WO2020199206A1

WO2020199206A1 - Transparent or semi-transparent material microscopic-defect detection system and method

Info

Publication number: WO2020199206A1
Application number: PCT/CN2019/081552
Authority: WO
Inventors: 王星泽; 闫静
Original assignee: 合刃科技（深圳）有限公司
Priority date: 2019-04-04
Filing date: 2019-04-04
Publication date: 2020-10-08
Also published as: CN111316090A

Abstract

Provided are a transparent or semi-transparent material microscopic-defect detection system and method, said detection system comprising: a coherent light source (10), used for emitted coherent beam scanning of a transparent or semi-transparent sample material; a photoelectric sensor (20), used for acquiring scattered light intensity imaging of said sample material; a controller (40), used for obtaining the scattered light intensity imaging corresponding to the scan position; performing a Fourier transform on the scattered light intensity imaging to obtain a corresponding amplitude spectrum and phase spectrum, and obtaining the cutoff frequencies of said amplitude spectrum and phase spectrum; said cutoff frequencies correspond to the scan position and corresponding relative position of acquisition; monitoring the scan position when the cutoff frequency does not fall within a frequency threshold range, taking the scan position to be a defect position, and determining a defect according to said defect position; the frequency threshold range is calculated in advance according to the Lorenz–Mie theory and is a reference value range based on the light scattering of an impure medium corresponding to the relative position of acquisition.

Description

Transparent or semi-transparent material micro defect detection system and method

Technical field

The invention relates to the technical field of industrial inspection, in particular to a system and method for detecting microscopic defects of transparent or translucent materials.

Background technique

Under normal circumstances, in the production process of products, common micro-defects such as bubbles, impurities, cracks, ripples, scratches, and bumps are often unavoidable. For products, these surface defects will seriously affect the performance and life of the product; for consumers, scratches are one of the most easily found and most intolerable defects. At present, the detection methods for scratches in the industrial inspection field are: 1. Use low-angle monochromatic light to illuminate the object; 2. Collect images by CCD/CMOS; 3. Use image processing software to process the images and defects Perform testing. Due to the limitations of the current lighting hardware, the traditional method can only detect scratches in a certain direction, but in practical applications, the direction of the scratches is uncertain, so it is easy to miss and misdetect.

There is also a detection method based on active structured light projection in the traditional technology. However, when performing appearance inspection on a specific transparent or semi-transparent surface, due to the use of optical imaging technology to perform appearance inspection on a transparent or semi-transparent surface, a local brightness saturation area is often generated due to too strong surface reflection, making this area unable to be segmented. Recognition, or the image cannot obtain surface defect information due to too strong transmission, so the commonly used optical three-dimensional scanning and two-dimensional imaging methods are difficult to detect microscopic defects on transparent or semi-transparent surfaces.

Summary of the invention

Based on this, in order to solve the problem that the commonly used optical three-dimensional scanning and two-dimensional imaging methods are difficult to detect microscopic defects on transparent or semi-transparent surfaces, and the missed detection rate and false detection rate are high, the present invention specifically proposes A system for detecting microscopic defects of transparent or translucent materials.

A microscopic defect detection system for transparent or translucent materials, including:

Coherent light source, used to emit a coherent beam to scan transparent or semi-transparent samples;

The photoelectric sensor is used to collect the scattered light intensity imaging of the sample, the photoelectric sensor, the coherent light source and the sample have a relative collection position in space, and the relative collection position includes at least a scattering angle and a State the distance of the seized material;

A controller for acquiring the scattered light intensity imaging corresponding to the scanning position; performing Fourier transform on the scattered light intensity imaging to obtain corresponding amplitude-frequency curve information and phase-frequency curve information, and obtain the amplitude-frequency curve Cut-off frequency of information and phase-frequency curve information; the cut-off frequency corresponds to the scanning position and the relative collection position corresponding to the scanning position;

Monitoring the scanning position when the cutoff frequency does not belong to the frequency threshold range, using the scanning position as a defect position, and determining the defect according to the defect position;

The frequency threshold range is a reference value range based on light scattering from an impure medium, which is calculated in advance according to the Lorenz-Mie theory and corresponds to the relative collection position.

In one of the embodiments, the system further includes a stage for carrying the inspection material;

The stage also includes an orthogonal X-direction and Y-direction movement mechanism for driving the inspection material to move in the X direction and the Y direction for scanning;

The position of the coherent light source, the photoelectric sensor, and the sample is relatively fixed during the scanning process.

In one of the embodiments, the controller is used to move the coherent light source and/or the photoelectric sensor to scan the sample;

The controller is also used to monitor the relative collection position of the photoelectric sensor with respect to the coherent light source and the sample, and obtain the frequency threshold range corresponding to the relative collection position.

In one of the embodiments, the number of the photoelectric sensors is N, and N is greater than or equal to 2. The N photoelectric sensors, the coherent light source, and the sample have N relative collection positions in space.

In one of the embodiments, during the scanning process, for a scanning position, there are corresponding N relative acquisition positions;

The controller is also used to monitor the N cutoff frequencies corresponding to the scattered light intensity imaging collected by the N relative acquisition positions corresponding to the scanning position, and if there is a cutoff frequency that does not belong to the corresponding frequency threshold range, the scanning position is recorded Is the defect location.

In one of the embodiments, the coherent light source is a laser light source, and the inspection material is scanned in a line scan manner.

In addition, in order to solve the problems that the commonly used optical three-dimensional scanning and two-dimensional imaging methods are difficult to detect microscopic defects on transparent or semi-transparent surfaces, and the missed detection rate and false detection rate are high, the present invention also proposes a problem. A method for detecting microscopic defects of transparent or translucent materials is based on the aforementioned controller.

A method for detecting microscopic defects of transparent or translucent materials, including:

Scan the sample with a coherent beam;

When scanning to a scanning position: collect the scattered light intensity imaging of the sample at the relative collection position; perform Fourier transform on the scattered light intensity imaging to obtain the corresponding amplitude-frequency curve information and phase-frequency curve information, and obtain all State the cutoff frequency of amplitude-frequency curve information and phase-frequency curve information;

The relative collection position includes at least a scattering angle and a distance to the sample; the cutoff frequency corresponds to the scanning position and the relative collection position corresponding to the scanning position;

In one of the embodiments, the scan is a line scan, and the scan direction includes orthogonal X direction and Y direction.

In one of the embodiments, there are N relative collection positions, and N is greater than or equal to 2;

The monitoring of the scanning position when the cutoff frequency does not belong to the frequency threshold range as the defect position includes:

For a scan position, if there is a cutoff frequency in the N cutoff frequencies corresponding to the scattered light intensity imaging collected at the N relative collection positions corresponding to the scan position, and there is a cutoff frequency that does not belong to the corresponding frequency threshold range, then the scan is recorded The location is the defect location.

In one of the embodiments, after determining the defect according to the defect location, the method further includes:

Perform verification on the defect, and update the frequency threshold range according to the verification result

Implementing the embodiments of the present invention will have the following beneficial effects:

Using the above-mentioned transparent or semi-transparent material micro-defect detection system and the transparent or semi-transparent material micro-defect detection method based on the controller in the system, the transparent or semi-transparent material is scanned by coherent light, and the scan is When scanning the position, collect the scattered light intensity distribution at a certain scattering angle and distance, and compare the scattered light intensity distribution with the reference value calculated according to the Lorenz-Mie theory, and the scattered light intensity distribution does not meet the reference value The scanning position is recorded as the defect position, so that after the scanning is completed, the size and type of the defect in the sample can be determined by integrating the collection of the defect position in the sample. Since the above-mentioned system and method collect scattered light, they will not be affected by the light transmittance of transparent or semi-transparent samples. At the same time, since the present invention compares the scattered light intensity distribution collected in real time with the theoretically calculated reference value, the false detection rate and the missed detection rate are low.

Description of the drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

among them:

Figure 1 is a schematic diagram of a transparent or translucent material micro defect detection system in an embodiment;

Figure 2 is a distribution diagram of the spatial scattered light intensity of the same refractive index and different defect sizes;

Figure 3 shows the amplitude-frequency curve and phase-frequency curve of the scattered light intensity distribution with a refractive index of 0.56 and a defect size of 6.6um after FFT transformation in an embodiment;

Figure 4 shows the amplitude-frequency curve and phase-frequency curve of the scattered light intensity distribution with a refractive index of 0.56 and a defect size of 13.2um after FFT transformation in an embodiment;

Fig. 5 is a schematic diagram of defect detection principle in an embodiment;

Fig. 6 is a flow chart of a method for detecting microscopic defects of transparent or translucent materials in an embodiment.

detailed description

The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention.

In order to solve the problem of the common optical three-dimensional scanning and two-dimensional imaging methods in the traditional technology, it is difficult to detect microscopic defects on transparent or semi-transparent surfaces, and the missed detection rate and false detection rate are high. As shown in FIG. 1, the present invention A system for detecting microscopic defects of transparent or semi-transparent materials is proposed, which includes a coherent light source 10, a photoelectric sensor 20, and a controller 30.

Among them, the coherent light source 10 is used to emit a coherent beam to scan a transparent or semi-transparent sample, and may be a laser, a laser diode, a laser array, or the like.

The photoelectric sensor 20 is used to collect the scattered light intensity imaging of the sample, and it may be a camera or other photosensitive elements that can convert light signals into electrical signals. The controller 30 may be a computer system based on the Von Neumann system that relies on the execution of a computer program, may be a single-chip microcomputer integrated on other components, or may be an independent personal computer, notebook computer, or server device.

The transparent or translucent material micro defect detection system of the present invention is based on the generalized Lorenz-Mie theory.

When light passes through any medium except vacuum, part of the light will deviate from the original propagation direction and diffuse to the surroundings, which is the phenomenon of light scattering. Generally, light scattering is divided into two types: pure medium light scattering and impure medium light scattering.

(1) Pure medium light scattering: It means that light will still be scattered when light propagates in a pure substance without any defects. This phenomenon is inherent to any substance itself, and has nothing to do with whether the communicator contains defects.

(2) Light scattering from impure media: refers to the scattering phenomenon of light when light is transmitted in materials containing defects (such as bubbles, impurities, cracks, ripples, scratches, bumps, etc.) when it encounters these defects. This kind of scattering phenomenon is not inherent to the substance itself, and the intensity of its scattered light is related to the nature of the defect. The frequency of the scattered light and the incident light are the same, and the intensity of the scattered light is also related to the wavelength of the incident light, specifically:

Among them, _Isca is the intensity of scattered light, λ is the wavelength of incident light, and a is a positive integer.

Based on the generalized Lorenz-Mie theory, the relationship between the properties of the laser beam (a kind of coherent light) and the relative position of the scatterer, the parameter that can express their positional relationship is defined as g _n , when the laser is incident on the scatterer, when The position of the scatterer is at the center of the beam. At this time, the expression of g _n is:

Among them, ω ₀ is the laser beam waist size of the laser beam.

Then, when the incident laser beam is scattered by the scatterer, the light intensity of a point P in space is small:

among them,

Is the spatial coordinate of point P with the scatterer as the reference system, r is the distance from point P to the scatterer, θ is the scattering angle of point P,

Is the azimuth angle of point P, q is the particle size parameter, m is the relative complex refractive index, ω ₀ is the laser beam waist size of the laser beam, and I ₀ is the incident light intensity. To simplify the calculation, we can make

Is 0, that is, the azimuth angle of point P can be defined as

The distance is r on the circle.

At this time, the light intensity of point P can be expressed as:

Among them, a _n , b _n , τ _n , π _n are the Mie scattering coefficients.

Since m and q are unknown in actual detection, they can pass

To control the photoelectric sensor to obtain different spatial light intensity combinations at different spatial receiving positions, that is, select multiple P points on the scatterer sphere to collect scattered light intensity information, and establish the relationship between spatial light intensity and defect information, and then use theoretical simulation The method combined with reality establishes a database model corresponding to the relative collection position of the scattered light collected by the photoelectric sensor (the aforementioned P point position) and the spatial distribution of the scattered light intensity, and then the scattered light collected by the photoelectric sensor at a certain relative collection position P The light intensity is compared with the reference scattered light intensity corresponding to the relative collection position P stored in the database model. According to the difference, it can be judged whether the scattered light intensity collected at the relative collection position P is the scattered light caused by the defect. Determine whether there are defects and the size of the defects.

When applied to the embodiment of the present invention, that is, the photoelectric sensor 20, the coherent light source 10 and the sample have a relative collection position in space, and the relative collection position includes at least the scattering angle and the distance to the sample. The controller 30 is used to obtain the scattered light intensity imaging corresponding to the scanning position; perform Fourier transform on the scattered light intensity imaging to obtain the corresponding amplitude spectrum and phase spectrum, and obtain the corresponding amplitude spectrum and phase spectrum. Cutoff frequency; The cutoff frequency corresponds to the scanning position and the relative acquisition position corresponding to the scanning position.

The controller 30 is also used to monitor the scan position when the cutoff frequency does not belong to the frequency threshold range as the defect position, and determine the defect according to the defect position; and the frequency threshold range is calculated in advance according to the Lorenz-Mie theory, The reference value range based on the light scattering of the impure medium corresponding to the relative collection position.

Refer to Figure 2, which shows the distribution of spatially scattered light intensity under defects with a material refractive index of 0.56 and a defect radius of 6.6um and 13.2um. As the size of the scatterer increases or becomes smaller, the value of the distance between the wave crests in the spatial range becomes smaller, more precisely, the value of the angle between the two wave crests becomes smaller and smaller. The distribution of the spatial scattered light intensity within a certain range is passed through FFT (Fourier transform), for the same refractive index (same defect material) and different q values (different defect sizes), as the q value increases, that is, as the defect size increases, the phase frequency and amplitude frequency The cutoff frequency of the curve also increases accordingly.

Referring again to Figures 3 and 4, Figures 3 and 4 show the amplitude-frequency curve and phase-frequency curve of the scattered light intensity distribution after FFT transformation. As the size of the scatterer becomes larger, that is, the size of the defect becomes larger, the cut-off frequency of the scattered light in the fixed space also increases monotonously, so in the actual inspection process, the cut-off frequency of the scattered light in the fixed space can be used to judge Defect size. In actual inspection, it is only necessary to compare the actual received scattered light cut-off frequency with the theoretically calculated simulation value to determine the existence and size of the defect.

Specifically, in this embodiment, the system further includes a stage 40 for carrying the inspection material. As shown in Fig. 1, the stage also includes orthogonal X-direction and Y-direction movement mechanisms for driving the inspection material to move in the X-direction and Y-direction for scanning. The coherent light source, the photoelectric sensor, and the sample are relatively fixed in position during the scanning process.

In this embodiment, the scanning process is a linear coherent light beam emitted by a coherent light source, and the stage 40 moves the animal in the X or Y direction along the X-axis or Y-axis, so that the coherent beam is oriented toward transparent or semi-transparent building materials. Carry out line scanning, and then use the photoelectric sensor to image the scattered light intensity of the object, process the image to obtain the scattered light intensity distribution map of the object, and perform FFT transformation on the scattered light intensity distribution map to obtain the amplitude-frequency curve and phase-frequency curve for real-time monitoring The cut-off frequency of the amplitude-frequency curve information and phase-frequency curve information generated by the scattered light intensity of the measured object through FFT transformation, and the cut-off frequency is a database corresponding to the relative collection position pre-calculated by the theoretical value and the spatial distribution of the scattered light intensity The frequency threshold range stored in the model is compared with the frequency threshold range corresponding to the relative acquisition position in real time. If the cut-off frequency is not within the frequency threshold range, the scanning position is recorded as the defect position.

For example, referring to Figure 5, when the stage 40 drives the inspection material to move along the X axis, when it reaches the x1 position, it is detected that the cutoff frequency is not in the corresponding frequency threshold range, and when it reaches the x2 position, the cutoff frequency is detected The recovery is in the corresponding frequency threshold range, which can be recorded on the X axis, and there is a defect in the position between x1 and x2. Similarly, when the stage 40 drives the inspection material to move along the Y axis, when it reaches the y1 position, it is detected that the cutoff frequency is not in the corresponding frequency threshold range, and when the y2 position is reached, the cutoff frequency is detected to be restored to the corresponding frequency threshold. In the range, it can be recorded on the Y axis, and the position in the interval between y1 and y2 is defective. Then, it can be determined that there are defects in the rectangular positions determined by the four vertices x1, x2, y1, and y2 of the sample.

In another embodiment, the controller 30 is used to move the coherent light source 10 and/or the photoelectric sensor 20 to scan the sample. The controller 30 is also used to monitor the relative collection position of the photoelectric sensor 20 with respect to the coherent light source 10 and the sample, and obtain the frequency threshold range corresponding to the relative collection position.

That is to say, a fixed stage can be used to scan the sample without moving the sample through the stage, but the scanning line can be moved by moving the coherent light source 10 to realize the scanning of the sample. But for the controller, the defect recognition algorithm is the same. It only needs to obtain the relative position of the real-time coherent light source 10 and the photoelectric sensor 20 to determine the relative acquisition position of the photoelectric sensor 20 to collect the scattering imaging at the time, which can be read from the aforementioned database. Take the corresponding frequency threshold range and compare it with the cut-off frequency of real-time monitoring to identify the defects of the inspected material.

In an embodiment, the number of photoelectric sensors may be N, and N is greater than or equal to 2. The N photoelectric sensors, the coherent light source, and the inspection material have N relative collection positions in space, and in the scanning process, for a scanning position, there are corresponding N relative collection positions.

The controller is also used to monitor the N cutoff frequencies corresponding to the scattered light intensity imaging collected by the N relative acquisition positions corresponding to the scanning position, and if there is a cutoff frequency that does not belong to the corresponding frequency threshold range, the scanning position is recorded as a defect position.

In other words, multiple photoelectric sensors can be set, and when scanning to a scanning position, it can be collected at multiple relative collection positions, as long as there is a cut-off frequency of the scattered light intensity distribution collected at a relative collection position and is not pre-stored in the database Within the frequency threshold range corresponding to the relative collection position, it is determined that the sample has a defect at the scanning position. In this way, a single photoelectric sensor collection may be affected by ambient light or other noises, which may cause missed detection. Multiple photoelectric sensors can be used to detect at multiple scattering angles, thereby preventing missed detection.

After inspecting the defects in the material, the controller 30 can verify the defects in the inspected material by other means, such as manual re-inspection, or re-inspection in other ways, so as to confirm the accuracy of the defect detection. In this case, it means that the theoretical reference value stored in the database is inaccurate. In this case, the frequency threshold range corresponding to the relative acquisition position of the wrongly detected defect in the database can be updated according to the re-inspection situation, so that the theoretical reference value can be updated The reference value can reduce false detections and improve the accuracy of detection during the next detection.

In order to solve the problems of conventional optical three-dimensional scanning and two-dimensional imaging methods, it is difficult to detect microscopic defects on transparent or semi-transparent surfaces, and the missed detection rate and false detection rate are high. As shown in FIG. 6, the present invention A method for detecting microscopic defects of transparent or semi-transparent materials is also proposed. The execution of the method relies on a computer program and is based on the aforementioned controller 30. Specifically, the method includes:

Step S102: Scan the sample through the coherent beam.

Step S104: When scanning to a scanning position: collect the scattered light intensity imaging of the sample at a relative collection position; perform Fourier transform on the scattered light intensity imaging to obtain the corresponding amplitude spectrum and phase spectrum, and obtain the Cutoff frequency of the amplitude spectrum and phase spectrum.

The relative collection position includes at least a scattering angle and a distance to the sample; the cutoff frequency corresponds to the scanning position and the relative collection position corresponding to the scanning position.

Step S106: Monitor the scan position when the cut-off frequency does not belong to the frequency threshold range as the defect position, and determine the defect based on the defect position; the frequency threshold range is calculated in advance according to the Lorenz-Mie theory, and is consistent with The reference value range based on the light scattering of the impure medium corresponding to the relative collection position.

In one embodiment, the scan is a line scan, and the scan direction includes orthogonal X direction and Y direction.

In one embodiment, there are N relative collection positions, and N is greater than or equal to 2. Monitoring the scanning position when the cutoff frequency does not belong to the frequency threshold range as the defect position includes:

In an embodiment, after determining the defect according to the defect location, the method further includes: verifying the defect, and updating the frequency threshold range according to the verification result.

The embodiments of the present invention will have the following beneficial effects:

Using the above-mentioned transparent or translucent material micro-defect detection system, and the transparent or translucent material micro-defect detection method based on the controller in the system, for transparent or translucent materials, scan them by coherent light, and then scan them. When scanning the position, collect the scattered light intensity distribution at a certain scattering angle and distance, and compare the scattered light intensity distribution with the reference value calculated according to the Lorenz-Mie theory, and the scattered light intensity distribution does not meet the reference value The scanning position is recorded as the defect position, so that after the scanning is completed, the size and type of the defect in the sample can be determined by integrating the collection of the defect position in the sample. The above-mentioned system and method will not be affected by the light transmittance of transparent or translucent samples due to the scattered light collected. At the same time, since it compares the scattered light intensity distribution collected in real time with the theoretically calculated reference value, the false detection rate and the missed detection rate are low.

The above-disclosed are only preferred embodiments of the present invention, which of course cannot be used to limit the scope of rights of the present invention. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope of the present invention.

Claims

A system for detecting microscopic defects of transparent or translucent materials, which is characterized in that it comprises:

Coherent light source, used to emit a coherent beam to scan transparent or semi-transparent samples;

The photoelectric sensor is used to collect the scattered light intensity imaging of the sample, the photoelectric sensor, the coherent light source and the sample have a relative collection position in space, and the relative collection position includes at least a scattering angle and a State the distance of the seized material;

A controller for acquiring the scattered light intensity imaging corresponding to the scanning position; performing Fourier transform on the scattered light intensity imaging to obtain corresponding amplitude-frequency curve information and phase-frequency curve information, and obtain the amplitude-frequency curve The cut-off frequency of the information and the phase-frequency curve information; the cut-off frequency corresponds to the scanning position and the relative collection position corresponding to the scanning position;

Monitoring the scanning position when the cutoff frequency does not belong to the frequency threshold range, taking the scanning position as a defect position, and determining the defect according to the defect position;

The frequency threshold range is a reference value range based on light scattering from an impure medium, which is calculated in advance according to the Lorenz-Mie theory and corresponds to the relative collection position.
The transparent or semi-transparent material microscopic defect detection system according to claim 1, wherein the system further comprises a stage for carrying the inspection material;

The stage also includes an orthogonal X-direction and Y-direction movement mechanism for driving the inspection material to move in the X direction and the Y direction for scanning;

The position of the coherent light source, the photoelectric sensor, and the sample is relatively fixed during the scanning process.
The transparent or translucent material microscopic defect detection system according to claim 1, wherein the controller is used to move the coherent light source and/or the photoelectric sensor to scan the inspection material;

The controller is also used to monitor the relative collection position of the photoelectric sensor with respect to the coherent light source and the sample, and obtain the frequency threshold range corresponding to the relative collection position.
The system for detecting microscopic defects of transparent or semi-transparent materials according to claim 1, wherein the number of said photoelectric sensors is N, and N is greater than or equal to 2; N of said photoelectric sensors and said coherent light source and The sampled material has N relative collection positions in space.
The system for detecting micro-defects of transparent or semi-transparent materials according to claim 4, characterized in that, in the scanning process, for a scanning position, there are corresponding N relative collection positions;

The controller is also used to monitor the N cutoff frequencies corresponding to the scattered light intensity imaging collected by the N relative acquisition positions corresponding to the scanning position, and if there is a cutoff frequency that does not belong to the corresponding frequency threshold range, the scanning position is recorded Is the defect location.
The microscopic defect detection system for transparent or translucent materials according to claim 1, wherein the coherent light source is a laser light source, and the inspection material is scanned in a line scan manner.
A method for detecting microscopic defects of transparent or translucent materials, based on the controller according to any one of claims 1 to 6, characterized in that it comprises:

Scan the sample with a coherent beam;

When scanning to a scanning position: collect the scattered light intensity imaging of the sample at the relative collection position; perform Fourier transform on the scattered light intensity imaging to obtain the corresponding amplitude-frequency curve information and phase-frequency curve information, and obtain all State the cutoff frequency of amplitude-frequency curve information and phase-frequency curve information;

The relative collection position includes at least a scattering angle and a distance to the sample; the cut-off frequency corresponds to the scanning position and the relative collection position corresponding to the scanning position;

Monitoring the scanning position when the cutoff frequency does not belong to the frequency threshold range, using the scanning position as a defect position, and determining the defect according to the defect position;

The frequency threshold range is a reference value range based on light scattering from an impure medium, which is calculated in advance according to the Lorenz-Mie theory and corresponds to the relative collection position.
8. The method for detecting microscopic defects of transparent or semi-transparent materials according to claim 7, wherein the scanning is a line scan, and the scanning direction includes orthogonal X direction and Y direction.
The method for detecting microscopic defects of transparent or semi-transparent materials according to claim 7, wherein there are N relative collection positions, and N is greater than or equal to 2;

The monitoring of the scanning position when the cutoff frequency does not belong to the frequency threshold range as the defect position includes:

For a scan position, if there is a cutoff frequency in the N cutoff frequencies corresponding to the scattered light intensity imaging collected at the N relative collection positions corresponding to the scan position, and there is a cutoff frequency that does not belong to the corresponding frequency threshold range, then the scan is recorded The location is the defect location.
The method for detecting microscopic defects of transparent or semi-transparent materials according to claim 7, wherein after determining the defect according to the defect position, the method further comprises:

The defect is verified, and the frequency threshold range is updated according to the verification result.