amount control units that receive light amount control signals output from the image processing unit 5 and adjust the light amounts of the light sources 2 and 2, and 8 and 8 are condensing lenses provided on the light emission side of the light sources 2 and 2. is there.

FIG. 3 is a perspective view for explaining the structure of the light source as an example of the light source.
D10 is provided side by side, and the light emitted from each LED 10 is made uniform by the diffusion plate 11 and shaped into slit light by the light shielding plate 12, and further, the slit light is given directivity to scatter light. In order to prevent this, a condensing lens 8 is provided.

The image processing section 5 receives an image pickup signal sent from the television camera 4, performs image processing, and performs processing such as character identification. As an example of character recognition in the image processing unit 5, for example, an input pattern after image processing is a mesh pattern, and mesh pattern data is searched from column to column from top to bottom and from bottom to top, and the searched black It is possible to adopt a method of performing character recognition based on the vertical and horizontal structural vectors obtained by the mesh and the white mesh, the area difference between black and white, the number of intersections with black, and the like.

Further, the image processing unit 5 has a contrast as shown in FIGS. 4 and 5 in order to evaluate optical conditions for causing the inspection object 1 to be illuminated favorably by the light sources 2 and 2. It has a measurement function and a function of measuring the degree of noise contamination at the time of character identification as shown in FIGS.

FIG. 4 is a table showing the results of measuring the contrast of light sources 2 and 2 (a pair of light sources are identified as A and B in FIG. 2) by the light amount and angle. The evaluation of the contrast is performed, for example, on the background portion and the character portion of the inspection object 1 on which the character is written as shown in FIG. G _c and g _{b at which the} maximum frequency of the
It can be determined that the greater the value of the density difference N _A = | g _c −g _b |, the higher the contrast and the better the optical conditions.

FIG. 6 is a table showing the measurement results of the character identification evaluation values based on the light amounts and angles of the light sources 2 and 2 (the light sources are distinguished as A and B in the same manner as described above). In the character identification evaluation, the inspection target 1 on which the characters (A, B, C, 0, 1, 2, 3) are written as shown in FIGS. 7A, when the optical condition of FIG. 7A is smaller than that of FIG. 7B in the optical condition, the discrimination as shown in FIGS. 8A and 8B is performed. It looks like a character evaluation value. FIG. 8A shows an evaluation value of the identification character corresponding to FIG.
(B) is an evaluation value for the identification character corresponding to FIG. 7 (b). As is clear from the figure, the inspection object 1 shown in FIG. 7 (b) is shown in FIG. 8 (b). Like the evaluation value,
Mean value V _B of the evaluation values of each character is lowered by noise mixing. Thus, it can be determined that the inspection object 1 in FIG. 7B has a higher noise mixing degree.

As the evaluation of the discrimination evaluation value, for example, a judgment value by the following attribute judgment method can be used. That is, the plurality of feature data of the input pattern after the image processing in the image processing unit 5 is compared with the membership functions of the plurality of feature data previously obtained from a large number of standard patterns, and the membership functions for all the feature amounts are compared. Of all the standard patterns, and the degree of attribute discrimination can be evaluated by fuzzy inference based on the total value. By performing this evaluation method in the same manner on the character recognition data, the degree of character identification can be evaluated.

[0023] As described above, the quality of the optical condition by the light source 2, 2 using the average value V _B of the identification evaluation value representing the contamination degree of the contrast measurement results V _A and the noise, the optical condition evaluation value V = V _By making a determination based on _A × V _B , it is possible to set the optical conditions for illumination of the inspection target 1 by the light sources 2 and 2 so that character identification in the image processing unit 5 is optimized. Optical condition evaluation value V '= V _A +
Even V _B, similar results are obtained.

For this reason, conventionally, even if the object to be inspected is the same, since the type or process is different, for example, the reflectance of the character portion changes, which affects the character identification and makes it difficult to identify the character. In some cases,
Although the work of changing the optical conditions of the light source was necessary, by automatically changing the optical conditions as described above as in the present embodiment, the effect of irregularly reflected light is suppressed, and the contrast is improved. Accurate character identification has become possible without the need for complicated work.

When character recognition in the image processing unit 5 cannot be performed by imaging under optical conditions that have already been set, the setting is changed to another optical condition so that imaging and recognition are performed again. You may.

FIG. 9 is a flowchart relating to the above-described optical condition change. In character recognition in the image processing unit 5, first, optical condition initial setting is performed (S1).
Based on the conditions, a control signal is output to the motor drive control unit 6 and the light amount control unit 7 to control the light emission angles or light amounts of the light sources 2 and 2. In this state, character recognition is performed in the image processing unit 5 (S2). However, as a result of the character recognition, if the image processing unit 5 cannot identify the object (S3), the motor drive control unit 6 and the light amount control unit 7 are changed from the image processing unit 5 to another optical condition. Then, a control signal is output (S4), the character recognition in the image processing unit 5 is executed again (S3), and the optical conditions are repeatedly set until the image processing unit 5 becomes identifiable.

It is also possible to previously register a plurality of data of optical conditions corresponding to inspection objects of different types or processes in the image processing unit 5 and to repeat the identification process as described above using the optical conditions. It is possible. Furthermore, several types of optical conditions are registered in advance in the order of high frequency based on the optical conditions for the recognition target with high frequency of character identification execution, and if it is impossible to identify by imaging with the set optical conditions, The imaging and the recognition may be performed again by changing the setting according to the above order to another optical condition.

FIG. 10A is a front view of a light source portion in the character recognition device according to the second embodiment of the present invention, and FIG. 10B is a side view of the light source portion of FIG. In the embodiment, a plurality of light sources 2 having the configuration described with reference to FIG.
), And the light emission angles are changed by each. Each light source 2 is independently controlled by the image processing unit 5.
It is configured to be able to control the amount of light by a control signal from, or to select and use a light source 2 having a required light emission angle. With this configuration, it is not necessary to use the angle changing drive motor 3 in the configuration of the first embodiment.

Also, the more the number of light sources installed, the more
It is possible to finely set the optical conditions for the inspection object 1. The number of light sources installed is appropriately set according to inspection specifications and the like.

[0030]

As described above, according to the present invention,
Without performing a complicated setting operation, it is possible to properly illuminate the imaging target site for imaging by the TV camera so that character identification according to various inspection targets is accurately performed by the image processing unit, Since the influence of irregularly reflected light can be suppressed and the optical environment such as contrast is improved, accurate character identification is realized.

[Brief description of the drawings]

FIG. 1 is a perspective view illustrating an entire character recognition device according to a first embodiment of the present invention.

2A is a front view of a light source portion in FIG. 1, and FIG. 2B is a side view of the light source portion in FIG.

FIG. 3 is a perspective view illustrating a configuration of a light source according to the first embodiment of the present invention.

FIG. 4 is a diagram showing measurement results.

FIG. 5 is a diagram for explaining the definition of contrast in the first embodiment of the present invention.

FIG. 6 is a view for explaining an evaluation function of a degree of noise contamination during character recognition according to the first embodiment of the present invention;

FIG. 7 is a view for explaining the degree of noise mixing during character recognition according to the first embodiment of the present invention;

FIG. 8 is a diagram for explaining an evaluation value of a degree of noise mixing during character recognition according to the first embodiment of the present invention;

FIG. 9 is a flowchart for setting optical conditions according to the first embodiment of the present invention.

FIG. 10 is a diagram for explaining a character recognition device according to a second embodiment of the present invention, where (a) is a front view of a light source portion,
(B) is a side view of the light source portion.

11A and 11B are views for explaining a conventional character recognition device, wherein FIG. 11A is a front view of a light source portion, and FIG. 11B is a side view of the light source portion.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Inspection object 2 Light source 3 Angle change drive motor 4 TV camera 5 Image processing part 6 Light quantity control part 7 Motor drive control part 8 Condensing lens 10 LED 11 Diffusion plate 12 Shield plate

Claims (6)

[Claims] 1. To automatically identify a serial number, etc.
In a character recognition device including an image processing unit and a television camera that images a target region and sends an imaging signal to the image processing unit, the irradiation angle and light amount of a light source with respect to the imaging target region are controlled by remote operation from the image processing unit. A character recognition device characterized in that the optical conditions are automatically changed by enabling the setting of the following. 2. The character recognition device according to claim 1, wherein the light emission angle of the light source is made variable by a driving device. 3. The character recognition device according to claim 1, wherein a plurality of light sources are provided at different light emission angles, and the light amounts of the respective light sources are independently changed. 4. The character recognition device according to claim 1, wherein the image processing unit sets an optimal optical condition based on contrast determination information and identification evaluation information. 5. In the character recognition in the image processing unit, if the image cannot be identified by the imaging under the set optical condition, the setting can be changed to another optical condition and the imaging and recognition can be performed again. The character recognition device according to claim 1, wherein 6. The image processing unit prepares a sequence of several types of optical conditions based on optical conditions for a recognition target having a high frequency of character identification execution, and cannot perform identification based on imaging using the set optical conditions. 2. The character recognition device according to claim 1, wherein, in such a case, the imaging and the recognition can be performed again by changing the setting according to the order according to another optical condition.