JPH05180781A - Method and apparatus for surface defect inspection - Google Patents

Abstract

PURPOSE:To achieve highly accurate detection of the defects even when there is a trend in the intensity of an image signal about any of an unevenly colored defect, a defect due to a fine change in the surface roughness and a dot-like or linear fine defect. CONSTITUTION:Light irradiates the surface of a steel plate 26 from a light source 10 and the surface of the steel plate 26 is photographed with a TV camera 12 to transmit an image signal. The image signal is converted into digital with an A/D converter and the image signal converted is stored in a memory 20 while undergoing a specified image processing with an image processing section 18. The processing section 18 smoothes the image signal and an offset value is added evenly to the smoothed image signal to determine a threshold. A difference is determined between the original image signal and the threshold and a positive value of the difference is binary coded. The presence of a surface defect is judged from an area factor alpha of a dark part 32A of the binary coded signal. The results of the judgment are printed out from an outputting section 24 or shown on a monitor.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surface defect inspection method and apparatus, and in particular, it is obtained by imaging the surface of an object to be inspected, which is suitable for use in, for example, inspection of surface defects of cold rolled steel sheet. The present invention relates to a surface defect inspection method and apparatus for detecting a surface defect of an object to be inspected from the image signal.

[0002]

2. Description of the Related Art Generally, in the inspection of surface defects of a cold-rolled steel sheet, a spot-shaped laser beam is irradiated on the surface of the cold-rolled steel sheet, as shown in, for example, JP-A-62-75234. Inspection using a laser scattering method or a laser diffraction method that utilizes the scattering and diffraction phenomenon of laser light caused by the unevenness of

It is also practiced to take a picture of the surface of the steel sheet with a television camera and inspect the image signal obtained thereby for defects on the surface. In such a surface defect inspection, for example, as shown in JP-A-56-77704, a binarization method using a threshold value is generally used, and the image signal obtained by the binarization is equal to or more than a set value or If the brightness is less than or equal to the set value, it is determined that there is a surface defect.

[0004]

However, in the inspection by the laser diffraction method or the laser scattering method, it is impossible to detect the irregular color defect existing on the surface of the inspection object (for example, cold rolled steel sheet). Moreover, it is difficult to detect a defect due to a minute change in surface roughness.

Further, conventionally, in the inspection technique in which the surface of a steel plate is imaged by the television camera and the image signal is binarized, diffracted light or scattered light due to the surface roughness of the steel plate surface is randomly distributed in the image signal. Since a signal is generated, it is difficult to detect a defect due to a minute change in surface roughness or a dot-like or linear minute defect. Further, there is a problem that it is difficult to accurately detect a defect because the image signal intensity is tilted due to the unevenness of the intensity of the illumination light and the photographing angle.

The present invention has been made in order to solve the above-mentioned conventional problems, and includes any of defects of uneven color, defects due to a minute change in surface roughness, and fine defects of dot or line. Can detect defects with high accuracy,
Moreover, it is an object of the present invention to provide a surface defect inspection method and apparatus capable of accurately detecting a defect even when an inclination of image signal intensity occurs.

[0007]

SUMMARY OF THE INVENTION The present invention is a surface defect inspection method for detecting a surface defect of an object to be inspected from an image signal obtained by imaging the surface of the object to be inspected. The above problem is solved by uniformly adding an offset value to the binarized image signal to obtain a threshold value, binarizing the original image signal with the obtained threshold value, and detecting the surface defect from the binarized image signal. To do.

Further, according to the present invention, in a surface defect inspection apparatus for detecting a surface defect of an object to be inspected, a light source for irradiating the object to be inspected with light and an image signal of the surface of the object to be inspected are picked up. For transmitting the image signal, an analog / digital converter for analog / digital converting the transmitted image signal, an image memory for storing the analog / digital converted image signal, and an image signal for smoothing the image signal. Means, means for uniformly adding an offset value to the smoothed image signal to obtain a threshold value, obtaining a difference between the original image signal and the threshold value, and binarizing the positive value of the difference. Means, a surface defect determining means for determining the presence or absence of a surface defect from the area of the binarized signal, and an output means for outputting the determination result by printing out or displaying an image. By It solves the shaft above problems.

In the present invention, the light source is rod-shaped,
In addition, continuous light may be emitted.

In the present invention, the light source is rod-shaped,
In addition, the light can be repeatedly irradiated at high speed.

[0011]

The inventor has
In the surface defect inspection technique of the object to be inspected using the laser diffraction method or the laser scattering method, in view of the fact that it is impossible to detect the irregular color defect, the technique of inspecting the surface defect by the image pickup means such as a television camera. Decided to adopt.

In the conventional defect inspection technique for binarizing an image signal by using such an image pickup means, a random signal () is generated in the image signal by the diffracted light or scattered light due to the surface roughness of the inspection object, for example, a steel plate. (Becomes noise) occurs, and it is difficult to detect defects due to minute changes in surface roughness and minute defects such as point-like or linear defects. In addition, it is difficult to accurately detect a defect when the intensity of the illumination light is uneven or the image signal intensity is inclined depending on the photographing angle.

Therefore, in order to remove the noise part in the image signal due to the diffracted light or scattered light and to correct the unevenness of the intensity of the illumination light or the inclination of the image signal intensity depending on the photographing angle, various examinations are made. The invention was invented.

That is, in the present invention, the image signal obtained by imaging the surface of the object to be inspected is smoothed, an offset value is uniformly added to the smoothed image signal to obtain a threshold value, and the obtained threshold value is used. The original image signal is binarized.

The smoothing process reduces the influence of randomly generated signals due to surface roughness. Further, as described above, the threshold value is calculated by adding the uniform offset value and binarized, so that the inclination of the image signal intensity due to the intensity unevenness of the illumination light or the like is corrected and only the defective portion is binarized. It will be.

A surface defect of the object to be inspected is detected from the binarized image signal. For example, the presence or absence of surface defects can be determined from the area of the binarized image signal.

Therefore, it is possible to accurately detect any of the irregular color defect, the defect due to the minute change of the surface roughness, the fine dot defect or the linear defect, and the image signal intensity. The defect can be accurately detected even when the inclination occurs.

The determination result can be printed on a sheet of paper by a printer or displayed on a screen of a display means (monitor) so that the inspector can perceive it.

If the light source for irradiating the object to be inspected with light is rod-shaped and irradiates continuous light, the surface defect can be detected from the image signal at an arbitrary timing.

If the light source is rod-shaped and is irradiated with light repeatedly at high speed, it is possible to irradiate the object to be inspected with relatively strong light and obtain a clear image signal of light and darkness. The accuracy of surface defect detection can be improved.

[0021]

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

This embodiment is a surface defect inspection apparatus having a structure as shown in FIG.

As shown in FIG. 1, this surface defect inspection apparatus mainly comprises a light source 10, a television camera 12, an analog / digital (A / D) converter 14, an image input section 16, an image processing section 18, The memory 20, the defect determination unit 22, and the output unit 24 are provided.

The light source 10 is a steel plate 26 which is an object to be inspected.
It is a rod-shaped light source for irradiating the strip-shaped light 28 on the inside. As the light source 10, a light source that emits continuous light or a light source that repeatedly emits light at high speed can be used. The reference numeral 30 on the steel plate 26 in FIG. 1 is a surface flaw (defect).

The television camera 12 captures an image of the surface of the steel plate 26 irradiated with the irradiation light 28. As the television camera 12, for example, an industrial television camera using a CCD (charge coupled device) as an image pickup device can be used.

The A / D converter 14 A / D converts the analog image signal output from the television camera 12 into a digital image signal.

The image input section 16 transmits the A / D converted image signal to the image processing section 18 and the memory 20.

The image processing section 18 smoothes the image signal,
An offset value is uniformly added to the smoothed image signal to obtain a threshold value, a difference between the original image signal and the threshold value is obtained, and a positive value of the difference is binarized to obtain an original image. 2
It is a process of performing a digitization process.

The memory 20 stores the original image signal input from the image input section 16 and the image signal processed by the image processing section 18.

The defect determining section 22 is for determining the presence or absence of a defect on the surface of the steel plate 26 from the image signal subjected to the image processing. Details of the defect determination performed by the determination unit 22 will be described later.

The output unit 24 prints the result of the defect determination on the paper surface by the printer or displays it on the screen of the monitor.

Here, an example of the smoothing method performed in the image processing section 18 will be described. That is, as shown in FIG. 2, 2 ⁿ pixels × 2 ⁿ pixels on the image signal (n = 0 to 9)
The value (density value or the like) of each pixel in the area (also referred to as a template) is smoothed, and the value obtained by the smoothing is put in the leftmost point (pixel) in the area. This area is
In the scanning direction indicated by arrow A in FIG. 2, when one scanning line is completed, the scanning line is moved to the next scanning line and the line is moved downward as indicated by arrow B.

FIG. 2 shows an area (pixels a to p) of 4 pixels × 4 pixels where n = 2, and in this area, the value U smoothed as shown by the following equation (1) is used. Into pixel a.

[0034]

[Equation 1]

However, fix means a fixed point (integer). If the value U is 8-bit data, 0 to 25
It is an integer value in the range of 5.

Usually, the value of n is n = 4-5.

The offset value is determined as follows.

The offset value is set to a value larger than the noise component of the original image signal and smaller than the signal strength of the defective portion. This offset value (a value within the range of 0 to 255 when the image signal is 8-bit data) is not set softly (automatically), but a fixed offset value can be added. The setting of the offset value can be arbitrarily changed depending on the lighting conditions. The offset value can be set to about 30-40.

The image processing section 18 is not provided with a smoothing processing circuit and a circuit having a function of giving an offset value, and a smoothing process and an offset value are given by driving a dedicated processor by microcode. It is carried out.

The presence / absence of a defect in the defect determining section 22 is determined as follows.

That is, assume that the binarized image signal (screen) output from the image processing section 18 is as shown in FIG. FIG. 3 shows an image signal having a screen of 512 pixels horizontally and 432 pixels vertically. Reference numeral 30A in FIG. 3 is a dark portion on the surface of the steel plate 26 (when the background is bright. When the background is dark, the bright portion).

In the defect judging section 22, the dark section 30
From the number of pixels A of A, the dark portion 30A for the screen of the image signal
If the area ratio is determined to be equal to or larger than a predetermined value, the dark portion 30A is determined to be a defect due to a surface flaw. That is, the number of pixels of the screen is 512 × 432, and the dark portion 30
Since the number of pixels of A is A, the area ratio α (%) is calculated by the following equation (2).

[0043]

[Equation 2]

In this case, for example, if the area ratio α is 5% or more, it is determined that the defect is due to a surface flaw.

Next, the operation of the embodiment will be described.

The surface defect inspection apparatus of the embodiment detects defects (surface defects) on the surface of the steel plate 26 by the procedure shown in FIG.

First, the television camera 12 captures an image of the steel plate 26 in the state where the light source 10 emits light (step 101).

Since the image signal output from the television camera 12 is an analog signal, it is converted into a digital signal by the A / D converter 14 (step 102).

Then, the digitally converted image signal (original image signal) is stored in the memory 20. An example of the original image signal thus stored is shown in FIG. It is assumed that there is a dark portion 32 due to a surface defect on the stored image signal. In this case, the luminance data of the image signal obtained along the scanning line EF in FIG. 5 is as shown in FIG.

Next, in the image processing section 18, the original image signal is smoothed by the above-mentioned FIG. 2 and the equation (1) (step 104). When the image signals shown in FIGS. 5 and 6 are smoothed, the luminance signal on the scanning line EF becomes as shown in FIG. 7, which is smoother in a blunted state than in FIG. ..

Next, an offset value is uniformly added to the smoothed image signal to obtain a threshold value (step 105). The threshold value obtained by adding the offset value to the smoothed image signal of FIG. 7 is as shown in FIG. 8, for example.

Next, the difference between the original image signal and the threshold value is obtained (step 106).

Then, the positive value of the obtained difference is binarized (step 107). The result of binarizing the positive value of the difference between the original image signal of FIG. 6 and the threshold value of FIG.
As shown in FIG. Further, when it is shown on the screen, it becomes as shown in FIG.

Next, the area ratio α of the dark portion 32A extracted by the binarized signal is calculated by the above equation (2) (step 108).

Next, the area ratio α is a predetermined value (for example, 5%).
Based on the above, it is determined whether or not there is a defect in the imaged steel plate 26 (step 109).

If the area ratio α is equal to or more than the predetermined value,
If the area ratio α is less than the predetermined value, it is determined that the steel sheet 26 has no defects (OK).

The output unit 24 prints out the determination result by the printer or displays it on the monitor screen.
Thereby, the inspector knows whether the steel plate 26 has a defect.

In the above-mentioned embodiment, the steel sheet is exemplified as the object to be inspected, but the object to be inspected for defects in the present invention is not limited to the steel sheet, and the objects to be inspected of various other materials are not limited thereto. The present invention can be used to detect defects in an inspection object.

[0059]

As described above, according to the present invention, it is possible to accurately detect defects such as irregular color defects, defects due to minute changes in surface roughness, and dot-like or linear minute defects. An excellent effect that the defect can be detected and the defect can be accurately detected even when the image signal intensity is tilted is obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing an entire configuration of a surface defect inspection apparatus according to an embodiment of the present invention, including a partial perspective view.

FIG. 2 is a plan view for explaining a smoothing process of the image processing apparatus in the apparatus of the embodiment.

FIG. 3 is a plan view for explaining a defect determination process of the image processing apparatus.

FIG. 4 is a flow chart showing a defect inspection procedure for explaining the operation of the apparatus of the embodiment.

FIG. 5 is likewise a plan view showing an example of an original image signal.

FIG. 6 is a diagram similarly showing an example of a luminance signal distribution on a scanning line in an original image signal.

FIG. 7 is a diagram similarly showing a luminance signal distribution after the image signal is smoothed.

FIG. 8 is a diagram similarly showing an example of a threshold value obtained by adding an offset value to the smoothed image signal.

FIG. 9 is a graph in which, similarly, the difference between the obtained image and the original image signal is obtained by the threshold value, and the positive value is binarized.
It is a diagram which shows the example of value-ized data.

FIG. 10 is a plan view showing an example of a binarized image signal screen.

[Explanation of symbols]

 10 ... Light source, 12 ... Television camera, 14 ... Analog / digital (A / D) converter, 16 ... Image input section, 18 ... Image processing section, 20 ... Memory, 22 ... Defect determination section, 24 ... Output section, 26 ... Steel plate (inspection object), 28 ... Irradiation light, 30 ... Surface flaw (defect), 30A, 32, 32A ... Dark part.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Office reference number FI Technical display location G06F 15/68 320 Z 8420-5L

Claims (4)

[Claims] 1. A surface defect inspection method for detecting a surface defect of an object to be inspected from an image signal obtained by imaging the surface of the object to be inspected, wherein the image signal is smoothed and uniformly offset to the smoothed image signal. A surface defect inspection method comprising adding a value to obtain a threshold value, binarizing an original image signal by the obtained threshold value, and detecting the surface defect from the binarized image signal. 2. A surface defect inspection apparatus for detecting a surface defect of an object to be inspected, for illuminating the object to be inspected with light, and for imaging the surface of the object to be inspected and transmitting an image signal. Image pickup means, analog / digital converter for analog / digital conversion of transmitted image signal, image memory for storing analog / digital converted image signal, means for smoothing image signal, smoothing Means for uniformly adding an offset value to the converted image signal to obtain a threshold value, and means for obtaining a difference between the original image signal and the threshold value and binarizing the positive value of the difference. And a surface defect determination means for determining the presence or absence of a surface defect from the area of the binarized signal, and an output means for outputting the determination result by printing out or displaying an image. Table Surface defect inspection device. 3. The surface defect inspection apparatus according to claim 2, wherein the light source is rod-shaped and emits continuous light. 4. The surface defect inspection apparatus according to claim 2, wherein the light source is rod-shaped and is repeatedly irradiated with light at high speed.