Spectrometer high-precision wavelength calibration method based on optical filter hardware interception

Abstract

The invention discloses a spectrograph high-precision wavelength calibration method based on optical filter hardware interception, and the method comprises the steps: placing an optical filter in an interference light path, limiting the spectrum range in a hardware manner, and enabling a camera to directly collect an interference spectrum signal with a sharp boundary; extracting wave crest and wave trough positions of the interference spectrum signal, calculating a local half cycle and fitting to obtain a half cycle change curve; calculating a phase gradient to obtain a wavenumber gradient; solving a relative function of wave numbers through an integral accumulation mode; establishing a mapping relation between a pixel coordinate and a wavelength by combining cut-off wavelength information of the optical filter, and completing wavelength calibration of the spectrograph; performing interpolation resampling on the interference spectrum signal based on a calibration result; and Fourier transform is carried out on the re-sampled interference spectrum signal to obtain depth information. According to the invention, the problem that the existing calibration method depends on an expensive light source, and is insufficient in precision or high in complexity can be solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of optical measurement and spectral analysis, and particularly to a high-precision wavelength calibration method for a spectrometer based on optical filter hardware interception. BACKGROUND

[0002] There is a wavelength calibration problem in the existing spectral domain optical coherence tomography (SD-OCT) system. Since the spectrometer has a nonlinear sampling in the wave number domain when collecting the interference spectrum signal, the frequency domain signal is expanded after Fourier transform, thereby affecting the imaging quality of the system. In order to realize accurate mapping from the wavelength domain to the wave number domain, an accurate "pixel coordinate-wavelength" corresponding relationship needs to be established. However, the spectrometer can only obtain the spectral intensity of each pixel point, and cannot directly obtain the corresponding wavelength information. The existing calibration method either relies on expensive multi-spectral light sources or lasers, which has high equipment requirements and large system complexity, or relies on ideal physical models, and the calibration accuracy is difficult to guarantee. Therefore, the existing technology has deficiencies in the accuracy, simplicity and applicability of wavelength calibration.

[0003] The traditional wavelength calibration method of the spectrometer usually uses a light source with multiple characteristic spectral lines as a calibration light source, makes it incident to the spectrometer, and identifies the pixel coordinate position corresponding to each known wavelength in the collected original data, thereby obtaining a set of "pixel coordinate-wavelength" discrete data. Subsequently, a cubic polynomial fitting is performed by using the least square method to establish a function relationship of "pixel coordinate-wavelength". In order to ensure the calibration accuracy, the calibration light source is required to have a large number of characteristic spectral lines with known wavelengths and stability. However, since the function relationship is fitted based on limited data points, there is inevitably a local deviation, and the requirements for the calibration light source and other equipment are high, so it is difficult to achieve high-precision wavelength calibration.

[0004] In the SD-OCT system, there is also a method of calibrating the wavelength of the spectrometer by a physical model, which avoids the dependence on the special calibration light source. However, its implementation relies on multiple idealized assumptions made on the optical path of the spectrometer and the broadband light source, which makes the actual system different from the ideal model, thereby leading to a large deviation of the calibration result.

[0005] In addition, the existing wavelength calibration method of the spectrometer based on interference spectrum improves the accuracy of wavelength calibration to some extent, and avoids the use of special calibration light sources. However, this method still needs to use a laser with known wavelength to determine the minimum wavelength and the maximum wavelength of the spectral range, which not only increases the complexity of the system structure, but also depends on additional equipment, limiting the application of this method in simple spectrometers.

[0006] In summary, the wavelength calibration methods of the existing spectrometer have the following shortcomings: the calibration method based on characteristic spectral line light source relies on high-quality calibration light source, requires high equipment and has limited fitting accuracy; the calibration method based on physical model needs multiple idealized assumptions, resulting in deviation between actual calibration result and true situation; and the calibration method based on interference spectrum improves accuracy to some extent and avoids the use of calibration light source, but still needs to use a laser with known wavelength to obtain the boundary of the spectral range, increasing the system complexity and dependence on additional large equipment. Therefore, the existing wavelength calibration methods still have shortcomings in accuracy, simplicity and equipment dependence, and need to be further improved. SUMMARY

[0007] The present application provides a high-precision wavelength calibration method for a spectrometer based on filter hardware interception, which is suitable for a spectral domain optical coherence tomography (SD-OCT) system. By intercepting the interference spectrum signal, fitting the half cycle, calculating the phase gradient, solving the wave number function and mapping the endpoints, an accurate correspondence between the pixel coordinates and the wavelength is established, thereby solving the problem of dependence on expensive light sources, insufficient accuracy or high complexity of the existing calibration methods.

[0008] A high-precision wavelength calibration method for a spectrometer based on filter hardware interception, comprising the following steps: (1) placing a filter in the interference light path to limit the spectral range in a hardware manner, so that the camera directly collects interference spectrum signals with sharp boundaries; (2) extracting the peak and valley positions of the interference spectrum signal, calculating the local half cycle and fitting to obtain the half cycle change curve; (3) calculating the phase gradient according to the half cycle change curve, and obtaining the wave number gradient therefrom; (4) based on the wave number gradient, the relative function of the wave number is obtained by integral accumulation; (5) according to the relative function of the wave number, the mapping relationship between the pixel coordinates and the wavelength is established in combination with the cutoff wavelength information of the filter, and the wavelength calibration of the spectrometer is completed; (6) after completing the wavelength calibration of the spectrometer, the interference spectrum signal is interpolated and resampled based on the calibration result; (7) Fourier transforming the resampled interference spectrum signal to obtain the depth information.

[0009] In step (1), the filter is placed in the interference light path, specifically: Long-pass and short-pass filters are placed in the interference light path, and the incident light beam passes through the two filters vertically before entering the interferometer and the spectrometer.

[0010] Preferably, the cutoff wavelength attenuation width of the filter is not more than 2 nm.

[0011] The specific process of step (2) is as follows: The pixel coordinates corresponding to the wave peak and wave trough in the interference spectrum signal are extracted through an automatic peak search algorithm; the coordinate difference of adjacent peaks and troughs, i.e., the local half period, is calculated; and then the coordinate difference of adjacent peaks and troughs is fitted using the least square method to obtain the half period change curve of the interference spectrum.

[0012] In step (3), the calculation of the phase gradient and the wave number gradient is extended to the continuous pixel coordinate domain through an interpolation method, and the discrete data is smoothed.

[0013] In step (4), the relative function of the wave number is obtained by integrating the wave number gradient along the pixel coordinate; since the pixel coordinate is discrete, the solution is realized by accumulation.

[0014] In step (5), the mapping relationship between the pixel coordinate and the wavelength is established, and the formula is as follows: ; Wherein, is the true wavelength at the pixel coordinate ; is the cut-off wavelength value of the short-pass optical filter; is the cut-off wavelength value of the long-pass optical filter; is the pixel coordinate of the end point of the sampling interval; represents the cumulative wave number increment from the starting pixel of the sampling interval to the th pixel; represents the total increment of the wave number in the entire sampling interval, which is used as a normalization denominator to map the local cumulative quantity to the overall proportional range of .

[0015] In step (6), a linear interpolation function is used to obtain the wave number change curve according to the calibration result and the linearly distributed wave number change curve, and the interference spectrum signal is subjected to interpolation resampling processing.

[0016] Compared with the prior art, the present application has the following beneficial effects: 1. Avoiding dependence on expensive light sources. The present application introduces long-pass optical filters and short-pass optical filters with known cut-off wavelengths into the optical path to realize hardware interception of the interference spectrum signal, which can directly determine the spectral range boundary without relying on single-frequency lasers or multi-spectral calibration light sources, thereby effectively reducing system complexity and equipment cost.

[0017] 2. Improving calibration accuracy and reliability. The cut-off wavelength parameters of the optical filters are stable, and the actual test shows that the attenuation width of the cut-off position is only about 2 nm, which can accurately limit the starting wavelength and the ending wavelength of the interference spectrum, ensuring the determination accuracy of the calibration boundary and improving the overall accuracy and repeatability of the wavelength calibration.

[0018] 3. Improve signal quality and imaging effect. By half cycle fitting, phase gradient calculation and wave number domain interpolation resampling of the interference spectrum, the method effectively eliminates the frequency domain expansion problem, makes the depth signal after Fourier transform more concentrated, the resolution is higher, and the imaging quality is significantly improved.

[0019] 4. Simple and easy to implement, strong popularization. The method of the present application does not need complex physical model assumption, the hardware structure is simple, the software processing steps are clear, and it is easy to implement. At the same time, the method is not only suitable for spectral domain optical coherence tomography system, but also can be popularized to other interference measurement and spectrum analysis systems based on spectrometer, and has good applicability and popularization value. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0021] Figure 1 A flow chart of a high-precision wavelength calibration method for a spectrometer based on filter hardware interception in an embodiment of the present application.

[0022] Figure 2 A schematic diagram of the optical path structure of the spectral domain optical coherence tomography system in the embodiment of the present application.

[0023] Figure 3 Cut-off wavelength attenuation curves of long-pass and short-pass filters used in the embodiment of the present application.

[0024] Figure 4 Interference spectrum signals collected by the system before and after placing the filter in the embodiment of the present application.

[0025] Figure 5 An intermediate result diagram in the spectrometer calibration based on the interference spectrum signal in the embodiment of the present application.

[0026] Figure 6 Interference spectrum wave number domain mapping results and Fourier transform results in the embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings of the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0028] It should be noted that, unless otherwise specified, the features in the following embodiments and implementation methods can be combined with each other.

[0029] like Figure 1 As shown, a high-precision wavelength calibration method for a spectrometer based on filter hardware interception includes the following steps: Step 1: Place a filter in the interference optical path to limit the spectral range in hardware, making the interference spectral boundaries sharp and ensuring the signal quality of the calibration area.

[0030] like Figure 2 As shown, the optical path structure of the spectral domain optical coherence tomography system includes a superradiative diode light source 1, a long-pass filter 2, a short-pass filter 3, a beam splitter 4, a first lens 5, a sample mirror 6, a second lens 7, a reference mirror 8, a grating 9, a third lens 10, and a camera 11. The beam emitted by the broadband light source carries the relative position information of the sample and the reference mirror after passing through the sample arm and the reference arm. This relative position information is then converted into interference spectral information by a spectrometer. The interference spectral signal acquired by the camera 11 can be expressed as: ; In the formula, it is assumed that the sample has a total of One reflective surface; and These are the reference mirror and the first The reflectivity of the reflective surface of the layer sample; The total luminous intensity of the light source; and These are the beam ratios reaching the reference arm and the sample arm, respectively, therefore we have ; Wave number; The optical path length of the reference arm beam; For the first The optical path length of the beam reflected back from the reflective surface of the sample layer; For the first The optical path of the beam reflected back from the reflective surface of the sample layer.

[0031] To more accurately calibrate the wavelength of the spectrometer system, the sample uses a mirror identical to the reference mirror to ensure high contrast in the interference fringes. Under this sample setting, after removing the DC and autocorrelation terms, the interference spectrum signal acquired by the camera can be expressed as: ; In the formula, and These are the reflectivities of the reference mirror and the sample mirror, respectively. and These represent the optical path lengths of the reference arm and sample arm beams, respectively. Since the grating in the spectrometer achieves spectral dispersion based on the diffraction angle, the sampling at the pixel coordinates by the camera does not correspond non-linearly to the wavenumber, resulting in chirped (non-linear phase) characteristics in the acquired interference spectral signal. Therefore, the interference spectral signal actually acquired by the camera can be modeled as: ; In the formula, The optical path difference between the reference arm and the sample arm beams; These are pixel coordinates; due to the non-uniform wavenumber sampling of the interference spectrum by the spectrometer, It is a nonlinear sampling function with respect to wavenumber, and Therefore, we have: ; In SD-OCT systems, superluminescent diodes (SLDs) with a Gaussian spectral distribution are typically used as the light source, with their energy concentrated primarily at the corresponding wavelengths across the full width at half maximum (FWHM). For example... Figure 4 As shown in (a), the contrast of the interference signal is poor at both ends of the interference spectrum, making it unsuitable for wavelength calibration and optical path difference reconstruction. Therefore, only the middle portion of the interference spectrum with better quality is retained. This truncation process is generally implemented through software; however, the starting and ending wavelengths after truncation cannot be obtained directly.

[0032] Traditional methods typically use a single-frequency laser of known wavelength as a reference, determining the starting and ending wavelengths acquired by the camera by obtaining interference spectral signals under two different optical path differences. However, this method requires an additional laser, increasing system cost and complexity.

[0033] To address the aforementioned issues, this invention introduces a long-pass filter and a short-pass filter into the interference optical path to define the effective wavelength range of the interference spectrum in hardware. The cutoff wavelength of the filter is a known parameter and can serve as a reference for determining the wavelength boundary. Preferably, the long-pass filter and short-pass filter define the start and end wavelengths of the spectrum, respectively, with an attenuation width at the cutoff wavelength not exceeding 2 nm to ensure the accuracy of wavelength boundary determination. In this embodiment, the long-pass filter and short-pass filter can be models FELH0825 and FESH0850 manufactured by Thorlabs, whose cutoff characteristics are shown in Figure 3. These filters enable high-precision, repeatable spectral boundary definition, thereby improving the reliability and convenience of the calibration process.

[0034] like Figure 2 As shown, the filters include a long-pass filter 2 and a short-pass filter 3. The incident light beam passes perpendicularly through the two filters before entering the interferometer and spectrometer. Ultimately, the camera 11 can directly acquire the interference spectral signal with sharp boundaries, such as... Figure 4As shown in (b), its corresponding sampling interval is... The corresponding pixel coordinate range is .

[0035] Step 2: Extract the peak and trough positions of the interference spectrum signal, calculate the local half-period, and fit the half-period variation curve.

[0036] right Figure 2 Interference spectral signals acquired by the camera in the system Perform extreme value detection to obtain the peak and trough positions arranged in ascending order of pixel size. ,like Figure 5 As shown in (a). Define the first The length of the local half-cycle and the first The pixel coordinates of each discrete sampling point are: ; ; In the formula, and The first and the The pixel coordinates of each peak and valley; For the first The pixel coordinates of the nth discrete sampling point, i.e., the nth The center pixel coordinates of each local half-cycle. Perform least squares fitting and on continuous pixel coordinates Up-interpolation resampling yields the half-period variation curve of the interference spectrum. The result is as follows Figure 5 As shown in (b).

[0037] Step 3: Calculate the phase gradient based on the half-cycle variation curve, and then obtain the wavenumber gradient.

[0038] Step 3.1, in each half-cycle Within, the phase increment is approximately Therefore, the phase gradient at each discrete sampling point is approximately: ; In the formula, Discrete sampling points Phase gradient at; The phase of the interference signal; These are pixel coordinates; For the first The pixel coordinates of a discrete sampling point; For the first The length of a local half-cycle.

[0039] Step 3.2: Using interpolation methods, the discretely sampled phase gradient values ​​are... Extended to continuous pixel coordinates This allows us to obtain the continuous phase gradient function corresponding to the pixel coordinates: ; In the formula, For continuous pixel coordinates The phase gradient function at a given point, and its relationship with the half-cycle variation curve. It is directly proportional to the reciprocal; This is an interpolation operator.

[0040] Step 3.3, based on the phase function expression in Step 1 The mathematical analytical form with phase gradient: ; By combining the phase gradient calculated from the half-period variation curve of the interference spectrum, the wavenumber gradient and the phase gradient can be directly correlated: ; therefore, Figure 5 The half-cycle variation curve shown in (b) Taking the reciprocal yields a curve proportional to the wavenumber gradient. Figure 5 (c)

[0041] Step 4: Based on the wavenumber gradient, obtain the relative function of the wavenumber by integral accumulation.

[0042] Step 4.1: The relative function of wavenumber can be obtained by integrating the wavenumber gradient along the pixel coordinates. Since the pixel coordinates are discrete, the solution is achieved by accumulation. ; In the formula, The wavenumber relative function in pixel coordinates The value at that location.

[0043] Step 4.2, define the initial value of the relative wavenumber function. And its total increment is: ; In the formula, These are the pixel coordinates of the endpoints of the sampling interval.

[0044] Step 5: Combine the cutoff wavelength information of the filter to establish the mapping relationship between pixel coordinates and wavelength.

[0045] Step 5.1: The wavelength values ​​corresponding to the endpoints of the sampling interval can be determined by using the parameters of the long-pass and short-pass filters. Therefore: ; In the formula, Coordinates of the endpoints of the sampling interval The corresponding wave value; This corresponds to the wavelength value, i.e., the cutoff wavelength value of the short-pass filter; Coordinates of the endpoints of the sampling interval The corresponding wave value; This corresponds to the wavelength value, i.e., the cutoff wavelength value of the long-pass filter.

[0046] Step 5.2, perform endpoint mapping: ; In the formula, pixel coordinates The true wavenumber at that location; This represents the total increment of the relative wavenumber function; pixel coordinates The relative wavenumber function value at that location; For the sampling interval, the first pixel coordinates; These are the pixel coordinates of the endpoints of the sampling interval.

[0047] Based on the quantitative relationship between wavelength and wavenumber This yields the pixel-to-wavelength mapping: ; In the formula, pixel coordinates The actual wavelength at that location; Indicates the distance from the starting pixel of the sampling interval to the... The cumulative wavenumber increment per pixel; This represents the total increase in wavenumber over the entire sampling interval, used as the normalization denominator to map the local cumulative values ​​to... Within the overall proportion range. The calibration results are as follows: Figure 5 As shown in (d).

[0048] Step 6: Complete the wavelength calibration of the spectrometer and perform interpolation resampling on the interference spectral signal based on the calibration results.

[0049] In this embodiment, the original interferometric spectral signal is interpolated and resampled to the wavenumber domain. Specifically, a linear interpolation function is used, and the interferometric spectral signal is interpolated and resampled based on the wavenumber variation curve obtained from the calibration results and the linearly distributed wavenumber variation curve. The comparison results of the interferometric spectral signal before and after interpolation and resampling are as follows: Figure 6 As shown in (a).

[0050] Step 7: Perform a Fourier transform on the resampled signal to obtain high-quality depth information.

[0051] The processed interference spectrum signal was subjected to a Fourier transform and compared with the signal without interpolation and resampling. The result is as follows: Figure 6 As shown in Figure (b), the method of the present invention effectively solves the problem of signal broadening in the frequency domain, thereby significantly improving the imaging quality of the system.

[0052] The embodiments described above provide a detailed explanation of the technical solutions and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A high-precision wavelength calibration method for a spectrometer based on filter hardware interception, characterized in that, Includes the following steps: (1) Place a filter in the interference optical path to limit the spectral range in hardware so that the camera can directly acquire the interference spectral signal with sharp boundaries; (2) Extract the peak and trough positions of the interference spectrum signal, calculate the local half-period and fit the half-period variation curve; (3) Calculate the phase gradient based on the half-cycle variation curve, and obtain the wavenumber gradient from it; (4) Based on the wavenumber gradient, the relative function of the wavenumber is obtained by integral accumulation; (5) Based on the relative function of wavenumber and combined with the cutoff wavelength information of the filter, establish the mapping relationship between pixel coordinates and wavelength to complete the wavelength calibration of the spectrometer; (6) After completing the wavelength calibration of the spectrometer, the interference spectrum signal is interpolated and resampled based on the calibration results; (7) Perform Fourier transform on the resampled interference spectrum signal to obtain depth information.

2. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1, characterized in that, In step (1), a filter is placed in the interference optical path, specifically as follows: A long-pass filter and a short-pass filter are placed in the interference optical path. The incident beam passes perpendicularly through the two filters and then enters the interferometer and spectrometer.

3. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1 or 2, characterized in that, The cutoff wavelength attenuation width of the filter does not exceed 2 nm.

4. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1, characterized in that, The specific process of step (2) is as follows: The pixel coordinates corresponding to the peaks and valleys in the interference spectrum signal are extracted by an automatic peak search algorithm; the coordinate difference between adjacent peaks and valleys, i.e., the local half-period, is calculated; and the coordinate difference between adjacent peaks and valleys is then fitted using the least squares method to obtain the half-period variation curve of the interference spectrum.

5. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1, characterized in that, In step (3), the calculation of phase gradient and wavenumber gradient is extended to the continuous pixel coordinate domain by interpolation method, and the discrete data is smoothed.

6. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1, characterized in that, In step (4), the relative function of wavenumber is obtained by integrating the wavenumber gradient along the pixel coordinates; since the pixel coordinates are discrete, the solution is achieved by accumulation.

7. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1, characterized in that, In step (5), the mapping relationship between pixel coordinates and wavelength is established, and the formula is: ； in, pixel coordinates The actual wavelength at that location; This is the cutoff wavelength value of the short-pass filter; This is the cutoff wavelength value of the long-pass filter; These are the pixel coordinates of the endpoints of the sampling interval; Indicates the distance from the starting pixel of the sampling interval to the... The cumulative wavenumber increment per pixel; This represents the total increase in wavenumber over the entire sampling interval, used as the normalization denominator to map the local cumulative values ​​to... Within the overall proportion range.

8. The high-precision wavelength calibration method for spectrometers based on filter hardware interception according to claim 1, characterized in that, In step (6), a linear interpolation function is used to perform interpolation and resampling processing on the interference spectrum signal based on the wavenumber variation curve obtained from the calibration results and the wavenumber variation curve of the linear distribution.

Images