JPS61256242A - Method and apparatus for inspecting flaw - Google Patents

Abstract

PURPOSE:To extract a flaw signal with good sensitivity even when the level of a detection signal varies gently and largely as compared with the change of a flaw signal and a noise component is further contained, by calculating the change quantity of an image signal within a definite time and judging a flaw if the absolute value thereof reaches a threshold value or more. CONSTITUTION:A magnetic disc 1 is mounted to a spindle 3 and a photoelectric element 6 is scanned by drive motors 4, 5. The detection signal from the element 6 is amplified by an amplifier 9 to be inputted to an A/D converter 10. Signal processors 16, 17 perform the judgement and renewal of the max. value and min. value of the detection signal and the operation of the difference between both values every time when the signal from the A/D converter 10 is inputted. Then, the absolute value of the change quantity of the detection signals within a gate signal time having a width equal to the size of a flaw generated in a mutually overlapped state through a frequency dividing circuit, an odd circuit 14 and an even circuit 15 is compared with the threshold value set by a thresh old value setting device 18 and, only when said absolute value is larger than the threshold value, the absolute value of the signal change quantity is written in a memory 20 by a writing order 30.

Description

[Detailed Description of the Invention] [Field of Application of the Invention] The present invention relates to a defect inspection method and apparatus for inspecting defects such as scratches and protrusions on a sample to be inspected, particularly a magnetic disk recording film, by imaging them with an imaging device. Regarding.

[Background of the invention]

When inspecting defects such as scratches and protrusions on a specimen to be inspected, such as a steel plate or a paint film on a metal disk, the specimen to be inspected is imaged with a television camera, etc., and the obtained video signal is converted to a threshold value of 2.
Defects such as scratches and protrusions are recognized by converting them into values. When the state of the inspection surface is stable, the detection signal from the normal surface is constant, so it is possible to extract defects by binarizing with a fixed threshold value. However, when the condition of the inspection surface is unstable, for example when the coating thickness of the magnetic disk coating surface changes, the detection signal S changes as shown in Figure 5, and defects are recognized using a fixed threshold. It is impossible to do so.

As a method to deal with this, for example, JP-A-54-56
There is a floating binarization method as shown in No. 58. This is done by reducing and smoothing the detection signal S, and then adding and subtracting a certain level, which is then binarized using a threshold value. According to this, patterns and minute defects can be detected, for example, in the case of defect inspection of printed circuit boards. However, in the case of the detection signal S shown in FIG. 5, when attempting to discriminate between minute defects I to H, even changes in the background are binarized, making it impossible to detect defects. In this case, as shown in FIG. 5, the reduced smoothed signal L1. By reducing L2 and adjusting the fixed level, defects A9 and 2 can be distinguished, but defects C and E cannot be detected. Threshold value L is required to detect defects C and H.
1', (L2#) must be as shown in FIG. In this case, in addition to the binarization of C and E, the R scene is also binarized, resulting in false detection.

In addition, since the floating dark value method generally utilizes the dynamic response of an electric circuit, it can be , the electrical time constant must be changed. This takes time.

[Purpose of the invention]

In view of the above-mentioned conventional problems, it is an object of the present invention to detect a detection signal when the level of the detection signal fluctuates more slowly and greatly than the change in the defect signal, and further includes a noise component that changes at the same period as the defect signal. Another object of the present invention is to provide a defect inspection method and an apparatus therefor, which are capable of extracting defect signals with high sensitivity.

[Summary of the invention]

In the present invention, when the imaging device scans and images the sample to be inspected relative to the sample, the defect signal obtained based on the defect has an indeterminate magnitude, but the amount of change within a unit time is compared to the change due to the background. Focusing on the fact that the value becomes extremely large, the absolute value of the difference between the maximum and minimum values of the video signal within a certain period of time, that is, the amount of change, is determined, and if the absolute value exceeds a certain threshold, it is determined as a defect. It is characterized in that a defect signal is sensitively detected from an image signal containing a noise component.

[Embodiments of the invention]

Embodiments of the present invention will be specifically described below with reference to FIGS. 1 to 4, taking magnetic disk coating inspection as an example.

First, the defect inspection method according to the present invention will be explained with reference to FIG. In FIG. 4, (A) shows a detection signal S including defect signals A to H, which is an example in which a change in reflectance of a sample to be inspected, such as a coated surface of a magnetic disk, is detected by a photoelectric element.

The detection signal S fluctuates more slowly and greatly than the changes in the defect signals I to H, and further includes a noise component that changes at a period comparable to that of the defect signals. These fluctuations in the detected video signal are caused by uneven coating of the paint film, fluctuations in the brightness of the illumination light, surface roughness of the paint film surface, and the like. (B) shows a detection signal S in which the periphery of defect signal 2 in (A) is expanded over a range of time t, and shows how defect signal 2 changes over an hour. In (B), minute fluctuations in the detected video signal S other than defect signal 2 are noise components. (
C) A gate signal having a gate width of ttx time W is scanned in the order of C, C4, L, and tn. It is important that the gate signals that are generated one after another overlap with each other to avoid overlooking defective signals.

In the example of (C), the absolute value of the signal change amount of the detected video signal S within the gate width W time is extracted every step of the way (thereby, the defect signals I to H are The key point of the present invention is to easily extract the signal.That is, by making the gate width W equal to the change time T of the defect signal, the gate width W can be easily extracted without being affected by gradual and large fluctuations in the detected video signal This means that only the absolute value of the amount of signal change within the time width W can be accurately extracted.

(D) shows the absolute value SL of the signal change amount within the gate width W time of the detected video signal S corresponding to the gate signal '*/m' and '*Q, and the defect signal The absolute value of the amount of change exceeding the threshold value in the gate signal of 4 and i when scanning the second part is 84. Based on the above three principles, the level of the detected video signal is determined by the change in the defect signal. It is now possible to extract only the defect signal with high sensitivity (and stability) even when it contains a noise component that fluctuates slowly and greatly compared to the defect signal and also changes at a period comparable to that of the defect signal.

1 to 6 illustrate an embodiment of the visual inspection apparatus according to the present invention, which scans the surface of a sample to be inspected, such as a magnetic disk coating surface, in a yJA spiral shape and detects defects using a photoelectric element.

FIG. 1 shows the overall configuration of this embodiment, in which 1 is an i-disk, 2 is a straight table, 5 is a varnished binder, 4, 5
is a motor, 6. is a photoelectric element, 7 is a linear encoder, 8
is a rotary encoder, 9 is an amplifier, 10 is an A/D converter, 11.12 is a counter, 15 is a frequency dividing circuit, 14f'
E, 0ctd circuit, 15 k'L evan circuit, 1
6.17 is a signal processor, 18 is a threshold value setter, 19 is an OR
The circuit 20 is a memory.

Defect detection is carried out by attaching the magnetic disk 1 to the spindle 5 provided on the @adc table 2.1. The photoelectric element 6 is scanned in a spiral manner by the drive motors 4 and 5 (Fig. 2).
Through this detection operation, light is transmitted from the element 6 to the amplifier 9 as shown in FIG. , spindle 5 from rotary encoder 8
(Magnetic disk 1) The elongation coordinates are output to the counter 12. FIG. 4(B) is an enlarged view of the second part shown in FIG. 4(A). The detection signal from the photoelectric element 6 is amplified by an amplifier 9 and then input to an A/D converter 10. A/
In the D converter 1o, the output signal from the rotary encoder 8 (FIG. 4(E)) is used as a clock, and the detected signal is converted into //Df and outputted to the signal processor 16 and the signal processor 17. Each time the EOC signal (FIG. 4 (F)) from the A/D converter 10 is input, the signal processors 16 and 17 judge and update the maximum value and minimum value of the detected signal, and calculate the difference. Time i! l115, ocLcL circuit 1
4. Compare the absolute value of the amount of change in the detected signal within the gate signal time created via the ay-leur circuit 15 with the dark value set by the threshold value setter 18, and only if it is larger than the threshold value, write it to the memory 20 by the blue filling command 50. Add the absolute value of the amount of signal change to . At this time, R15g from counters 11 and 12
, θ coordinates are used as defect position information. Frequency dividing circuit 15
The output signal from the odd circuit 14 is shown in FIG. 4 (G), and the output signal from the odd circuit 14 is shown in FIG. 4 (H). - indicates the output signal from the circuit 15 in FIG. 4(I). These are shown in Figure 4 (C
) is a signal for creating the gate signal shown in ocL.
The gate signals of A, m, shown in FIG. 4(C) are controlled by the d signal, and the gate signals of j, shown in FIG. 4(C) are controlled by the ttpal signal.
Creates a gate signal for In this embodiment, the period of the branch clock is 4, the period of the Od(signal is W), the period of the #?$3 signal is W, and the phase difference between the odCL signal and the 1?-C signal is I. Fig. 4 (F) shows the EOe signal input from the A/ID converter 1G.
This is a signal output every time /D conversion is completed, as shown in Figure 4 (F
) The output is delayed by Δt from the encoder output indicated by K.

The configuration and operation of the signal processors 16 and 17 will be explained in detail with reference to FIG. Since the signal processor 16 and the signal processor 17 are completely the same in structure and operation except that they operate with a T period difference in gate signal, only the signal processor 16 will be explained in terms of degrees Celsius.

In the figure, 21.22 are latch circuits, 25.24°2
6 is a comparator, 25 is arithmetic unit, 27.28.29 & engineering AN
This is the D circuit.

The A//D converted detection signal (hereinafter referred to as A7. signal) is sent to the latch circuits 21 and 22 and the comparator 25°24i.
C is input. Latch circuit 216ヱ, ocLtL circuit 1
Every time the add signal (Figure 4 (H)) is output from 4,
The latch is cleared to zero and the maximum value is set every time the latch circuit 221 outputs the adct signal (FIG. 4(H)). 25 comparators The data (MAX) latched in the error latch circuit 21 and the A7.
Constantly compare signal data, A7. When the signal data is large (A/D>MAX), the comparator 25 outputs a signal @1", and
fL (N) is ANDed, this data is latched in the latch circuit 21, and updated. In the converter 24, the data (MXN) latched in the latch circuit 22 and the A' output from the A'D converter 10 are , /6 signal data are constantly compared, and if A7. signal data is small (A'D (MIN
), a signal of "1" is output from the comparator 24, and the AN
The D gate 28 performs an AND operation with the EOe signal "1", and this data is latched into the latch circuit 22 and updated.

That is, these latches are performed when the logical product of the latch signals of the comparators 25 and 24 and the EOC signal of the transformer 10 (which is performed by the AND circuit 2728) becomes @1''.

As a result of the above operation, the MAX value of the odd M signal (A7. signal after the latch is cleared to zero by FIG. Figure 4 (H) The MIN value of the A filter signal after setting the maximum value in the latch is output from K. The hw value of this A7. signal and the iI threshold are input to the calculator 25, and the next odd The difference between the MIN value and the MIN value of the signal that continues until the signal (Fig. In the comparator 26, the calculation unit 25
The absolute value of the amount of signal change within the gate width outputted from the calculator 25 is compared with the threshold set by the threshold value setter 18, and if the absolute value of the amount of signal change outputted from the arithmetic unit 25 is large (MAX -
AND gate 2
9 and outputs it, ANDs it with the "1" signal of od and d-, and stores it in the memory 2o via the OR circuit 19. At this time, the R-seated image and θ coordinate of the detection position are also taken in from the counters 11 and 12 of the linear encoder 7 and rotary encoder 8, and used as defect position information. memory 2
Data is written to O by an AND circuit 29 using the input signal from the comparator 26 and the odd signal from the odd circuit 14.
The logical AND of the d signal (FIG. 4(H)) is performed by the write command 5o.

In the above explanation, only the signal processor 16 that performs signal processing based on the add signal has been explained, but in this embodiment, as shown in FIG. The difference between the signal processor 16 and the signal processor 17, in which the signal processor 17 is installed, is as described above (odd
, ct multiplied signal (Fig. 4 (H)) and gygn signal (Fig. 4 (H))
1) The phase of (1)) is shifted by seven periods. this is,
The gate signals are overlapped with each other to prevent defective signals from being overlooked, and the amount of overlap is the + of gate @W.
Anything other than that is fine.

The point is at yourself, j? - It is sufficient that the processing of the files alternates in time.

As described above, according to the present embodiment, since the above-mentioned defect signal extraction is realized using only a simple circuit configuration, in addition to the above-described effects, the defect detection speed can be increased.

In the present embodiment, the inspection method has been explained using a magnetic disk as an example, but it is obvious that the present invention can be applied to inspection of samples other than 8K disks.

Further, in this embodiment, the α and MIN values of the detected video signal are calculated using A/DK conversion, but they can also be performed using an analog method such as peak hold. Further, in this embodiment, the gate is generated based on the encoder clock, but it is not necessarily necessary to use this. That is, when a motor with a constant rotational speed is used, even if A/D conversion, gate generation, etc. are performed asynchronously or locked by a crystal oscillator or the like, no problems occur. This will result in a cheaper configuration, but the defect position information will be more inaccurate than in the embodiment.

〔Effect of the invention〕

As explained above, according to the present invention, gate signals having a width equivalent to the size of a defect are generated in real time so as to overlap with each other, and the absolute value of the amount of change in the detection signal within the gate width is When the set threshold is exceeded, it is determined to be a defect signal, so the level of the video signal obtained from the imaging device fluctuates more slowly and greatly compared to the change in the defect signal, and furthermore, the level of the video signal obtained from the imaging device fluctuates more slowly and with a period similar to that of the defect signal. Even when changing noise components are included, defect signals can be extracted with high sensitivity without being overlooked.Furthermore, according to the present invention, defects can be inspected with high sensitivity using a simple circuit configuration.

[Brief explanation of drawings]

FIG. 1 is an overall configuration diagram showing an embodiment of the defect inspection method according to the present invention, FIG. 2 is a plan view for explaining the method of scanning a sample to be inspected, and FIG. 5 is a signal diagram of the signal shown in FIG. 1. FIG. 4 is a diagram for explaining the internal structure and operation of the processor, FIG. 4 is a diagram for explaining the present invention in detail, and FIGS. 5 and 6 are signal waveform diagrams for explaining the conventional example. 1...-Ki disk, 2... Straight table, 5...
- Spindle, 4, 5...Motor, 6...Optical straight tube, 7...Linear encoder, 8...Rotary encoder, 9...Amplifier, 10...A7. converter, 11
, 12... Counter, 15... Frequency divider circuit, 14...
・ocht circuit, 15...at6 circuit, 16.17
... Signal processor, 18... Threshold value setter, 19...
OR circuit, 20...Memory, 21.22...Latch circuit, 25j24.26...Comparator, 25...Arithmetic unit, 27, 28, 29...AND circuit, 60...
・Soaking command. Fig. 2 Fig. 7 Fig. 3 EOCO L, )υ Fig.

Claims (1)

[Claims] 1. The inspection sample and the imaging device are scanned relative to each other, gate signals having a width equivalent to the size of the defect are generated in an overlapping manner, and the above-mentioned imaging is performed within the gate signal time. A defect inspection method characterized in that a defect is determined when a difference between a maximum value and a minimum value of a video signal obtained from a device exceeds a set threshold value. 2. An imaging device that scans and images relative to the sample to be inspected, a gate generation circuit that generates gate signals by overlapping them, and a video signal obtained from the imaging device within each gate generated by the gate generation circuit. The present invention is characterized by being provided with an arithmetic circuit that calculates a difference signal between the maximum value and the minimum value of , and a comparison circuit that compares the difference signal calculated by the arithmetic circuit with a preset threshold value and determines it as a defect. Defect inspection equipment.