CN113218516A - Near-infrared speckle wavemeter - Google Patents
Near-infrared speckle wavemeter Download PDFInfo
- Publication number
- CN113218516A CN113218516A CN202110072037.5A CN202110072037A CN113218516A CN 113218516 A CN113218516 A CN 113218516A CN 202110072037 A CN202110072037 A CN 202110072037A CN 113218516 A CN113218516 A CN 113218516A
- Authority
- CN
- China
- Prior art keywords
- wavelength
- speckle
- photoelectric detector
- sampling
- infrared
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005070 sampling Methods 0.000 claims abstract description 27
- 230000003287 optical effect Effects 0.000 claims abstract description 22
- 238000009826 distribution Methods 0.000 claims abstract description 21
- 239000013307 optical fiber Substances 0.000 claims description 21
- 238000012545 processing Methods 0.000 claims description 12
- 238000001514 detection method Methods 0.000 claims description 9
- 238000010586 diagram Methods 0.000 claims description 6
- 238000000034 method Methods 0.000 claims description 6
- 238000013527 convolutional neural network Methods 0.000 abstract description 14
- 238000005259 measurement Methods 0.000 abstract description 10
- 230000008859 change Effects 0.000 abstract description 3
- 239000006185 dispersion Substances 0.000 abstract description 3
- 230000003321 amplification Effects 0.000 abstract 1
- 230000000694 effects Effects 0.000 abstract 1
- 238000000605 extraction Methods 0.000 abstract 1
- 238000003199 nucleic acid amplification method Methods 0.000 abstract 1
- 238000012549 training Methods 0.000 description 7
- 229910000530 Gallium indium arsenide Inorganic materials 0.000 description 6
- 238000012360 testing method Methods 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 238000000691 measurement method Methods 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 229920003229 poly(methyl methacrylate) Polymers 0.000 description 2
- 239000004926 polymethyl methacrylate Substances 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000011056 performance test Methods 0.000 description 1
- 239000004038 photonic crystal Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J9/00—Measuring optical phase difference; Determining degree of coherence; Measuring optical wavelength
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Spectrometry And Color Measurement (AREA)
Abstract
A near-infrared speckle wavelength meter is mainly used for measuring the laser wavelength of a near-infrared band at high speed and accurately. The wavelength meter mainly utilizes a speckle pattern generated by an intermodal dispersion effect of the scattering optical waveguide to demodulate the wavelength of light to be measured. In the speckle pattern collecting part, a down-sampling photoelectric detector is used for quickly collecting far-field speckle distribution of the emergent end face of the scattered light waveguide, and analog signals caused by wavelength change in a channel are collected through a collecting card after trans-impedance amplification. In the wavelength extraction process, original down-sampling photoelectric detector data is up-sampled into pictures, and then the wavelength is rapidly extracted through a trained convolutional neural network. The near-infrared wavelength meter collects the light intensity distribution of speckles by using a down-sampling photoelectric detector, replaces an expensive I nGaAs camera used by a common speckle wavelength meter, and has the advantages of low cost, high measurement speed, high resolution and strong anti-interference capability.
Description
Technical Field
The invention belongs to the field of laser spectrum characteristic parameter measurement, and discloses a device for measuring a wavelength value of unknown laser.
Technical background:
laser as a light source is widely applied in modern science and technology and engineering practice due to the characteristics of good monochromaticity, strong directivity, small dispersion and the like. The wavelength is one of the important parameters of laser and is also an important index in applications such as optical fiber communication, sensing and precision measurement. Therefore, the method can accurately and quickly measure the laser wavelength and has important significance in the field of basic research and application of optics. The traditional commercially available wavemeters are classified according to the measurement principle, and mainly include fizeau interference type wavemeters, fabry-perot interference type wavemeters and michelson interference type wavemeters, the basic working principle of which is based on light interference, but generally require an area array CCD to detect interference fringes and perform complex image processing. Some wavemeters, such as fabry-perot interferometric and michelson interferometric, also require a built-in reference laser ([1] north ethnic university, a laser wavelength measurement device and method based on interferometric approach [ P ]. china: CN201911173953.7, 2019, 2, shanghai optical precision mechanical institute of chinese academy of sciences [ P ]. china: CN201210193437.2, 2012). The factors make the existing wavelength meter have higher cost and large volume, and the anti-interference capability limited by the principle cannot be applied in large scale. In recent years, a speckle-based means for measuring wavelength has emerged, and the basic principle is to calculate the wavelength of incident light by obtaining and analyzing the speckle shape and the change of intensity distribution based on the speckle image generated after the light to be measured passes through the scattering medium. It provides more flexibility in the choice of dispersive elements such as disordered photonic crystals, random scattering media and bragg fiber arrays that have been used. It is widely adopted at present to use multimode optical fibers as speckles generated by a scattering medium, and a CCD camera is used to record a speckle pattern to measure the wavelength of incident light. There are many areas to be improved in combination with the existing laser wavelength measuring devices: first, there is no effective balance between high resolution and volume, cost, and sensitivity. If high resolution is sought, the price of volume, cost and sensitivity is paid; whereas if high sensitivity and low cost are required, the resolution is not very satisfactory. Secondly, most wavemeters need to use area array or linear array CCD to receive optical signals, and the photoelectric material with the best detection efficiency for near-infrared wave band is only InGaAs at present, but the area array or linear array CCD of InGaAs is extremely expensive, and tens of thousands to hundreds of thousands of lines are used, so that the price of the near-infrared wavemeter is high. Thirdly, due to the manufacturing principle, the electrical signals of the linear array or the area array CCD can only be read out sequentially. Even if the photoelectric response speed of the InGaAs single-pixel receiver can reach hundreds of megabits or even GHz, all spectral information can be acquired only after the pixels are sequentially read out one by one. This greatly reduces the detection rate of the spectrometer. Aiming at the performance short board of the near infrared wavemeter, the invention provides the near infrared speckle wavemeter which is low in cost, small in size, high in measurement speed, high in resolution and strong in anti-interference capability.
The invention content is as follows:
the invention provides a near-infrared speckle wavelength meter, which is a wavelength real-time rapid measuring device with low cost, high resolution and strong anti-interference capability, and consists of a scattering optical waveguide and a down-sampling photoelectric detector. A scattered light waveguide element is used for generating a speckle pattern, and a down-sampling photoelectric detector is used for collecting far-field speckle intensity distribution of an emergent end face of the scattered light waveguide. The collected speckle intensity distribution data is subjected to up-sampling and bicubic interpolation to enhance characteristic information, and finally, the wavelength is rapidly extracted through a trained convolutional neural network. The near-infrared wavelength meter uses a down-sampling photoelectric detector to collect the light intensity distribution of speckles, replaces an expensive InGaAs camera used by a common speckle wavelength meter, and has the advantages of low cost and higher measurement speed; meanwhile, the convolutional neural network is utilized to ensure high resolution and strong anti-interference capability.
The technical solution of the invention is as follows:
the utility model provides a simple and easy low-cost wavelength real-time measuring device which characterized in that: the device is composed of an optical fiber interface (1), a single optical fiber collimator (2), a first lens (3), a scattering optical waveguide (4), a constant temperature device (5), a second lens (6), a down-sampling photoelectric detector (7), a signal processing unit (8), a data acquisition card (9) and a computer (10).
The optical fiber interface (1) is connected with the single optical fiber collimator (2);
the single optical fiber collimator (2) and the scattering optical waveguide (4) collimate and focus light and couple the light into the scattering optical waveguide through a first lens (3);
the lens (6) collimates and focuses the divergent light emitted from the scattering optical waveguide (4);
the down-sampling photoelectric detector (7) is welded on the printed circuit board and is connected with the signal processing unit (8);
the scattering optical waveguide (4) is wrapped by a constant temperature device (5);
the signal processing unit (8) is connected with the data acquisition card (9);
the data acquisition card (9) is connected with the computer (10) through a transmission line;
preferably, the optical fiber interface (1) is an FC/APC interface;
preferably, the single optical fiber collimator (2) is a GRIN lens type optical fiber collimator;
preferably, the scattering optical waveguide (4) is a rectangular core optical fiber, the Numerical Aperture (NA) is 0.16, the core size is 120 x 80 μm, the outer diameter of the optical fiber is 125 μm, and the length is 1 m;
preferably, the material of the first lens (3) and the second lens (5) is PMMA (polymethyl methacrylate);
preferably, the downsampling photoelectric detector (7) is an InGaAs four-quadrant detector and consists of four fan-shaped photoelectric detectors, the total photosensitive surface is a circle with the diameter of 2mm, and the response wavelength is as follows: 900-1700nm, cut-off frequency of 30 MHz;
preferably, the signal processing unit (8) is composed of a digital-analog acquisition part and a signal processing part. The invention has the beneficial effects that:
1. the measuring system of the technology of the invention does not need to use an expensive InGaAs camera, uses a down-sampling photoelectric detector to collect the light intensity distribution of speckles, has low cost, small volume and higher refreshing speed which can reach dozens of megabytes, and the spectral refreshing rate of the currently best known linear array CCD spectrometer is only 20-30 kHz.
2. The down-sampling photoelectric detector compresses speckle images with a large amount of information into a small amount of speckle intensity distribution data, thereby greatly reducing the data amount required to be processed by a computer.
3. Compared with the traditional wavelength measurement method, the laser wavelength measurement method based on deep learning has the advantages that the resolution is improved to 4fm (0.5mHz), and the wavelength demodulation speed is improved by at least ten times. In addition, the method has the characteristic of high robustness, ensures the measurement reliability in a complex environment, and widens the application range of the traditional laser wavelength measurement method.
Description of the drawings:
FIG. 1 is a block diagram of the apparatus of the present invention;
FIG. 2 is a graph of data collected by a downsampled photodetector according to the present invention;
FIG. 3 is a speckle intensity distribution plot of the present invention after upsampling;
FIG. 4 is a schematic diagram of the deep learning convolutional neural network topology model of the present invention;
FIG. 5 is a graph of the accuracy and loss of the convolutional neural network training and testing of the present invention;
FIG. 6 is a schematic flow chart of the operation of the present invention;
Detailed Description
The present invention is described in further detail below with reference to the attached drawings.
FIG. 1 is a schematic diagram of a simple and low-cost wavelength real-time measurement device. The device mainly comprises an optical fiber interface (1), a single optical fiber collimator (2), a first lens (3), a scattering optical waveguide (4), a constant temperature device (5), a second lens (6), a down-sampling photoelectric detector (7), a signal processing unit (8), a data acquisition card (9) and a computer (10).
Laser is emitted to a first mirror (3) through a single optical fiber collimator (2) and then coupled into a rectangular optical fiber (4) with the length of one meter, light beams in the optical fiber are emitted to a second lens (6) and then detected by a down-sampling photoelectric detector (7), detected electric signals are processed by a signal processing unit (8) and then collected by a data acquisition card (9), and finally the laser wavelength to be detected is obtained through analysis of a wavelength output module of a computer (10). Wherein the scattering optical waveguide (4) is arranged in the thermostatic device (5) to isolate the influence of the external temperature change. The lens, the scattering optical waveguide and the down-sampled photodetector remain coaxial.
The scattering optical waveguide causes different optical path differences of incident light of different wavelengths transmitted therein and generates different speckles at an output port thereof. The specific type is not limited, and a rectangular optical fiber is preferable in the present invention.
The down-sampling photoelectric detector is used for performing down-sampling photoelectric detection on the speckle images. In the present invention, the shape, type, arrangement and number of pixels of the down-sampling photodetector are not limited. The down sampling is to count the number of pixels of the area array CCD, and the common area array CCD is composed of tens of thousands to millions of pixels. The pixel of the down-sampling photoelectric detector in the patent is only a few to ten pixels, and when the pixel is used for detecting speckle images, the pixel is equivalent to compressing abundant image information into a small amount of intensity distribution data, namely, the speckle images are down-sampled. A four-quadrant photoelectric detector is preferably selected in the invention, so that the speckle image can be uniformly divided into four parts to be respectively detected and output electric signals.
FIG. 2 is a graph of data collected by a downsampled photodetector, and in the present invention, the speckle pattern produced by the scatter waveguide is detected by the downsampled photodetector. The speckle is downsampled to correspond to four parts. Thereby compressing the speckle image with a large amount of information into four data. When the wavelength of the input light is tuned, the intensity distribution of the speckle pattern changes due to the dispersion, and therefore the acquired data also changes. FIG. 2 shows the intensity variation of the detection pixels with 5pm step size in the range of 1547nm-1553 nm. It can be seen that the intensity of each channel varies significantly depending on the wavelength.
Fig. 3 is a speckle intensity distribution graph after up-sampling, and as mentioned above, wavelength information is extracted from the down-sampled speckle intensity distribution data in the present invention. The speckle pattern is compressed into four separate outputs by a downsampled photodetector. The four speckle intensity values are stored in a 2 x 2 matrix. As shown in fig. 3, the original matrix is then extended from 2 x 2 to 100 x 100 using upsampling to maximize the difference between speckle patterns, thereby improving resolution. The upsampling includes two steps of power coding and bicubic interpolation. After normalization processing is carried out on the four original intensity data values, the powers of 1-144 of the four original intensity data values are calculated:
FIG. 4 is a schematic diagram of a deep learning convolutional neural network topology model. The invention applies a deep learning algorithm to reconstruct the wavelength of the near-infrared light wave. The convolutional neural network based on the Alexnet network is tried for the first time, and the constructed convolutional neural network is trained by using speckle intensity distribution values collected under different wavelengths as training samples. The training schedule is shown in fig. 5(a), which shows the accuracy and loss of the validation data set. It can be seen that as the number of iterations increases, the loss and accuracy quickly approach 0 and 100. After a number of iterations they tend to converge gradually. Finally, the accuracy and loss of our constructed convolutional neural network is 100% and 0.0004. In addition, the accuracy of the network is tested by a selected test image, where the wavelength of the light waves of the test set is known but has never been trained. The results of the test are shown in fig. 5(b), and the graph can be regarded as a matrix in which rows and columns represent the predicted wavelength and the true wavelength, respectively. Diagonal and off-diagonal elements correspond to correctly and incorrectly classified wavelengths, respectively. The value of the diagonal elements represents the number of correctly classified images from a given image. The test results show that the resolution of the wavemeter is better than 0.5MHz (4 fm).
As shown in fig. 6, is an operation flowchart of the present invention, which specifically includes the following steps:
in a first step, the speckle intensity distributions of different wavelengths are recorded by means of a down-sampling photodetector (7)
Laser generated by a tunable laser enters a wavelength measuring device through an FC/APC interface (1), each quadrant of a four-quadrant photoelectric detector can receive optical signals by adjusting a lens (6), the optical signals pass through a first quadrant detection domain (a second quadrant detection domain), a third quadrant detection domain (a fourth quadrant detection domain), then are converted into analog signals through a signal processing unit (8), and finally are collected by a data collection card (9). By changing the wavelength of the light source and recording the intensity distribution signals corresponding to different wavelengths.
Secondly, constructing and training a convolutional neural network by using the collected speckle intensity distribution values
The digital signal sets corresponding to different recorded wavelengths are normalized, expanded by power polynomials and integrated and coded into a 12 x 12 data set. And expanding the obtained result to 100 multiplied by 100 through bicubic linear interpolation, and training the constructed convolutional neural network by using speckle intensity distribution values acquired under different wavelengths as training samples. And collecting a speckle intensity distribution data set corresponding to wavelength values except the wavelength of the training set as a test set to perform performance test on the trained convolutional neural network.
Thirdly, measuring the wavelength of the light to be measured by using a convolutional neural network
Light to be measured enters the wavelength measuring device through the FC/APC interface, the device transmits four collected analog signals to a computer, and the wavelength is rapidly extracted through a convolutional neural network which is input, trained and tested, so that the near-infrared light wave wavelength measurement based on deep learning is realized.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (6)
1. A near-infrared speckle wavelength meter mainly comprises a single optical fiber collimator, a scattering optical waveguide, a lens, a down-sampling photoelectric detector and a signal processing unit; the method is characterized in that: the light with the wavelength to be measured sequentially passes through a single optical fiber collimator, a scattered light waveguide and a lens to form a speckle pattern, enters a down-sampling photoelectric detector, is converted into an electric signal and then enters a signal processing unit;
the down-sampling photoelectric detector is used for performing down-sampling photoelectric detection on the speckle image;
the scattering optical waveguide causes different optical path differences of incident lights with different wavelengths transmitted therein and generates different speckles at an exit port thereof.
2. The near-infrared speckle wavemeter of claim 1, wherein: the scattering optical waveguide is a multimode optical fiber.
3. The near-infrared speckle wavemeter of claim 1, wherein: the stray light waveguide is enclosed by a thermostatic device.
4. The near-infrared speckle wavemeter of claim 1, wherein: the down-sampling photoelectric detector is a combination of a plurality of photoelectric sensors which have the same performance and the same external dimension and operate independently, and the sensing surface of the integrated down-sampling photoelectric detector is provided with a circular or rectangular plane.
5. The near-infrared speckle wavemeter of claim 1, wherein: the scattering optical waveguide is coaxial with the lens and is aligned with the center of the sensing surface of the down-sampling photodetector.
6. The near-infrared speckle wavemeter of claim 1, wherein: and performing power polynomial expansion and integrated coding on the photoelectric signals acquired from the downsampling photoelectric detector by adopting power multiplication coding, and performing bicubic interpolation on the coded data to further expand a data set. The method can enhance the characteristic information of the speckle distribution diagram and improve the signal-to-noise ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110072037.5A CN113218516A (en) | 2021-01-20 | 2021-01-20 | Near-infrared speckle wavemeter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110072037.5A CN113218516A (en) | 2021-01-20 | 2021-01-20 | Near-infrared speckle wavemeter |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113218516A true CN113218516A (en) | 2021-08-06 |
Family
ID=77084433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110072037.5A Pending CN113218516A (en) | 2021-01-20 | 2021-01-20 | Near-infrared speckle wavemeter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113218516A (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112097925A (en) * | 2020-10-26 | 2020-12-18 | 杭州菲柏斯科技有限公司 | Optical fiber speckle wavelength meter based on polarization enhancement |
-
2021
- 2021-01-20 CN CN202110072037.5A patent/CN113218516A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112097925A (en) * | 2020-10-26 | 2020-12-18 | 杭州菲柏斯科技有限公司 | Optical fiber speckle wavelength meter based on polarization enhancement |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US4678332A (en) | Broadband spectrometer with fiber optic reformattor | |
| CN106323471B (en) | A kind of hypersensitive spectral method of detection and system based on DLP technology and compressive sensing theory | |
| WO2017070997A1 (en) | High-speed three-dimensional microscopic imaging system and method | |
| CN108007582B (en) | Laser wavelength detection method and device based on Rayleigh Brillouin scattering | |
| CN108827471B (en) | Diffraction element, high-resolution spectrometer and spectrum detection method | |
| CN101923052B (en) | Infrared spectrum type MEMS gas sensor based on filter structure light splitting | |
| CN201589623U (en) | Demodulation device for fiber grating sensors | |
| CN110632002A (en) | Aperture coding spectrum detection device based on compressed sensing | |
| CN109540293A (en) | A kind of spectrum detection device based on fiber grating | |
| CN112697274A (en) | Single-capture spectrum measurement method and device | |
| CN109696409A (en) | A kind of spectral scan probe | |
| CN113063735A (en) | Surface plasmon resonance sensing system based on chromatic dispersion | |
| CN113418902A (en) | On-chip Raman detection system | |
| CN113218516A (en) | Near-infrared speckle wavemeter | |
| CN106840008A (en) | A kind of optical fiber distance measurement system and measuring method | |
| CN103968943A (en) | Accurate measurement method for signal-to-noise ratio of fiber optic spectrometer | |
| CN111272285A (en) | Polarization state measurement analyzer with high-speed real-time response | |
| CN201130028Y (en) | Optical fibre grating sensor demodulation instrument | |
| CN206670570U (en) | A kind of optical fiber distance measurement system | |
| CN112986177A (en) | Near-infrared optical fiber spectrometer based on novel spectrum acquisition structure | |
| CN108151877A (en) | A kind of micro-hole spectrometer and spectrum reconstruction method | |
| CN111458028B (en) | Built-in chip spectrum appearance module is carried to cell-phone | |
| CN211148422U (en) | Aperture coding spectrum detection device based on compressed sensing | |
| CN209264116U (en) | A kind of spectrum detection device based on fiber grating | |
| CN115078266A (en) | Optical system and design method thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210806 |
|
| WD01 | Invention patent application deemed withdrawn after publication |