JPH09138200A - Method for determining surface defect of strip material - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a method for determining the defect of an image equivalent to that of visual inspection and to eliminate erroneous determination of defect due to contamination of steel plate which may present over a wide range. SOLUTION: A defect signal detected by an optical means is subjected to image processing at a control section 20 and then the type and class of defect are determined at an execution processing section 22. For all defects detected in a section of preset constant length and the type and class thereof are determined, the number of defects is counted for each type and class at the defect accumulating section 2 in macro processing section 1. When each count exceeds a threshold value being set for each type and class of defect at a setting section 3, an execution processing section 4 sets the defect in heavier class.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surface defect judging method for judging a surface defect generated on the surface of a strip-shaped inspected object such as a cold-rolled steel plate, aluminum or the like. The present invention relates to a method of re-determination so as to match a human visual inspection image, and a method of eliminating erroneous determination of defects due to stains and the like on steel plates that are often present over a wide range.

[0002]

2. Description of the Related Art When manufacturing strips such as steel plates, surface defects are inspected in order to guarantee the quality of the products, and for example, the type and grade of surface defects are determined. As an apparatus for inspecting such surface defects, an optical inspection apparatus having a configuration shown in FIG. 6 is known.

The inspection apparatus includes a laser transmitter 10 as a light source and a rotary mirror 12 for scanning a laser beam from the laser transmitter 10 in the width direction of a steel plate (strip) 11.
And a condenser lens 14 for condensing the laser light reflected on the surface of the steel plate 11 at one point, and receives the reflected laser light condensed by the condenser lens 14 via a spatial filter (mask) 16. And a photoelectric converter 18.

[0004] Further, the above-mentioned inspection apparatus comprises the above-mentioned photoelectric converter 1
A control unit 20 for performing image processing and predetermined pre-processing on the output signal from the control unit 8; extracting a characteristic parameter of the defect signal from the signal processing output from the control unit 20; And an arithmetic processing unit 22 for comparing the characteristic parameters for each type of defect to determine the type and class of the defect and for editing the determined defect signal. By outputting the edited defect information, the inspection result is printed and a development table is created.

In the above inspection apparatus, the steel plate 11 scans the laser beam while traveling in the direction of the arrow, and the reflected light at that time is received by the photoelectric converter 18 and the control unit 20
In FIG. 7, the output signal from the photoelectric converter 18 is processed according to the profile method shown in FIG.

That is, first, the traveling steel plate 11 is shown in FIG.
Prescribed size, for example, 250mm length x 1 width
The image processing unit 26 is divided into 00 mm image processing units 26, and each of the image processing units 26 normally has a side 2 to 2 as shown in FIG.
It is configured as an aggregate of pixels having a size of about 5 mm.
Each of these pixels corresponded to each other by comparing the signal value of each pixel due to the defect of FIG. 7 (A) with preset levels r _{1 to} r ₆ as shown in FIG. 7 (C). r ₁
Has a digital value of ~r _6, constitutes a digital map 26A as a whole.

Next, with respect to the digital map 26A, the number of pixels having signal values of r _{1 to} r ₆ is integrated in the vertical direction to create a surface defect profile consisting of a histogram shown in FIG. 7D. And from the above profile,
A width W, a length H, and an area S are extracted as parameters representing the characteristics of the surface defect.

Further, through the same processing, for each image processing unit obtained by fractionating the entire steel sheet 11, the number of defects having a length H equal to or less than a certain reference, the number of point-like defects NA, a certain reference From the number of defects having the length H, the number of linear defects NB, the defect occurrence position PP from the end of the steel plate, and the signal level polarity ratio VI (r ₁ , r ₁ , shown in FIG. 7C)
Various parameters are extracted, such as the sum of the pixels taking the signals r ₂ and r ₃ divided by the total number of defective pixels.

The various defect characteristic parameters obtained in this way are then compared in the arithmetic processing section 22 of the inspection apparatus with the judgment conditions in each of the defect type judgment logics prepared in advance, and the defect type is judged. Will be Further, the defect class is determined by comparing the defect level calculated from the signal levels r _{1 to} r _{6 with} a threshold for defect class determination prepared in advance for each type of defect. Become.

[0010]

As described above,
In the conventional inspection apparatus, the type and the grade of the defect are determined only based on the comparison between the characteristic parameter of the detected defect and the previously prepared defect type determination logic and the grade determination threshold value.

By the way, the surface defect inspection apparatus is intended for visual inspection substitution, and the judgment criteria equivalent to those of the visual inspector are required for the pass / fail judgment of metal strip products by the apparatus. Should match the visual image of the inspector as much as possible.
That is, in visual inspection, for example, for a certain type of defect, if a large number of light defects are concentrated in a certain range of the plate to be inspected, the inspector judges the degree to be heavy and, conversely, In the case where a slight defect is sporadically generated, it is determined to be a harmless defect. On the other hand, the conventional inspection apparatus having the above-mentioned configuration performs the inspection / arithmetic processing for each image processing unit, and there is a problem that the determination result is different from the visual image of the inspector looking over the entire plate. .

Further, the conventional surface defect device has the following principle.
Since the laser reflection pattern from the inspected object is detected and processed, defects such as stains that are often distributed over a wide range on the inspected object or discoloration of the steel plate due to acid in the pickling line in the steel plate manufacturing process There was a problem that it was erroneously determined.

The present invention has been made in order to solve the above-mentioned problems, and makes it possible to make a defect judgment of an image equivalent to that of a visual inspector, and to eliminate the erroneous judgment of a defect due to stains of a steel plate that often exist over a wide range. It is intended to provide a way.

[0014]

[Means for Solving the Problems] Among the above-mentioned problems, the problem of realizing an automatic inspection equivalent to a visual image is to analyze the reflected light from the surface of an object to be inspected by a device, and analyze the type of surface defect of the object to be inspected. In the method of determining the surface defects of the strip-shaped inspected object that determines the grade and the grade, the number of occurrences of those defects is determined for all the defects that are detected within the preset fixed length section and for which the type and grade are determined. And count by grade,
When the counted number of defects exceeds a preset threshold for the number of defects and the number of defects generated for each class, the problem is solved by providing a function of increasing the class of defects. To be done.

Further, the above-mentioned problem is set in advance in a method for determining a surface defect of a belt-like body by analyzing the reflected light from the surface of the belt-like body to be inspected and determining the type and grade of the surface defect of the belt-like body to be inspected. The number of occurrences of these defects is counted for each type and grade for all the defects that have been detected within the fixed length section and whose type and grade have been determined, and the number of each counted defect is set in advance. When the threshold value for the number of defects generated and the number of occurrences for each grade is not exceeded, the defect is solved by providing a function of reducing the grade of the defect.

As described above, in the visual inspection, when a large number of light defects are concentrated in a certain range of the plate to be inspected with respect to a certain type of defect, the inspector determines the degree. Is determined to be a harmless defect, and conversely, when a slight defect of a small degree occurs, it is determined to be a harmless defect. Therefore, by providing the function as described above, a determination closer to a visual inspection can be performed. .

Further, among the above-mentioned problems, the problem of erroneous determination of a defect due to dirt or the like on the object to be inspected over a relatively wide range is to analyze the reflected light from the surface of the object to be inspected, and to analyze the surface of the object to be inspected. In the surface defect determination method of the band-shaped body for determining the type and grade of the defect, the number of all the defects detected within the preset fixed length section, and the type and the grade of which are determined, is a threshold value set in advance. It is solved by providing a function of changing the type of defect when the value exceeds.

Plate stains and discolorations appear over a wide range of the object to be inspected, whereas the inspection apparatus determines that many defects are present. On the contrary, many defects other than plate stains and discoloration do not appear over a wide range of the inspected object. Therefore, when defects are detected in a wide range of a predetermined number or more, it is determined that there are not many defects but plate stains or discoloration, and the defect type is changed.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings. The grades of defects are A, B, C, D, and E in order from the lightest one.

FIG. 1 is a macro judgment process of a steel plate applied to a surface defect judgment method for a strip-shaped object to be inspected according to an embodiment of the present invention (re-decision of defect grade and type is called macro judgment process. ) A block diagram showing a basic configuration.

In this embodiment, an inspection apparatus having a function substantially the same as the optical inspection apparatus shown in FIG. 7 and a macro judgment processing function is applied.

First, in FIG. 1, the control unit 20 and the arithmetic processing unit 22 are the same as those shown in FIG.
Before that, the optical system shown in FIG. 6 is also connected. Then, as described in Japanese Patent Application Laid-Open No. 4-110758, a defect feature parameter is extracted in the arithmetic processing unit 22 and the type and grade of the defect are determined.

Then, the defect determination result and the defect occurrence position information are output to the macro processing unit 1. The macro processing unit 1 accumulates the number of defects generated within a section of a preset fixed length for each defect type / grade in a defect occurrence number accumulating unit 2 and prepares the defect occurrence number setting unit 3 in advance. The arithmetic processing unit 4 compares the threshold value with the threshold value, thereby re-determining and outputting the defect type and grade according to the setting. Then, after the subject has traveled for a certain distance, the integrated value of the generated defects is reset, and the number of generated defects in the next certain section is integrated.

Next, a concrete example of the surface defect judgment method by the macro judgment processing of the strip according to the present embodiment when it is actually applied to the continuous processing line of the steel sheet will be described.

The specific procedure is as follows. (1) First, as shown in FIG. 2, a steel plate is equally divided in the width direction to each have a width of 1 W to 6 W, and 250 m in the steel plate traveling direction.
The image is divided for each m, and a portion represented by oblique lines is set as an image processing unit.

(2) The number of occurrences of each defect is counted for each defect type and grade, and for all defects in a predetermined length section, for example, 1 m from the inspection start position. However, in this case, since the number of image processing units in a certain section is 24, the maximum number of defects counted in a certain section length is 24.

(3) The type and grade of defects are re-determined by a preset macro processing algorithm using the above aggregated values in the plate 1 m.

In the above (2), when the defect judgment result by the apparatus before the macro processing within the 1 m section is, for example, FIG. 3 (a), the macro processing algorithm of the above (3) is processed as shown in FIG. 3 (b). ), The determination result after macro processing is as shown in FIG. 3 (c). That is, the number of gouges A is six, which exceeds the set number (threshold) of 5, so the grade is increased by one grade and gouge A is changed to gouge B.

If the device determination result before the macro processing is as shown in FIG. 4 (a), the macro processing algorithm shown in FIG.
By (b), the judgment is made again as shown in Fig. 4 (b). That is, since the number of gouges A is 2 and the set number (threshold value) is 2 or less, the grade is lowered by one grade and the gouges A are changed to defect-free.

As a result of these, an inspection result closer to the visual inspection result image can be obtained. Similarly, when the device determination result before the macro processing is as shown in FIG. 5A, the total number of defects exceeds 20, so that the macro processing algorithm shown in FIG.
By (b), the judgment is made again as shown in FIG. 5 (b). This eliminates the problem of misjudgment due to stains or discoloration that adheres to a wide area of the steel sheet.

[0031]

As described above, according to the present invention, the number of occurrences of defects which are detected in a section having a preset constant length and whose type and grade are determined are counted by type and grade, and the count is performed. The defect grade is re-determined by comparing the number of generated defects with a preset threshold for the number of defects and the number of defects generated for each grade, so it is more equivalent to the visual image of the inspector. The criteria can be used to judge the grade of defects.

Further, the total number of defects generated, whose types and grades have been determined within a predetermined fixed length section, is counted, and the counted number of defects is set to a preset threshold value. When the value exceeds the limit, a function of determining plate stain is provided, so that the problem of erroneous determination of defects due to plate stain and discoloration can be solved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a basic configuration of a steel plate macro determination process applied to an embodiment of the present invention.

FIG. 2 is an explanatory diagram when a macro determination process is applied to a steel plate.

FIG. 3 is a diagram illustrating a specific example of macro determination processing. (a) Defect judgment result before macro processing (b) Macro processing algorithm (c) Defect re-judgment result after macro processing

FIG. 4 is a diagram showing a specific example of macro determination processing. (a) Defect determination result before macro processing (b) Defect re-determination result after macro processing

FIG. 5 is a diagram showing a specific example of macro determination processing. (a) Defect determination result before macro processing (b) Defect re-determination result after macro processing

FIG. 6 is a schematic configuration diagram of an optical inspection apparatus applied to conventional surface defect detection.

FIG. 7 is an explanatory diagram showing a profile method for extracting a defect feature parameter.

[Explanation of symbols]

 1 Macro processing unit 2 Defect occurrence number integration unit 3 Defect occurrence number setting unit 4 Arithmetic processing unit 20 Control unit 22 Arithmetic processing unit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Katsutoshi Ito 1-2-1, Marunouchi, Chiyoda-ku, Tokyo Nihon Steel Pipe Co., Ltd.

Claims (3)

[Claims] 1. A surface defect determination method for a belt-shaped body for determining the type and grade of a surface defect of the belt-shaped inspection body by image-analyzing reflected light from the surface of the belt-shaped inspection body, wherein a predetermined length is set. For all the defects detected in the section of which the types and grades have been determined, the number of occurrences of those defects is counted by type and grade, and the number of each counted defect is set in advance. A method for determining a surface defect of a band-shaped object, wherein when the threshold value for the number of occurrences for each type and grade is exceeded, the defect grade is increased for the defect. 2. A surface defect determination method for a belt-shaped body, wherein the type and grade of the surface defect of the belt-shaped body to be inspected is analyzed by image analysis of reflected light from the surface of the belt-shaped body to be inspected. For all the defects detected in the section of which the types and grades have been determined, the number of occurrences of those defects is counted by type and grade, and the number of each counted defect is set in advance. A method for determining a surface defect of a strip, which comprises reducing the defect grade of the defect when the threshold value for the number of occurrences for each type and grade is not exceeded. 3. A surface defect determination method for a belt-shaped body, wherein the type and grade of the surface defect of the belt-shaped body to be inspected is analyzed by image analysis of reflected light from the surface of the belt-shaped body to be inspected. When the number of all the defects detected in the section of which the types and grades are determined exceeds a preset threshold value, the defect type is changed, and the surface defect determination method of the band-shaped body is characterized. .