JPH04145351A - Defectiveness inspection device - Google Patents

Abstract

PURPOSE:To improve defectiveness detecting precision, by conducting inspection by using a different inspection condition set at every lane or block. CONSTITUTION:At a memory means 60, optimum inspection condition sets used at every lane, I, II,....VI are memorized. A scanning lane discrimination means 64 discriminates which lane a scanning position on a main scanning line 32 is in. A selecting means 66, when a lane signal beta is 'I', selects an inspec tion condition set I opposing this lane I from the means 60. Also, when the signal beta becomes 'II', the means 60 selects an inspection condition II. Thus, inspection conditions I, II - VI to be selected are changed in opposition to the scanning position being changed to respective lanes I, II - V. A changeover means 68 delivers to respective portions inspection conditions selected by means of the means 66 by conducting reading from the means 60.

Description

Detailed Description of the Invention (Industrial Application Field) The present invention can be obtained by scanning the surface of a rolled or long inspection object such as a steel plate, aluminum plate, copper plate, paper, or plastic film. The present invention relates to a defect inspection device that inspects the presence or absence of defects using image signals.

(Technical Background of the Invention) A defect inspection device is known that optically scans the surface of a wafer board, paper, or the like to detect scratches on the surface or internal defects. Conventionally, the image signal obtained by scanning the inspection object is compared with a preset threshold value, and if the image signal falls outside the range determined by this threshold value, it is determined that there is a defect.

On the other hand, in the case of hot-rolled steel coils, plastic films, etc., or long objects to be inspected, the surface texture in the width direction usually changes slightly depending on processing conditions such as rolling conditions. Normally, the ground noise is larger near both edges than near the center in the width direction, and the degree of waving and roughness is also thicker.Furthermore, depending on the rolling conditions, periodic scratches may appear at certain positions in the width direction. Sometimes.

Further, in order to improve the accuracy of defect discrimination, image signals are usually subjected to various types of signal processing. For example, filtering processing using a differential filter or density changing processing that changes the density distribution range of the image signal is performed. The optimum conditions for such signal processing also change depending on the ground noise, undulations, and roughness of the surface of the object to be inspected.

Even in such cases, conventional equipment requires a certain threshold value and other inspection conditions (hereinafter simply referred to as inspection conditions) as described above.
Since defects were detected using , if appropriate inspection conditions were used for the area near the center of the object to be inspected, ground noises that occur frequently near both edges would be judged as defects. On the other hand, if appropriate inspection conditions were used near both edges, there was a problem in that inspections would be omitted near the center.

In addition, if a scratch appears at an approximately fixed position in the width direction or at a known periodic location, it may be planned in advance to cut off and discard the area around this area. Although there is no need to detect this defect, the conventional device faithfully detects this defect, which is rather troublesome.

(Objective of the Invention) The present invention has been made in view of the above circumstances, and is capable of inspecting a test object without being affected by ground noise or surface undulation or roughness that occurs near a fixed position in the width direction of the object to be inspected. It is an object of the present invention to provide a defect inspection device that can detect defects with high precision and, if the range to be cut off is known in advance, can prevent defects in this range from being detected.

(Structure of the Invention) According to the present invention, this object is to provide a defect inspection device that detects defects in the inspection target by comparing an image signal obtained by scanning the inspection target with predetermined inspection conditions. scanning lane discrimination means for dividing the inspection area into a plurality of long lanes in the feeding direction and determining the lane in which the scanning position is placed; memory means for separately storing inspection conditions corresponding to each lane; This is achieved by a defect inspection apparatus characterized in that the defect inspection apparatus includes a selection means for reading inspection conditions corresponding to the above from the memory means, and detects defects using the selected inspection conditions.

Here, a plurality of test conditions corresponding to each lane stored in the memory means may be stored corresponding to at least one of the type of test object and the test state. In addition, inspection conditions may be set so that no defects are detected for at least one lane.

Instead of lanes, the inspection area to be inspected is divided into a plurality of blocks partitioned in the width direction and the feed direction, and the inspection conditions corresponding to each block are separately stored in the memory, and the block being scanned is divided into blocks. It is also possible to read out test conditions corresponding to the above from the memory means.

(Function) The scanning lane (
block) is determined by the determining means. When the lane (block) containing the scanning position is determined, test conditions such as threshold values corresponding to this lane (block) are read out from the memory means. Using these inspection conditions, the presence or absence of defects in the corresponding lane (block) is inspected.

(Embodiment) FIG. 1 is a block diagram of a first embodiment of the present invention, FIG. 2 is a diagram showing an example of a set of test conditions stored in a memory means,
3 to 5 are explanatory diagrams of signal processing. That is, FIG. 3 is an explanatory diagram of spatial filtering processing, FIG. 4 is a diagram illustrating an example of a differential filter, and FIG. 5 is an explanatory diagram of density conversion characteristics of an image signal. FIG. 6 is a diagram showing an example of dividing lanes or blocks to be inspected.

In FIG. 1, reference numeral 1o indicates an object to be inspected, such as a steel plate, paper, or plastic film, and this object 10 to be inspected is sent from a supply roll 12 to a take-up roll 14. This take-up roll 14 is driven by a take-up motor 16. The traveling speed (feeding speed) of the test object 10 being wound up by the motor 16 is detected by a speed detector (pulse generator) 18 . While the inspection object 10 is being transported, an image on the surface thereof is read by an image detection means 20 using a flying spot method. The image detection means 20 scans a scanning beam ρ made of laser light emitted from a laser light source 22 using a rotating mirror (a polygonal mirror) rotated by a motor 24.
While the inspection object 10 is scanned at a constant speed (main scanning) in the width direction by the mirror 26, the light reflected by the surface of the inspection object 10 is sent to a pair of light receivers 30 by the light receiving rod 28.
(30a, 30b) and receives the light. That is, the light receiving rod 28 is connected to the main scanning line 32 of the scanning beam β.
When the reflected light is received, it is totally reflected on the inner surface of the light receiving rod 28 and guided to both ends thereof.
Photoreceiver 3 such as a photomultiplier (photomultiplier tube)
0, the amount of received light is detected. The image signal output from each light receiver 30 is amplified by a preamplifier and a main amplifier, and is also waveform-shaped to form an analog image signal atxa shown in FIG.
* becomes.

The level of each signal a I s a 2 decreases as it moves away from the scanning position on the main scanning line 32 of the scanning beam β, and conversely, the level increases as it approaches the scanning position. Therefore, in this embodiment, , both signals a + , a 2
are added by the adding means 34, and the influence of the change in scanning position on the main scanning line 32 is removed, resulting in a signal a3.

The output adjustment means 36 adjusts the gain and offset so that the image signal a3 has a desired amplitude and output level. Inspection conditions such as gain g, offset of, and other data used in signal processing by this output adjustment means 36 are input from memory means 60, which will be described later.

This output-adjusted signal a4 is converted into a digital signal a in the A/D conversion means 38. For example, it is converted into a 256-level density signal a5. and bulk memory 40
is stored in memory. Here, data such as the number of gradations for A/D conversion are input from memory means 60, which will be described later.

This bulk memory 40 has a storage capacity equal to the product (N The digital image signal a5 is stored in . That is, it has a ring buffer structure in which when the storage area of the bulk memory 40 becomes full, the oldest data is sequentially overwritten. Here, the inspection width gate signal α indicates the scanning width on the main scanning line 32, and is detected by a pair of optical sensors 35. It is produced by the outputs α1 and α2 of (35a, 35b). This gate signal α becomes "1°" when the optical sensor 35a detects the scanning beam β, and the optical sensor 35a
When 5b detects it, it returns to "0°" and eventually becomes "1°" within the inspection width in the main scan.

Further, this gate signal α can also be generated by detecting both edges of the object to be inspected by signal processing from the image signal a4.

On the other hand, the signal a4 output-adjusted by the output adjustment means 36 is subjected to signal processing in the signal processing means 42. In this case, analog signal processing is performed, but the output a of the A/D conversion means 38 is
, may be digital signal processing in which the signal processing means 42 processes the signals.

This means 42 performs filtering processing using, for example, a differential filter. To explain this differential processing using the digital signal processing method described above, for example, as shown in FIG. Find the product of these and use their sum as the output value. This operation is performed in rask scanning order from the upper left pixel to the lower right pixel.

As a spatial filter used in this filtering process, a filter having weighting coefficients shown in FIG. 4 can be employed. This uses two filters: a filter Δ, f for row direction differentiation and a filter Δ, f for column direction differentiation.
The final output value is the sum of the absolute values of the outputs of each filter, or the larger value of the outputs of both filters.

Here, the action of the differential filter has been explained using an example of digital signal processing, but in the case of analog signal processing, the image signal for each scan may be differentially processed using a known analog differential circuit.

As a result of such filtering processing, periodic fluctuations (low frequency components) included in the signal a3 due to changes in the optical path length of the scanning beam ρ, changes in the angle of incidence, changes in the texture of the surface to be inspected, etc. are removed, and the image The contour in at least one direction of χ and y directions is emphasized, and subsequent defect detection accuracy is improved.

The signal processing means 42 may perform density conversion in addition to or in place of this differential processing. The conversion table used for this density conversion has the characteristics shown in FIG. 5, for example. In FIG. 5, the horizontal axis represents the density of the input density signal y, and the vertical axis represents the output density signal Y after density conversion. These density signals y1Y have, for example, 256 gradations. In addition, this table assumes a constant density (for example, an intermediate density of 127) within a certain density width a on both sides of the background density C, and outside this width a, the maximum (255) and minimum density (0) are converted with a slope b. It is something to do. This density conversion can eliminate noise in the background area and improve defect detection accuracy. Note that the set values a, b, and c used for this density conversion are stored in a memory means 60 to be described later.
Input from

In the signal y that has been subjected to the filtering process using the differential filter as described above, the low frequency components present in the signal a4 are removed, the defective portion is emphasized, and the background noise is erased by the density conversion.

The defect detection means 44 compares this signal a6 with a predetermined threshold value (or range) υ, and determines a defect when the signal a6 is outside the range determined by this threshold value. The signal 2 indicating this defect is sent to the address determining means 46, and the address Ad of this defect is determined.

Threshold value (or range) used in this defect detection means 44
υ is sequentially read out from the memory means 6o (described later) as the scanning position on the main scanning line 32, that is, the lane in the main scanning direction changes as will be described later, and is set to different threshold values υ1, υ2, etc. for each lane. is used to inspect for defects.

The address Ad is obtained from the operation control section 48. That is, the operation control unit 48 drives the motor 16 so that the feed speed of the inspection object 10 matches the command speed (7) inputted from the memory means 60 (described later). Also, the scanning beam β
The motor 24 is driven so that the scanning speed in the main scanning direction matches the command speed Vx similarly input from the memory means 60. The operation control unit 48 counts the speed pulses from the speed detector 18 to determine the Y coordinate in the sub-scanning direction, and also calculates the X coordinate in the main scanning direction from the rotational speed of the motor 24 and the elapsed time based on the inspection width gate signal. These coordinates (X, Y) are output to the address determining means 46 as the coordinates of the scanning position.

When the address Ad of the defect is determined in this manner, the bulk memory 40 transfers the area containing this address Ad, that is, the area containing the defect, to the frame memory 50. This frame memory 50 contains the signal a5 as well as the coordinates of the defect (X
, Y) may be stored simultaneously. .

The contents of the frame memory 50 are further subjected to signal processing in a signal processing means 52 as necessary. For example, the defective portion is enlarged or reduced to a magnification (Xn) inputted from the memory means 60 described later. And this image is CR
It is displayed on a monitor 54 such as T. The address Ad of the defect may be displayed here together with the image of the defect. This image and address Ad can be stored on a video, magneto-optical disk, or
It may also be recorded on a floppy disk, hard disk, memory tape, etc.

On the other hand, when the address determining means 46 determines the defective address Ad, the signal processing means 56 then performs classification processing, for example. This is to classify defects by determining the degree and type of the defect such as the size and spread of the defect determined by the defect detection means 44. This classification result is output to an output means 58 such as a printer or an external process computer connected via a data line. Data used for this classification process is input from memory means 60, which will be described later.

The data used for the above processing, that is, the inspection conditions are supplied from the memory means 60. As shown in FIG. 2, this memory means 60 stores a set of optimum inspection conditions to be used for each lane I, . In other words, the inspection object 10 is divided into six lanes, for example, as shown in FIG. do.

In this example, the inspection conditions for each lane are as follows:
For example, steel plate A, steel plate B, etc. are stored in memory corresponding to the type of inspection target. Here, different test conditions may be stored in memory depending on the test state of the test object 10, for example, the state where the surface is dry, oily, colored, etc. The test conditions to be memorized include the threshold value, the gain g used in the output adjustment means 36, and the offset 0. , the number of gradations k used in the A/D conversion means 38, the filter Δ used in the signal processing means 42,
f, Δyf, density conversion setting values a, b, c, scanning speed V, , V, , magnification (Xn), classification processing data, etc.

These setting values are input manually or automatically from the condition input means 62. For example, while viewing the image on the monitor 54 or the results printed on the output means 58, the inspector manually inputs preferred setting values for each inspection condition of the inspection object 10 using a keyboard or the like. Further, some setting values may be automatically determined. For example, the gain g and the offset of can be automatically set so that the output level of the output adjustment means 36 falls within a predetermined range.

Reference numeral 64 denotes a scanning lane determining means, which determines in which lane the scanning position 1 on the main scanning line 32 is located. That is, the inspection object 1o is virtually divided into a plurality of lanes, ■, ~■ in the width direction as shown in FIG. This is determined based on the elapsed time since then, and it is determined which lane the determined scanning position is in. Then, a lane signal β indicating the lane currently being scanned is sent to the selection means 66.

The selection means 66 selects the test condition set I corresponding to the lane I from the memory means 60 when the lane signal β is "■". Further, when the lane signal β becomes °゛■°゛, the memory means 60 selects the test condition ■. In this way, as the scanning position changes from lane E to U to ■, the inspection conditions to be selected are changed from ■ to ■. The switching means 68 reads out the inspection conditions selected by the selection means 66 from the memory means 60 and sends them to each section.

In addition, if the types and inspection conditions of the 10 inspection targets change, a group of inspection conditions for each lane corresponding to the changes is selected manually or automatically, and the same operation is performed using the inspection conditions of each lane in that group. .

In the embodiments described above, groups of different inspection conditions are also stored in memory in response to changes in the type of inspection target or the inspection status. However, in the present invention, each lane is It may also be possible to store only the inspection conditions of.

In addition, each lane is divided by a predetermined length in the feeding direction of the inspection object 10, and each lane is divided into blocks; On the other hand, the memory means 60 may store test conditions corresponding to each block.

In this case, the lane discrimination means 64 discriminates the block based on the inspection width gate signal α and a signal indicating the feed amount of the inspection object 10 (obtained from the operation control means 48 or the like). During scanning, the corresponding test conditions are read out from the memory means 60 for each block and tested.

In some lanes or blocks, inspection conditions such as the threshold value υ may be set so that no defects are detected.

The above embodiment uses a flying spot method in which the laser beam is reflected on the surface of the inspection target, but the inspection target is irradiated with a rod-shaped light source arranged in the width direction, and the reflected light is sent to the receiver via a rotating mirror. It may be a flying image type reading method, or a type reading a line sensor or an area sensor.

Although the above embodiments have been described as detecting defects appearing on the surface of the object to be inspected, the present invention also includes detecting defects inside by scanning the surface.

For example, internal defects in a steel plate may be detected using a magneto-optical effect. This detects the leakage magnetic field from a defect when the inspection object is magnetized with an alternating magnetic field as a change in the polarization of reflected light.

(Effects of the Invention) As described above, the present invention divides the inspection area to be inspected into a plurality of lanes long in the feed direction or a plurality of blocks partitioned in the width direction and the feed direction, and reads out inspection conditions corresponding to changes in the scanning position from a memory means in which appropriate inspection conditions are stored in advance;
Inspection is performed using different inspection conditions for each lane or block. For this reason, due to formation noise, waving, roughness, etc. that appear at approximately fixed positions in the width direction due to the rolling process, etc., the optimal inspection conditions for detecting defects may differ for each lane or block. Also, each lane can be inspected using the optimum 4° inspection conditions. Therefore, the defect detection accuracy is improved (claims (1) and (4)).

In addition, prepare multiple inspection conditions for each lane or block depending on the type of inspection target and inspection condition.
It is also possible to make it possible to change the inspection conditions in response to changes in the type of inspection target or inspection status. Section (2)).

Furthermore, inspection conditions can be set so that defects are not detected for some lanes or blocks. In this case, defects should not be detected in areas that include flaws that appear periodically near a predetermined position in the width direction or at predetermined intervals, and which will be cut off and discarded later. This eliminates the need to detect defects unnecessarily (Claim (3)).

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a first embodiment of the present invention, FIG. 2 is a diagram showing an example of a set of test conditions stored in a memory means,
FIG. 3 is an explanatory diagram of spatial filtering processing, FIG. 4 is a diagram illustrating an example of a differential filter, and FIG. 5 is an explanatory diagram of density conversion characteristics of an image signal. FIG. 6 is a diagram showing an example of dividing lanes or blocks to be inspected. 10... Inspection object, 20... Image detection means, 60
...Memory means, 62...Inspection condition input means, 64
...Scanning lane (block) discrimination means, 66...Selection means.

Claims (4)

[Claims] (1) In a defect inspection device that detects defects in the inspection target by comparing an image signal obtained by scanning the inspection target with predetermined inspection conditions, the inspection area of the inspection target is divided into multiple lanes long in the feeding direction. scanning lane discriminating means for determining the lane in which the scanning position is placed; memory means for separately storing inspection conditions corresponding to each lane; and scanning lane discriminating means for separately storing inspection conditions corresponding to each lane; and a selection means for
A defect inspection apparatus characterized in that a defect is detected using the selected inspection conditions. (2) The defect inspection apparatus according to claim 1, wherein a plurality of inspection conditions corresponding to each lane stored in the memory means are stored in memory corresponding to at least one of a type of an inspection object and an inspection state. (3) The defect inspection apparatus according to claim (1), wherein inspection conditions are set so that no defects are detected for at least one lane. (4) In a defect inspection device that detects defects in the inspection target by comparing an image signal obtained by scanning the inspection target with predetermined inspection conditions, the inspection area of the inspection target is partitioned in the width direction and the feed direction. scanning block discriminating means for dividing the block into a plurality of blocks and discriminating the block in which the scanning position is placed; and memory means for separately storing inspection conditions corresponding to each block;
What is claimed is: 1. A defect inspection apparatus comprising: selection means for reading out inspection conditions corresponding to a block being scanned from said memory means, and detecting defects using said selected inspection conditions.