JPH0843077A - Surface inspection method and apparatus - Google Patents

Abstract

PURPOSE:To objectively and quantitatively measure such surface characteristics as scratch and roughness. CONSTITUTION:The surface of an object 4 to be measured is scanned by making a small probe 3 into contact with the surface, the force acting on the probe 3 in this case, is detected as each component force of X, Y, and Z by a force detection part 2, and the scratch of the object 4 to be measured is evaluated by the neural network part of a judgment part 10 according to the amount of waveform characteristic of the component forces.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a surface inspection device for detecting minute scratches on an object to be measured and for evaluating surface roughness and smoothness.

[0002]

2. Description of the Related Art When detecting minute scratches on the surface of an object to be measured or when evaluating surface roughness or smoothness,
A skilled worker directly touches the surface of the object to be measured with a finger or the like, or compares a feeling of contact by using a standard product that is facilitated as necessary, and makes a judgment based on the comparative feeling.

In such an inspection by the human sense, it is difficult to make an objective and quantitative judgment even if the comparison with a standard product is made. For this reason, for example, in the inspection of functional parts, it is impossible to determine a strict evaluation standard in association with the performance inspection.

Further, there is a possibility that the judgment may vary due to the difference in the sharpness of the individual's sense, the difference in the physical condition and the mood. On the other hand, there is a stylus type surface profile measuring device that automates surface profile measurement.
This measuring device two-dimensionally scans the surface of the object to be measured with a stylus, and obtains information on the surface shape of the object to be measured from the movement of the stylus at this time.

However, this measuring apparatus can obtain only the information on the surface shape of the object to be measured, and it is difficult to detect the surface roughness and smoothness of the object to be measured.
Therefore, in order to detect surface roughness and smoothness, it is necessary for a person to directly touch the object to be measured with a finger or the like to make a determination. Further, since the probe is two-dimensionally scanned on the surface of the object to be measured, the evaluation of a wide range of the object to be measured is
It takes a lot of time.

[0006]

As described above, it is difficult to make an objective / quantitative judgment by an inspection based on human senses. Further, the measuring device can obtain only the information on the surface shape of the object to be measured, and it is difficult to detect the surface roughness and smoothness of the portion to be measured. Therefore, it is an object of the present invention to provide a surface inspection method and an apparatus therefor capable of objectively and quantitatively measuring surface characteristics such as surface scratches and roughness.

[0007]

According to a first aspect of the present invention, a fine measuring element is brought into contact with the surface of an object to be measured to scan, and the force acting on the measuring element at this time is detected as at least a bidirectional component. The surface inspection method is intended to achieve the above object by detecting the surface of the object to be measured based on the detected directional components.

According to a second aspect of the present invention, a minute probe that scans the surface of the object to be measured in contact with it, a force detecting means for detecting the force acting on the probe as at least two-direction components, and this force. A surface inspecting apparatus, which comprises: a determination unit that detects the surface of the object to be measured based on each direction component detected by the detection unit, and that achieves the above object.

According to the third aspect, the probe is a point contact type, a line contact type, or a contact type depending on the contact form with the surface of the object to be measured.
Mount one of the surface contact types. According to the fourth aspect, a plurality of minute measuring elements that scan the surface of the object to be measured in contact with each other, and a plurality of force detecting means that detect forces acting on these measuring elements as at least two-direction components, respectively.
And a measuring head comprising a base on which the measuring element and the force detecting means are integrated and arranged one-dimensionally or two-dimensionally,
A surface inspecting apparatus, which comprises: a determination unit that detects the surface of the object to be measured based on at least two-direction components at each of the plurality of measurement points detected by the measurement head.

According to the fifth aspect, the determining means has a neural network for determining at least the surface roughness of the object to be measured from the waveform characteristics of at least two-direction components of the force acting on the probe.

According to a sixth aspect of the present invention, the judging means has a neural network for spectrally analyzing at least two-direction components of the force acting on the tracing stylus, and judging at least the roughness of the surface of the object to be measured from the spectral analysis result. ing.

[0012]

According to the first aspect of the present invention, the surface of the object to be measured is detected by detecting the force acting on the surface of the object to be measured as a component in at least two directions when a minute measuring element is brought into contact with the surface of the object to be scanned. By detecting, it is possible to objectively and quantitatively measure surface features such as scratches and roughness on the surface of the object to be measured.

According to a second aspect of the present invention, a fine measuring element is brought into contact with the surface of the object to be measured for scanning, and the force acting on the measuring element at this time is detected by the force detecting means as at least two-direction components, The determination means detects the surface of the object to be measured based on each of the detected directional components.

According to the third aspect, any one of a point contact type, a line contact type, and a surface contact type is attached as the probe according to the contact form with the surface of the object to be measured. Claim 4
According to this, when the measuring head in which a plurality of measuring elements are arranged is scanned in contact with the surface of the object to be measured, at least two-direction components acting on these measuring elements are detected by each force detecting means, and a plurality of components are detected. The surface of the object to be measured is detected based on each direction component at the measurement location.

According to the fifth aspect, the determining means determines at least the roughness of the surface of the object to be measured using the neural network from the waveform characteristics of at least two-direction components of the force acting on the measuring element.

According to the sixth aspect, the judging means spectrally analyzes at least two-direction components of the force acting on the measuring element, and judges the roughness of at least the surface of the object to be measured by the neural network from the result of the spectral analysis.

[0017]

【Example】

(1) Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of a surface inspection device. A probe 3 is attached to the detector support mechanism 1 via a force detector 2.

Of these, as the detection portion support mechanism 1, for example, an arm mechanism or a compliance mechanism is used. or,
One of a point contact type, a line contact type, and a surface contact type (multipoint contact type) is attached to the tracing stylus 3 depending on the contact form with the surface of the DUT 4.

FIG. 2 shows each type of the probe 3. Each of the tracing styluses 3a and 3b is a point contact type, of which the tracing stylus 3a is formed in a conical shape and the tracing stylus 3b is formed in a pyramid shape.

The tracing stylus 3c is a line contact type and is formed in a knife edge shape. Also, each probe 3d-
3f is a surface contact type, and the tracing stylus 3d is formed in a smooth planar shape, the tracing stylus 3e is formed in a sawtooth surface shape, and the tracing stylus 3f is formed in a planar shape with many protrusions.

The stylus 3 does not necessarily have to be a rigid material such as metal or diamond. Particularly, in the case of the surface contact type, in consideration of the contact property with the surface of the object 4 to be measured, an elastic body such as rubber is used. It is best to

The force detector 2 scans the force F acting on the measuring element 3 in the normal direction component (Z component) when the measuring element 3 is scanned in contact with the surface of the object 4 to be measured. It has a function of detecting as a three-component force of a directional component (X component) and a vertical component (Y component).

A computer 6 is connected to the force detector 2 via an amplifier 5. The calculator 6 is provided with respective input units 7 to 9 and determination unit 10 for each XYZ component force.

Of these, the determination unit 10 has a function of evaluating surface features such as minute scratches on the surface of the object 4 to be measured, surface roughness, and smoothness based on the XYZ component forces in each direction. There is.

FIG. 3 is a specific configuration diagram of the determination unit 10. The feature extraction unit 11 has a function of inputting each XYZ component force output input through each of the input units 7 to 9 and extracting the feature amount 11a of the waveform of the component force output.

The neural network unit 12 uses the XYZ
Of the XYZ component force output waveforms extracted by the feature extraction unit 11 and provided with a hierarchical neural network for scratches, surface roughness, smoothness, etc. of the DUT 4 with respect to the feature amounts of the respective component force output waveforms of 11a, information processing is performed based on these waveform feature amounts 11a to detect scratches on the DUT 4, and has a function of evaluating surface features such as surface roughness and smoothness of the DUT 4. There is.

Next, the operation of the device configured as described above will be described. One of the point contact type, the line contact type, and the surface contact type of the probe 3 is attached to the force detection unit 2 depending on the contact form with the surface of the DUT 4.

The probe 3 is scanned on the surface of the DUT 4 while the force detector 2 and the probe 3 are supported by the detector support mechanism 1. A force F acts on the measuring element 3 by the scanning of the measuring element 3 on the surface of the DUT 4.

The force detection unit 2 detects the force F acting on the probe 3 as each component force of XYZ and outputs each electric signal corresponding to these component forces. Each electric signal of each component force of XYZ is amplified to a level that can be easily processed by the amplifier 5 and sent to each input section 7 to 9 of the computer 6.

The feature extraction unit 11 in this computer 6 is
From the XYZ component forces input through the input units 7 to 9, the feature amount 11a of the waveform of the component force output is extracted. Next, the neural network unit 12 receives the characteristic quantities 11a of the waveforms of the XYZ component force outputs extracted by the characteristic extracting section 11, performs information processing based on the characteristic quantities 11a of these waveforms, and scratches the DUT 4. Is detected, and surface characteristics such as roughness and smoothness of the surface of the DUT 4 are evaluated.

As described above, in the first embodiment,
A minute probe 3 is brought into contact with the surface of the object to be measured 4 to scan the force, and the forces acting on the probe 3 at this time are detected as XYZ component forces, and a neural network is obtained from the waveform feature quantities of these component forces. Since the scratches and the like on the DUT 4 are evaluated by using the portion 12, the surface features that are sensuously captured such as scratches, roughness and smoothness on the surface of the DUT 4 can be objectively and quantitatively measured. Can be measured.

Further, since the probe 3 has a construction in which any one of a point contact type, a line contact type, and a surface contact type is mounted, the probe of the optimum type according to the contact form with the surface of the object 4 to be measured. 3 can be selected for highly accurate measurement.

Therefore, the sensory features such as the presence or absence of surface defects, the roughness and the smoothness of the object to be measured 4 can be objectively and quantitatively evaluated without being influenced by the human sense. Further, since the evaluation standard is constant, it can be applied to the inspection of functional parts, and the actual and strict evaluation based on the performance test and the correlation can be performed. (2) Next, a second embodiment of the present invention will be described.

FIG. 4 is a block diagram of the surface inspection apparatus. A plurality of force detectors 21 are two-dimensionally arranged on the elastic base 20, and the force gauge 2 is attached to each of the force detectors 21.
2 is attached to form the measuring head 23.

One of the point contact type, the line contact type, and the surface contact type shown in FIG. 2 is attached to these measuring elements 22 depending on the contact form with the surface of the object to be measured. Similarly to the above, each force detection unit 21 measures the force acting on the measuring element 22 when the measuring element 22 is scanned in contact with the surface of the object to be measured, the Z component, the X component, and the Y component. It has a function of detecting as a three-component force.

A computer 25 is connected to the measuring head 23 via an amplifier 24. The calculator 25 is provided with respective input units 26 to 28 and determination unit 29 for each XYZ component force.

These input units 27 to 28 are used for the respective force detection units 2
Each component force detected by 1 is input respectively. The determining unit 29 has a function of evaluating surface features such as minute scratches on the surface of the object to be measured, surface roughness, and smoothness based on the XYZ component forces in each direction. Similar to the example, it has the functions of the feature extraction unit and the neural network unit.

The feature extraction unit has a function of extracting the feature amount of the waveform of each XYZ component force output detected by each force detection unit 21. The neural network unit receives the characteristic amounts of the waveforms of the XYZ component force outputs of the force detecting units 21 extracted by the characteristic extracting unit, performs information processing based on the characteristic amounts of these waveforms, and scratches the object to be measured. It has a function of detecting the surface roughness and evaluating surface characteristics such as roughness and smoothness of the surface of the object to be measured.

With such a configuration, a plurality of measuring elements 2
2 is scanned on the surface of the object to be measured, these measuring elements 2
A force acts on each of the two. At this time, the plurality of force detection units 21 respectively detect the forces acting on the plurality of tracing stylus 22 in XY directions.
It detects as each component force of Z, and outputs each electric signal according to these component forces.

Each electric signal of each component force of these XYZ is
It is amplified to a level that can be easily processed by the amplifier 24 and sent to each of the input units 26 to 28 of the computer 25. This calculator 2
The feature extraction unit in No. 5 is the XYZ detected by the plurality of force detection units 21 input through the input units 26 to 28.
From each component force of, the feature amount of the waveform of these component force outputs is extracted.

Next, the neural network unit 12 determines the XY for each of the plurality of force detection units 21 extracted by the feature extraction unit.
Receiving the characteristic quantities of the waveforms of the respective component force outputs of Z, performing information processing based on the characteristic quantities of these waveforms to detect flaws on the object to be measured,
In addition, surface characteristics such as roughness and smoothness of the surface of the object to be measured are evaluated.

As described above, in the second embodiment,
Since the measuring head 23 having a plurality of measuring elements 22 arranged thereon is scanned on the surface of the object to be measured, at the same time, scratches on the surface of the object to be measured, scratches and roughness on the surface of the object to be measured, smoothness, etc. It is possible to objectively and quantitatively measure the surface features that have been captured sensuously, and to perform high-speed measurement of surface features.

The plural measuring elements 22 arranged on the measuring head 23 can achieve the same effect even if they are arranged one-dimensionally. (3) Next, a third embodiment of the present invention will be described. The same parts as those in FIG. 3 are designated by the same reference numerals, and detailed description thereof will be omitted.

FIG. 5 is a block diagram of a computer in the surface inspection apparatus. The spectrum analysis unit 30 performs a spectrum analysis on each component force of XYZ acting on the tracing stylus 3 input through each input unit 7 to 9, and uses the result of this spectrum analysis as an input to the neural network unit 12 for learning. It has a function.

The neural network unit 12 has a function of inputting the result of spectrum analysis and learning the human evaluation result as a teacher signal. With such a configuration, when the probe 3 is scanned on the surface of the DUT 4, a force acts on the probe 3, and this force is applied to the force detector 2.
Is detected as each component force of XYZ, and each electric signal corresponding to these component forces is sent to each input section 7 to 9 of the computer.

In this computer, the feature extraction unit 11 extracts the feature amount 11a of the waveform of the component force output from each of the XYZ component forces input through the input units 7 to 9,
Further, the neural network unit 12 performs information processing based on the characteristic amount 11a of the waveform of each component force output of XYZ to detect the scratches on the DUT 4, and the surface of the DUT 4 such as roughness and smoothness. Evaluate the features.

On the other hand, the spectrum analysis unit 30 performs a spectrum analysis on each component force of XYZ acting on the tracing stylus 3 input through each of the input units 7 to 9. FIG. 6 shows the result of such spectrum analysis, in which the size of the quadrangle indicates the size of the signal at the corresponding time and frequency.

The spectrum analysis unit 30 uses the neural network unit 1 for learning the result of the spectrum analysis.
Input to 2. As a result, the neural network unit 12 learns by using the spectrum analysis result as an input and the human evaluation result as a teacher signal, and changes the weighting coefficient of the hierarchical neural network, for example.

As described above, according to the third embodiment, in addition to the effects of the first embodiment, the spectrum analysis unit 30
Learning of the neural network unit 12 can be performed using the spectrum analysis result of. (4) Next, a fourth embodiment of the present invention will be described. The same parts as those in FIG. 1 are designated by the same reference numerals, and detailed description thereof will be omitted.

FIG. 7 is a block diagram of the surface inspection apparatus. This apparatus is applied to the feel test of the edge portion 40 as the object to be measured. The calculator 41 presses the pressing direction (Z direction) against the edge portion 40 of the probe 3 detected by the force detection unit 2.
2 component force in the scanning direction (corresponding to the Y component).

That is, the computer 41 calculates the Z component force and the Y component force.
The input units 42 and 43 for inputting the component force and the determination unit 44
It has each function of. The determination unit 44 includes the input units 42,
Input each component force output of ZY input through 43,
A feature extraction unit that extracts the feature amount of the waveform of the component force output, and information processing based on the feature amount of the waveform of each ZY component force output extracted by the feature extraction unit to perform edge processing.
It has the functions of a neural network section for detecting scratches on the edge and evaluating surface characteristics such as roughness and smoothness of the surface of the edge section 40.

With such a configuration, when the feel test is performed on the edge portion 40, the Z component force and the Y component force acting on the probe 3 are input to the calculator 41. This calculator 41
The determination unit 44 of extracts the feature amounts of the waveforms of these component force outputs, performs information processing based on the feature amounts of these waveforms to detect scratches on the edge portion 40, and also makes the surface of the edge portion 40 rough and smooth. Evaluate surface features such as size.

As described above, according to the fourth embodiment, when the feel test is performed on the edge portion 40, by using the Z component force and the Y component force acting on the probe 3, the scratch on the edge portion 40 can be obtained. Surface characteristics such as detection and surface roughness of the edge portion 40, smoothness, etc. can be evaluated.

The present invention is not limited to the above embodiments, but may be modified as follows. For example, the tracing stylus 3 is normally in contact with the surface of the DUT 4 by the detection unit supporting mechanism 1 such as the arm mechanism or the compliance mechanism as described above by a substantially constant pressing force, and the responsivity of the tracing stylus 3 is increased. Is uniquely determined by the dynamic characteristics of the detection part support mechanism 1 and cannot be varied.

On the other hand, as shown in FIG.
An actuator 50 such as T is provided and driven by an actuator driving device 51, and the Z component force detected by the force detection unit 2 is sent to a compensating element 53 through an amplifier 52 to change the transfer characteristic of this compensating element 53. Thus, the deviation from the target value may be sent to the actuator driving device 51 to perform feedback control. This allows the dynamic characteristics of the tracing stylus 3 to be set arbitrarily.

[0056]

As described above in detail, according to the present invention, it is possible to provide a surface inspection method and apparatus for objectively and quantitatively measuring surface characteristics such as surface scratches and roughness.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing a first embodiment of a surface inspection apparatus according to the present invention.

FIG. 2 is a diagram showing each type of probe.

FIG. 3 is a specific configuration diagram of a determination unit.

FIG. 4 is a configuration diagram showing a second embodiment of the surface inspection apparatus according to the present invention.

FIG. 5 is a functional block diagram of a computer showing a third embodiment of the surface inspection apparatus according to the present invention.

FIG. 6 is a diagram showing a spectrum analysis result.

FIG. 7 is a configuration diagram showing a fourth embodiment of the surface inspection apparatus according to the present invention.

FIG. 8 is a configuration diagram showing a modified example of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Detection part support mechanism, 2 ... Force detector, 3 ... Measuring element, 4 ...
Object to be measured, 6 ... Calculator, 10 ... Judgment unit, 11 ... Feature extraction unit, 12 ... Neural network unit, 20 ... Base,
21 ... Force detection unit, 22 ... Measuring element, 23 ... Measuring head, 2
5 ... Calculator, 29 ... Judgment unit, 30 ... Spectrum analysis unit,
41 ... Calculator, 44 ... Judgment unit.

Claims (6)

[Claims] 1. A fine measuring element is brought into contact with the surface of the object to be measured to scan the surface, and the forces acting on the measuring element at this time are detected as at least two directional components, and based on each detected directional component. A surface inspection method for detecting the surface of the object to be measured. 2. A minute measuring element that scans the surface of the object to be measured in contact with the surface of the object, force detecting means for detecting forces acting on the measuring element as at least two-direction components, and the force detecting means. And a determination means for detecting the surface of the object to be measured based on each direction component. 3. The surface inspection according to claim 2, wherein the probe is attached with any one of a point contact type, a line contact type, and a surface contact type according to a contact form with the surface of the object to be measured. apparatus. 4. A plurality of minute measuring elements for scanning the surface of an object to be measured in contact with each other, a plurality of force detecting means for detecting forces acting on these measuring elements as at least two-direction components, and these measurements. A measuring head composed of a base in which the child and the force detecting means are integrated and arranged one-dimensionally or two-dimensionally, and the measured object based on at least two-direction components at a plurality of measuring points detected by the measuring head. A surface inspection apparatus comprising: a determination unit that detects the surface of an object. 5. The determination means includes a neural network for determining at least the surface roughness of the object to be measured from the waveform characteristics of at least two-direction components of the force acting on the probe. 4. The surface inspection device according to 4. 6. The determination means includes a neural network that spectrally analyzes at least two-direction components of the force acting on the probe, and determines at least the surface roughness of the object to be measured from the result of the spectral analysis. The surface inspection apparatus according to claim 2 or 4.