Micro device morphology measurement and calibration method based on Michelson interference

Abstract

The invention provides a micro device morphology measurement and calibration method based on Michelson interference, and the method employs the Michelson interference principle to carry out calibration under the conditions of limited precision and uncertain step pitch of a piezoelectric ceramic controller. Through a system based on a white light interference light path structure, plane mirrors are placed on two sides of a piezoelectric ceramic driving system to form two sets of Michelson interferometers. The Michelson interferometer on the first side is used for obtaining interference fringes on the surface of a tiny device, the step pitch of the piezoelectric ceramic is accurately calculated according to the change relation of the interference fringes in combination with digital image processing, and then the step pitch of the piezoelectric ceramic is calibrated through the Michelson interferometer on the other side. Compared with a Michelson interferometer without a calibration system, the method has the advantages that the step pitch of the piezoelectric ceramic can be accurately controlled, a clearer and more accurate interference image can be obtained, and the surface appearance restoration capability of the system on a tiny device can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of micro-device surface morphology detection technology using white light interference image acquisition technology, and in particular to a method for calibrating a piezoelectric ceramic controller (PZT) when measuring the surface morphology of an optical fiber connector and a semiconductor material using Michelson interference under the condition of limited accuracy and uncertain step distance. Background Art

[0002] White light interference: White light is a complex light mixed with multiple monochromatic lights. The interference phenomenon caused by light is manifested as a stable interference phenomenon of strong and weak distribution when coherent lights with the same frequency, constant phase difference and consistent vibration direction meet. Therefore, in the field of micro-device morphology measurement, using the principle of white light interference to shoot the surface can adapt to micro-devices with unclear surface height fluctuations, obtain several interference images, and restore the surface morphology 3D image after computer algorithm analysis.

[0003] Michelson interferometer: The principle of Michelson interferometer is that after a beam of incident light is split into two beams by a beam splitter, it is reflected by a plane mirror for reflection to obtain an image of intensity distribution. In the actual detection process, the piezoelectric ceramic controller in the optical path structure is used to control the plane mirror to move up and down with equal steps. At the same time, the reflector in another Michelson interferometer with He-Ne laser as the light source also moves, and the interference produces parallel and equally spaced light and dark stripes. As the reflector moves, the interference stripes will increase or decrease to a certain extent. Using the known wavelength of light and the movement of the interference stripes, the measured distance can be calculated.

[0004] Piezoelectric ceramics: Piezoelectric ceramics are a type of ceramic material that converts mechanical energy into electrical energy based on the piezoelectric effect. When deformed by an external force, polarization occurs inside the ceramic, and positive and negative charges appear on the surface, generating a voltage output. When voltage is applied to a piezoelectric ceramic element, deformation occurs, but the deformation amount is not necessarily accurate. Therefore, the distance moved by the piezoelectric ceramic controller (PZT) in the optical path interference system may have errors, thus affecting the restoration results of the micro-device morphology measurement.

[0005] He-Ne laser: Interference measurement technology is generally used in laser ranging, such as interferometric ranging or interferometric vibration measurement, so a laser with good coherence is needed to ensure ranging accuracy and measurement stability. A laser with good coherence has a longer coherence length, that is, the light wave has a higher spatial coherence, which can provide more stable phase information and help reduce measurement errors. The coherence of a laser is related to the wavelength of the light wave it emits. Lasers with longer coherence lengths are usually in the visible or near-infrared wavelength range (between 400 nanometers and 1000 nanometers). The most traditionally used laser is a semiconductor laser, but its emission spectrum is wide and the phase relationship of the spectrum is unstable, resulting in a shorter coherence length. It is also more sensitive to temperature changes and has poor stability. In contrast, the wavelength of the He-Ne laser is 632.8 nanometers, so the coherence length is longer and the stability is stronger, which is suitable for laser ranging. Summary of the invention

[0006] The purpose of the present invention is to provide a method for calibrating a piezoelectric ceramic controller (PZT) with limited accuracy and uncertain step distance when measuring the surface morphology of optical fiber connectors and semiconductor materials using Michelson interference in the field of micro-device surface morphology restoration.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] An image acquisition calibration system based on a piezoelectric ceramic controller whose step distance is inaccurate due to multiple factors, the method performs piezoelectric ceramic controller movement step distance feedback based on a white light interference optical path structure, comprising the following steps:

[0009] S1. Instrument assembly: Install plane mirrors on both sides of the piezoelectric ceramics. Use a white light source on one side with a Michelson interferometer to measure the surface of tiny devices. Use a He-Ne laser on the other side to detect the step length of the piezoelectric ceramics and obtain the size of the piezoelectric ceramic step length error. The purpose of using the He-Ne laser is to improve the measurement accuracy and stability.

[0010] S2. Calibration step: The piezoelectric ceramic moves while driving the plane mirror (moving mirror) to move. The light emitted from the light source is divided into two beams by the beam splitter, one of which is directed to the moving mirror, and the other is reflected to the fixed mirror after compensation. The two beams of light meet again at the beam splitter and form interference. The interference image is captured by a CCD camera. The change of the interference fringes satisfies the relationship between the translation distance d and the number of fringe movements N: d = Nλ / 2, λ is the wavelength of the incident light. The optical path difference corresponding to each fringe spacing is λ / 2. After digital image processing, the fringe spacing is subdivided to improve the calibration accuracy of the displacement corresponding to the CCD pixel. The piezoelectric ceramic step can be calculated based on the number of pixels and the number of indents in the interference image captured by the CCD camera, and the calibration operation can be performed.

[0011] S3. Micro-morphology measurement: As the piezoelectric ceramic moves up and down with equal steps, interference occurs after the beam splitter converges. Every time the piezoelectric ceramic moves, an interference image is captured using a CCD camera and the interference image is transmitted to a computer.

[0012] S4. Processing of interference images of micro devices: CCD camera takes 100 interference images of the surface of the device to be tested, and extracts the pixels at the same position of the 100 images, which is called interference curve. The interference curve is transformed into time domain and frequency domain by Fourier transform, and the accurate end face image white light interference curve is calculated by genetic algorithm.

[0013] The interference curve is fitted to obtain the envelope and the highest point of all interference curves is found, which is the relative height value of the fiber optic connector and the semiconductor material surface at the pixel point. The 3D restored image of the surface is obtained by identifying noise points through neural network and filtering restoration. Compared with the Michelson interferometer without a calibration system, the Michelson interferometer using this calibration method can achieve precise control of the step size of piezoelectric ceramics, so that clearer and more accurate interference images can be obtained, and the system's ability to restore the surface morphology of tiny devices can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 :White light interference detection of micro device surface morphology structure diagram

[0015] Figure 2 :Optical path structure system flow chart

[0016] Figure 3 :Micro device surface morphology detection flow chart DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0018] The present invention provides a method for calibrating a piezoelectric ceramic (PZT) controller when measuring the surface morphology of an optical fiber connector and a semiconductor material by using Michelson interference due to the limited precision and uncertain step distance of the piezoelectric ceramic itself. Figure 1 As shown, it includes the following steps:

[0019] S1. Instrument structure: In this device, there is a piezoelectric ceramic device, which acts as a controller and moves in equal steps through voltage regulation. A plane mirror is placed on each side of the device. These two plane mirrors are used to reflect light and form the optical path of two Michelson interferometers. When light hits the surface of the device, the reflected light passes through the Michelson interferometer on the first side to form an interference pattern. By observing and analyzing these interference patterns, the morphology, height and other parameters of the device surface can be inferred. Using digital image processing technology, the step length of the piezoelectric ceramic in the Michelson interference process on the first side is calculated and compared with the standard step length value. If the actual step length is not within the error range, the Michelson interferometer on the other side is used to correct the step length of the piezoelectric ceramic. Since piezoelectric ceramics are a material that can deform under the action of an electric field, by adjusting the optical path of the Michelson interferometer to observe the changes in the interference pattern for feedback and control, it is possible to achieve a slight adjustment of the step length of the piezoelectric ceramic and achieve precise correction of the step length.

[0020] S2. Calibration test: When the PZT is extended or shortened under the action of an external voltage, the test reflector fixedly connected to the probe is pushed and translated by the PZT under the probe. The light emitted from the light source is divided into two beams by the beam splitter, one of which is emitted to the moving mirror, and the other is reflected to the fixed mirror after compensation. The two beams of light meet again at the beam splitter and form interference, and the fringe shape is interference straight fringes. By using a CCD camera to capture the image of the interference fringes, it is convenient to transmit it back to the computer for processing and calculation of data such as optical path difference.

[0021] The corresponding white light interference fringes will be scanned on the CCD target surface. The change of the interference fringes satisfies the relationship between the translation distance d and the number of fringe movements N, d = Nλ / 2, where λ is the wavelength of the incident light. Each fringe spacing corresponds to an optical path difference of λ / 2. After digital image processing, the fringe spacing is subdivided, and the four-step phase shifting technology is combined: I1 (x,y)=a(x,y)+b(x,y)cosφ(x,y) (1) I2 (x,y)=a(x,y)+b(x,y)cos[φ(x,y)+π / 2] (2) I3 (x,y)=a(x,y)+b(x,y)cos[φ(x,y)+π] (3) I4 (x,y)=a(x,y)+b(x,y)cos[φ(x,y)+3π / 2] (4) By combining equations (1) to (4), we can obtain the measured fixed phase function φ(x, y): According to this measurement method, the elongation data of PZT can be obtained, and the relationship curve between the elongation of PZT and the change of voltage can be obtained. The calibration accuracy of the displacement corresponding to the CCD pixel can be improved. The piezoelectric ceramic step distance can be calculated and the calibration operation can be performed.

[0022] S3. Micro device interference image processing: Figure 1 The CCD camera in the middle and lower part normally takes 125 photos of the device under test, i.e., the optical fiber connector or the surface of the semiconductor material. The image pixels are 564*1344 and the format is bmp. The interference data is obtained by extracting the gray value of each pixel in each picture to form an interference image.

[0023] The image has a total of 758,016 pixels. The feature dimension reduction step transforms the interference image from the time domain image to the frequency domain image through Fourier transform. The frequency domain image is resolved by genetic algorithm to determine the effective interference data, which can reduce the calculation time to a certain extent.

[0024] S4, end face image restoration: The interference data after step S3 is used through neural networks such as BP and GRNN to identify surface noise points. Noise points are generally defective parts such as surface dust and oil stains. Then, the noise points are removed through mean filtering, Gaussian filtering and other methods to obtain a 3D restored image of the surface of the device to be tested, i.e., the optical fiber connector or semiconductor material. The parameters such as protrusions and depressions in the image are obtained through data analysis.

[0025] The specific implementation method described above roughly illustrates the technical solution of the present invention. It should be understood that the above is only the optimal embodiment of the present invention and is not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A micro-device morphology measurement and calibration method based on Michelson interferometry, characterized in that: include: A plane mirror is placed on each side of the piezoelectric ceramic to form the optical path of two Michelson interferometers; when light is irradiated onto the surface of the device to be tested, the reflected light passes through the Michelson interferometer on the first side to form an interference pattern; Using digital image processing technology, calculate the step length of the piezoelectric ceramics in the Michelson interferometer process on the first side, compare it with the standard step length value, and formulate a calibration measurement of the step length of the piezoelectric ceramics using the Michelson interferometer on the other side if the actual step length is not within the error range; By analyzing the Michelson interference image, reducing the dimension of the features, identifying the noise points through a neural network and filtering and restoring it into a 3D image, we can obtain parameters such as the relative height of the device surface and the depth of the depression.

2. The method according to claim 1, characterized in that: Building a Michelson interference optical path structure includes: adding another Michelson interferometer on the basis of the original Michelson interferometer and placing a plane mirror on both sides of the piezoelectric ceramic to form the optical path of two Michelson interferometers. The light is reflected by the surface of the device to be tested, resulting in an optical path difference and Michelson interference to form an interference image.

3. The method according to claim 1, characterized in that The piezoelectric ceramic controller step distance is affected by many factors and the image acquisition calibration system is inaccurate, including: In the steps, the specific method of the calibration operation is: calculate the piezoelectric ceramic step length and compare it with the standard step length value. If the actual step length is not within the error range, use the Michelson interferometer on the other side to correct the piezoelectric ceramic step length under the pre-distortion correction algorithm.

4. The method according to claim 1, characterized in that Image acquisition and calibration system based on piezoelectric ceramic controller step size inaccurate due to excessive data, including: In the above steps, the specific method of using genetic algorithm to calculate the accurate end face image white light interference curve is: transform the interference curve by Fourier transform, and use genetic algorithm to calculate the accurate end face image white light interference curve. After feature dimension reduction, neural network identification of noise points and filtering and restoring into 3D image, the relative height of the device surface, the depth of the depression and other parameters can be obtained.

Images