A method for combining signals from optical imaging spectrometers employing optical fibers

Abstract

The present application relates to the technical field of merging of optical imaging spectrometer signals, and particularly relates to a method for merging of optical imaging spectrometer signals using optical fibers. The technical scheme comprises the following method steps: light source illumination and data acquisition, using a light source to uniformly illuminate the entrance of the optical fiber, and a 0th order data is generated by the optical fiber spectrometer detector; spatial dimension data selection and peak value identification, selecting a column of spatial dimension data from the 0th order data, and using an AMPD algorithm to identify the peak value and distinguish the spatial dimensions corresponding to different optical fibers. The present application ensures accurate data acquisition by reasonably selecting the light source and adjusting the intensity, efficiently distinguishes the optical fibers using the AMPD algorithm, has strong noise resistance, flexibly adapts to different transmission conditions using the super-Gaussian fitting, and automatically adjusts the merging range using the new baseline selection standard to improve the signal quality and accuracy. Moreover, the present application can directly determine the merging area from the 0th order data, improves the processing efficiency, reduces the noise, and provides reliable support for spectral analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of merging signals of optical imaging spectrometers, and particularly relates to a method for merging signals of optical imaging spectrometers using optical fibers. BACKGROUND

[0002] An optical imaging spectrometer using optical fibers (hereinafter referred to as fiber-optic spectrometer) is a spectral analysis instrument that uses optical fibers as a light transmission medium to conduct external light sources into the spectrometer for analyzing the intensity distribution of light at different wavelengths. The core diameter of the fiber outlet matches the slit width of the spectrometer, and through precise optical axis alignment and divergence angle control, it ensures that the fiber output light can effectively enter the slit. When there are multiple optical fibers, due to the cladding thickness of the optical fibers playing a physical isolation role in the optical fiber array, the light entering the spectrometer slit through the optical fibers is not a continuous light beam, but is composed of multiple segmented light beams. Ideally, the CCD image plane of the spectrometer presents segmented uniform and clear spots in the spatial dimension.

[0003] In fact, due to problems in the manufacturing process and assembly process of optical fibers, the CCD image plane of the spectrometer presents segmented non-uniform spots in the spatial dimension. When processing the light intensity signals collected by the CCD, the spatial dimension binning method is usually adopted to improve the signal-to-noise ratio of the signals. For non-uniform spots, the selection of different numbers of signal merging in the spatial dimension affects the final data quality.

[0004] A fiber-optic spectrometer is an optical analysis instrument based on fiber-optic transmission technology, used to guide light signals through optical fibers into a spectrometer for analyzing and measuring information such as light wavelength and intensity. Fiber-optic spectrometers are currently the most common type of spectrometer. The flexibility of optical fibers allows them to be easily integrated into various measurement scenarios, adapting to different experimental environments and application requirements; compared to traditional spectrometers, fiber-optic spectrometers are generally more compact and can be designed as portable devices, suitable for field and mobile monitoring. Fiber-optic spectrometers are used in environmental monitoring, industrial process control, medical and biological fields, astronomy and remote sensing, chemical analysis and experimental research, etc.

[0005] The fiber-optic spectrometer is composed of a fiber-optic delivery system, a slit, a dispersive element, a detector, and a data processing system. The fiber-optic delivery system typically consists of a custom optical lens, a fiber-optic cable, and a fiber-optic interface. The fiber-optic cable is responsible for collecting and transmitting optical signals. It usually includes a core for light transmission and a cladding for isolation and reflection. The fiber-optic interface collects light from the light source at the input end and transmits it to the slit of the spectrometer at the output end. When the light beam output by the fiber-optic cable enters the spectrometer, it passes through the slit for spatial filtering, limiting the width of the light beam to improve spectral resolution. The dispersive element is usually a grating that breaks down the incoming white light into different wavelengths of light, forming a spectrum. The detector is usually a CCD or CMOS sensor that detects different wavelengths of light after dispersion and converts them into electrical signals. The data processing system converts the electrical signals output by the detector into digital signals for processing and analysis of spectral data.

[0006] After the light signal transmitted by the fiber-optic cable passes through the slit of the spectrometer, a two-dimensional light spot image is formed on the CCD. One dimension of the CCD records the wavelength information of the spectrum (dispersion dimension), while the other dimension records the spatial information of the light transmission in the fiber-optic array (i.e., the light signals transmitted by different positions in the fiber-optic array).

[0007] In the fiber-optic spectrometer, the coupling relationship between the fiber-optic output end, the slit, and the imaging on the CCD image plane is crucial to the performance of the spectrometer, especially in terms of spectral resolution, light flux, and imaging quality. The coupling relationship between these components is as follows: 1. Coupling between the fiber-optic output end and the slit. 2. Imaging relationship between the slit and the CCD image plane.

[0008] The output end of the fiber-optic cable transmits light to the dispersive system of the spectrometer through the slit, which limits the light flux and improves the spectral resolution in this process. The matching between the fiber-optic output end and the slit is as follows: 1. Matching between the fiber-optic core diameter and the slit width: the core diameter of the fiber-optic cable determines the size of the light beam transmitted by the fiber-optic cable, while the width of the slit limits the angle and intensity of the light beam entering the spectrometer. 2. Beam collimation: the light output by the fiber-optic cable is usually divergent and needs to be collimated by a collimating lens to form a parallel light beam that passes through the slit. After the collimated light beam passes through the slit, it enters the dispersive system. The degree of collimation affects the transmission efficiency of the light and the quality of the imaging.

[0009] 3. Flatness and alignment accuracy of the fiber-optic end face: the flatness of the fiber-optic output end (such as the quality of cutting or polishing) and its alignment accuracy with the slit affect the coupling efficiency.

[0010] The slit, dispersive element, and detector in the spectrometer together determine the resolution of the spectrometer and the quality of the spectral image. Imaging of the slit on the detector image plane: the width of the slit not only determines the spectral resolution but also affects the imaging of the light beam on the CCD.

[0011] Without considering the spectrometer itself, the coupling of the fiber and the slit can lead to poor coupling effect due to the following factors, resulting in low transmission efficiency of optical signals and non-ideal spot formation on the CCD image plane. 1. The core diameter of the fiber does not match the width of the slit: If the core diameter of the fiber is too small, and the slit is too wide, the light beam entering the spectrometer will be divergent or insufficient, resulting in uneven shape and intensity distribution of the spot on the CCD; on the contrary, if the core diameter of the fiber is too large, part of the light beam cannot enter the slit, causing signal loss. The spot on the CCD image plane is blurred, incomplete or weakened. 2. The alignment error of the fiber outlet and the slit: If the fiber outlet end is not accurately aligned with the slit, the light beam will deviate from the predetermined light path, causing part of the light to fail to pass through the slit into the spectrometer. The spot on the CCD deviates or disappears partially, resulting in incomplete light beam. 3. The arrangement of the fiber array is not uniform: If the arrangement of multiple fibers is not uniform or incorrect, it may cause the light beam to be partially blocked or deviated when entering the slit. This improper arrangement can cause irregular spot array on the CCD, even causing intensity distribution difference between the spots. The spot array is irregular or the spacing is uneven, some spots are strong and some spots are weak. 4. The divergence angle of the fiber does not match the acceptance angle of the slit: If the divergence angle is too large, the light beam entering the slit of the spectrometer will be excessively divergent, and part of the light may be blocked by the slit; if the divergence angle is too small, the acceptance range of the slit cannot be fully utilized, resulting in a decrease in coupling efficiency. The intensity of the spot on the CCD is weakened, the spot is diffused or too concentrated. 5. The quality problem of the fiber outlet end surface: If the fiber outlet end surface has scratches, contamination or damage, light transmission will occur scattering and diffraction, resulting in irregular light beam, affecting the coupling efficiency. The shape of the spot on the CCD image plane is irregular, the edges of the spot are blurred or noisy, and the overall intensity of the spot array is reduced. 6. Micro-bending or stretching of the fiber: If the fiber is subjected to mechanical stress (such as excessive bending, stretching) during transmission, it may cause changes in the internal light path of the fiber, resulting in transmission loss and change in beam divergence angle, ultimately affecting the coupling. The intensity of the spot on the CCD is reduced or uneven, the shape of the spot becomes irregular, and even the fiber does not output light. 7. Problems with the fiber fixing and alignment device: If the fixing device between the fiber and the slit is loose or the alignment mechanism is unstable, it may cause slight deviation or vibration of the fiber relative to the slit, affecting the coupling effect. The position of the spot is unstable, and the shape and intensity of the spot change over time. 8. Fiber cladding reflection problem: If the cladding material and structure of the fiber are not ideal, or the fiber is not properly contacted with the external environment (such as damaged or contaminated cladding), it may cause the reflection efficiency inside the fiber to be reduced, and the light beam transmission process to be loss. The intensity of the spot on the CCD is generally low, and even some fibers do not output light at all.

[0012] As Figure 1Fig. 4 shows the curves of single light ray in the spatial dimension of the two-dimensional CCD image plane when the fiber entrance is uniformly illuminated by sunlight. (a) The output of the fiber is relatively smooth and uniform, and the light intensity distribution has no obvious fluctuations or unevenness. Such output means that the coupling of the fiber is good, and the light is uniformly distributed in the spatial dimension without obvious light loss or interference. (b) The light intensity fluctuates with the change of the spatial position, but the overall still has good continuity. (c) The fluctuation becomes larger, and the light intensity at different positions appears obvious difference, indicating that there is more significant coupling failure or fiber surface defect in the fiber, which causes different transmission efficiency of the light beam in different regions. (d), (e), (f) As the coupling effect becomes worse, the spatial distribution of light intensity gradually presents Gaussian distribution. The light intensity distribution appears a central peak, but the two sides have attenuation, indicating that the coupling of the fiber gradually concentrates in a certain area, which may be due to the inconsistency of the fiber core and the cladding or the misalignment of the geometric center of the fiber. The light intensity is concentrated in a narrow area, and the light intensity on both sides decays rapidly, which may reflect that the fiber has more serious non-uniformity in the transmission process, or even part of the light loss. (g) The light intensity spatial distribution of three different fibers in succession. It can be clearly seen that there is a large difference in light intensity between different fibers.

[0013] The light in the fiber enters the slit and forms a two-dimensional image on the detector after passing through the beam splitter. For most spectral analysis, the wavelength information is the key research object, and the spatial distribution of the fiber does not carry information useful for spectral analysis. Therefore, the CCD image in the spatial dimension is merged, the spatial information related to the position of the fiber is compressed, and the wavelength information of the spectrum is focused. The CCD image plane is usually affected by noise (such as readout noise, thermal noise, etc.) and signal intensity fluctuations. Summing the signals of multiple pixels corresponding to the same wavelength can increase the signal intensity, while the noise is usually randomly distributed. After merging, the proportion of noise to signal decreases, reducing the random noise of the measurement, thereby enhancing the signal quality.

[0014] For a spectrometer, the one-dimensional spatial scale of the slit corresponds to the spatial dimension in the two-dimensional detector image. When there is no fiber, the spatial dimension is continuous, and continuous spatial dimension is usually used for merging. When a fiber is used, the spatial dimension is separated by the fiber cladding. One merging method is to select a single fiber spatial dimension from the image according to the dark line generated by the fiber cladding, and then average the spatial dimension. Another merging method is to perform Gaussian fitting on the curve of the spatial dimension of a single fiber at a single wavelength, and select the part above the full width at half maximum as the region for merging the spatial dimension.

[0015] In summary, the present application proposes a merging method for the signal of an optical imaging spectrometer using a fiber. SUMMARY

[0016] The purpose of the present application is to solve the problem of the optical imaging spectrometer using optical fiber in the background art, which is caused by the manufacturing and assembly process of optical fiber, and the coupling effect is not good due to various factors in the coupling of optical fiber and slit, resulting in non-uniformity of CCD image spot, affecting the transmission efficiency of optical signal and data quality, and the existing space dimension merging method has defects and cannot effectively deal with the problem of non-uniform spot and noise, and a merging method for signals of an optical imaging spectrometer using optical fiber is provided.

[0017] The technical scheme of the present application is a merging method for signals of an optical imaging spectrometer using optical fiber, comprising the following method steps:

[0018] Light source illumination and data acquisition, using a light source to uniformly illuminate the entrance of the optical fiber, and the optical fiber spectrometer detector generates 0-level data;

[0019] Space dimension data selection and peak value identification, selecting a column of space dimension data illuminated from the 0-level data, and adopting an AMPD algorithm for peak value identification to distinguish different space dimensions corresponding to different optical fibers;

[0020] Super Gaussian fitting and baseline selection, super Gaussian fitting is performed on the data of a single optical fiber, and the part above the baseline is selected as the space merging area according to the optimized baseline

[0021] Space dimension merging, merging the space dimensions of the optical fibers into one row, and the merging mode is to average multiple rows.

[0022] Optionally, in the light source illumination and data acquisition, specifically comprising:

[0023] The light source is selected in relation to the wavelength parameter of the spectrometer, and for the ultraviolet wavelength spectrometer, any one of sunlight, mercury lamp and xenon lamp can be used, and the intensity of the light source is adjusted by adjusting the integration time of the detector to ensure that the imaging of the spectrometer detector does not appear overexposure.

[0024] After the light source illuminates the entrance of the optical fiber, the light passes through the optical fiber, the slit and the beam splitter to form a two-dimensional spot on the detector, generating 0-level data.

[0025] Optionally, in the space dimension data selection and peak value identification, when the AMPD algorithm is applied to the 0-level data, the light intensity value is taken as negative and then calculated, the cladding between adjacent optical fibers causes a decrease in light intensity, which is a trough, and the boundary between the two optical fibers is found after taking the negative value.

[0026] Optionally, in the super Gaussian fitting and baseline selection, after distinguishing different optical fibers, the data of a single optical fiber in the space dimension is obtained, and super Gaussian fitting is performed on the data,

[0027] The expression of super Gaussian is ​

[0028] where parameters ω, k determine the shape and width of the super-Gaussian,

[0029] When k=2, the line type is a standard Gaussian;

[0030] When k<2, the line type gradually sharpens in the middle;

[0031] When k>2, the line type gradually flattens in the middle;

[0032] The baseline is selected as a reference for the merging of the spatial dimension data, and the baseline selection criteria The baseline expression is similar to the half-width expression, but the value is affected by the line type parameter k. When the value of k is greater than 2, the baseline is smaller than the half-width, and the selected spatial dimension quantity is smaller than the spatial dimension quantity selected by the half-width; when the value of k is less than 2, the opposite is true.

[0033] Optionally, the spatial dimension merging specifically includes intersecting the baseline with the fitting curve and solving after the baseline is fitted, obtaining the part above the baseline as the selection of the merged spatial dimension, and finally merging the spatial dimensions corresponding to the optical fibers in a multi-row average manner.

[0034] Compared with the prior art, the present application includes at least one of the following beneficial technical effects:

[0035] Selecting a suitable light source and adjusting the intensity ensures that the detector imaging is not overexposed, and accurate and clear light signal data is obtained. Uniformly illuminating the optical fiber entrance ensures that the brightness of different optical fibers is averaged, avoiding the situation that some optical fibers are not illuminated, and providing stable and reliable original data for subsequent processing.

[0036] The AMPD algorithm is used to distinguish different optical fibers, and has no hyperparameters, no need for parameter adjustment, good adaptability, strong noise resistance, low requirement for periodicity, wide applicability, and accurate identification of optical fiber boundaries.

[0037] It is suitable for different optical fiber transmission conditions, can present different line types according to actual conditions, and accurately describes various transmission conditions. Through parameter adjustment, customized fitting can be achieved, improving the adaptability and fitting accuracy for different conditions.

[0038] The new baseline selection criteria can automatically adjust the spatial dimension merging range according to different transmission conditions, improve signal quality and accuracy, and reduce errors. The spatial dimension merging area of each optical fiber can be obtained directly from 0-level data, improving data processing efficiency, focusing on wavelength information, reducing noise, and enhancing signal quality.

[0039] This invention ensures accurate data acquisition by rationally selecting light sources and adjusting their intensity; it employs the AMPD algorithm to efficiently distinguish optical fibers with strong noise resistance; the super-Gaussian fitting flexibly adapts to different transmission conditions; the new baseline selection standard can automatically adjust the merging range, improving signal quality and accuracy; it can also directly determine the merging region from level 0 data, improving processing efficiency, reducing noise, and providing reliable support for spectral analysis. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of a signal combining method for an optical imaging spectrometer using optical fiber.

[0041] Figure 2 A flowchart illustrating a method for combining signals from an optical imaging spectrometer using optical fibers;

[0042] Figure 3 Image of sunlight from an ultraviolet fiber optic spectrometer;

[0043] Figure 4 This is a solar image from a spectrometer;

[0044] Figure 5 The results of distinguishing single optical fibers are shown in the figure.

[0045] Figure 6 Peak finding results for AMPD;

[0046] Figure 7 For different k values, the supergaussian line type is shown.

[0047] Figure 8 To generate a super-Gaussian fitting plot for a single fiber;

[0048] Figure 9 This is the full width at half maximum (FWHM) plot after super-Gaussian fitting;

[0049] Figure 10 This is a graph showing the trend of baseline and half-width at half-maximum as a function of k value, as proposed in this embodiment.

[0050] Figure 11 The result of baseline fitting proposed in this embodiment Figure 1 ;

[0051] Figure 12 The results of distinguishing single optical fibers are shown in the figure.

[0052] Figure 13 The result of baseline fitting proposed in this embodiment Figure 2 . Detailed Implementation

[0053] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0054] Example

[0055] like Figure 2 As shown in the figure, this embodiment proposes a method for combining signals from an optical imaging spectrometer using optical fibers, based on... Figure 3 The process, combined with practical application cases, is explained in detail to illustrate the processing procedures and details in the technical solution.

[0056] The fiber optic inlet is typically connected to a specially designed optical lens, which is uniformly illuminated by a light source. The choice of light source depends on the wavelength parameters of the spectrometer. Taking a fiber optic spectrometer in the ultraviolet band as an example, the light source can be sunlight, a mercury lamp, or a xenon lamp. The intensity of the light source should ensure that the spectrometer's detector imaging does not result in overexposure, which can be adjusted by adjusting the detector's integration time. Uniformly illuminating the fiber optic inlet ensures even brightness across different fibers and prevents some fibers from being unilluminated.

[0057] After the light source illuminates the fiber optic inlet, the light passes through the fiber, slit, and beam splitter, forming a two-dimensional light spot on the detector.

[0058] like Figure 3 This image shows the image taken by a two-dimensional CCD detector of an ultraviolet fiber optic spectrometer using the sun as a light source. The spectrometer has 49 fibers. In the image, the horizontal axis represents the spectral dimensions. Figure 4 The blue curve in the image represents the Fraunhofer line from 300nm to 400nm; the vertical axis represents the spatial dimension, indicating that adjacent fibers are separated in space due to the cladding. For illustrative purposes, Figure 4 This is an image of the detector's spatial dimensions being uniformly illuminated when there is no optical fiber.

[0059] like Figure 3 As shown, the CCD is uniformly illuminated in the spatial dimension (vertical axis). Based on the CCD's level 0 data, a set of data is selected from it. Figure 5 The data (in one column) is illuminated in the spatial dimension. Periodic peak finding is performed on this data set to distinguish different optical fibers. This embodiment uses the AMPD algorithm, which is applied to periodic or quasi-periodic signals. The algorithm itself has (almost) no hyperparameters, requires no parameter tuning, and has good adaptability to signals. The only assumption is that the signal is periodic or quasi-periodic.

[0060] It has strong noise resistance, and as we will see later, it is not very demanding in terms of periodicity.

[0061] Figure 3 To illustrate the results of distinguishing individual optical fibers using the AMPD algorithm, adjacent fibers are plotted in different colors. The spatial dimension of spectral data is a column selected from the level 0 data, such as... Figure 6 The red curve in the image.

[0062] It should be noted here that when the 0-level data is applied to the AMPD algorithm, the light intensity value needs to be taken as negative, and then the AMPD algorithm is applied. The reason is that the AMPD is to find the peak value, and the cladding between the adjacent optical fibers causes the light intensity to decrease, which is a trough. After the light intensity value is taken as negative, the AMPD algorithm can find the boundary between the two optical fibers.

[0063] Figure 5 The middle blue curve is the result of taking the negative value of the spatial dimension data, and the red marker points are the peak values calculated by the AMPD algorithm. According to the position of the peak value, different optical fibers can be distinguished, and then the result of Figure 7 is obtained.

[0064] After distinguishing different optical fibers, the spatial dimension data of a single optical fiber can be obtained, and the hypergaussian fitting is performed.

[0065] The principle and method of hypergaussian fitting are briefly described here.

[0066] The expression of hypergaussian is:

[0067]

[0068] The parameters ω and k determine the shape and width of the hypergaussian. When k = 2, the line type is a standard Gaussian. Figure 8 is the hypergaussian line type corresponding to different k values. When k < 2, the line type gradually sharpens in the middle, and when k > 2, the line type gradually flattens in the middle.

[0069] Figure 9 The results of hypergaussian fitting of all single optical fiber spatial dimension data are shown in the figure. The fitting results show that the hypergaussian line type better describes the effect of optical fiber transmission under different conditions.

[0070] Next, the baseline needs to be selected as the reference for merging the spatial dimension data. Generally, in the application of spectral resolution, the half-width of Gaussian and hypergaussian fitting is often selected as the reference of resolution. The half-width expression of hypergaussian is:

[0071]

[0072] As a comparison, the hypergaussian half-width is selected as the baseline to obtain the result as Figure 9 .

[0073] Figure 10 The blue broken line in the middle is the half-width after hypergaussian fitting. If the area above the blue broken line is selected as the part of spatial merging, the weak part of the optical fiber transmission edge will also be included. When the optical fiber transmission state is good, this inclusion is unnecessary, and it will lead to the decrease of the overall intensity after merging.

[0074] To this end, the present embodiment proposes a new baseline selection:

[0075]

[0076] The baseline expression is similar to the expression of the half-height width, but the value is also affected by the linear parameter k, Figure 10 The results of the above formula and the half-height width with the change of k value are compared, and ω = 1.

[0077] As Figure 11 shown, it is obvious from the figure that when the value of k is greater than 2, the baseline proposed by the present method is smaller than the half-height width, and the selected spatial dimension is smaller than the spatial dimension selected by the half-height width. When the value of k is less than 2, the opposite is true.

[0078] Figure 12 The blue broken line in the figure shows the results of using the new baseline. For most optical fibers, the flatter part of the line is selected.

[0079] As Figure 13 and ​ show a set of fitting results under the condition that the transmission state of the optical fiber is poor (k value is less than 2). After fitting the baseline, the baseline can be intersected with the fitting curve and solved, and the part above the baseline can be used as the selection of the combined spatial dimension. The results of the part are shown in Table 1:

[0080] Table 1 Corresponding spatial dimensions of some optical fibers in the CCD

[0081]

[0082]

[0083] After obtaining the spatial dimensions corresponding to the optical fibers, the merging method is to average multiple rows.

[0084] The present application selects appropriate light sources and adjusts the intensity to ensure that the detector imaging does not appear overexposure, which can ensure that accurate and clear optical signal data is obtained. By uniformly illuminating the optical fiber entrance, the brightness between different optical fibers is averaged, avoiding the situation that part of the optical fiber is not illuminated, thereby providing a more stable and reliable original data basis for subsequent signal processing.

[0085] Among them, the AMPD algorithm is used to distinguish different optical fibers with multiple advantages. First of all, this algorithm has almost no super parameter and does not need to be adjusted, and has good self-adaptability to the signal, which greatly reduces the complexity and technical threshold of the operation. Secondly, it has strong anti-noise ability, and even in the presence of certain noise interference, it can accurately identify the boundaries between different optical fibers, improving the accuracy of optical fiber differentiation. The characteristics of not being too high in periodicity also make this algorithm have more extensive applicability in various practical application scenarios.

[0086] In addition, the super-Gaussian fitting method can adapt to different fiber transmission conditions. When the transmission condition is good, the approximate rectangular pulse line type; when the transmission condition is poor, the approximate triangular pulse line type. This flexibility enables the method to more accurately describe various situations of actual fiber transmission, whether in an ideal state or in a certain problem, and can effectively fit the optical signal to provide a more accurate model for subsequent processing.

[0087] By adjusting the shape and width of the super-Gaussian through parameters, customized fitting can be performed according to the specific fiber transmission characteristics, further improving the adaptability and fitting accuracy of different situations.

[0088] It is worth noting that the new baseline selection criteria It has important significance. It can automatically adjust the range of spatial dimension merging according to the changes of line type parameters, fully considering the signal characteristics under different fiber transmission conditions. When the fiber transmission condition is good, the spatial dimension of the baseline selection is narrower, so that the part of the optical signal that is strong and uniform is more concentrated, improving the quality and accuracy of the signal; when the fiber transmission condition is poor, the baseline is wider, and more spatial dimension data is selected to reduce the error, ensuring that useful signals can be obtained as much as possible under adverse conditions.

[0089] Further, the intersection of the baseline and the fitting curve is solved to obtain the merged spatial dimension selection, and then the multi-row average merging method is performed, which can effectively compress the spatial information related to the fiber position, focus on the wavelength information of the spectrum, reduce the random noise of the measurement, and enhance the signal quality. At the same time, this method can directly obtain the spatial dimension merging area used by each fiber from the 0-level data of different types of fiber spectrometers, improving the efficiency and automation level of data processing.

[0090] The above specific embodiments are only a few optional embodiments of the present application, and based on the technical solutions of the present application and the related inspiration of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. A method for combining signals from an optical imaging spectrometer using optical fibers, characterized in that, The following steps are included: Light source illumination and data acquisition: The light source is used to uniformly illuminate the fiber optic inlet, and the fiber optic spectrometer detector generates level 0 data. Spatial dimension data selection and peak identification: Select a column of illuminated spatial dimension data from level 0 data, and use the AMPD algorithm to identify the peak and distinguish the spatial dimension corresponding to different optical fibers. Super-Gaussian fitting and baseline selection: Super-Gaussian fitting is performed on single-fiber data, based on an optimized baseline. The area above the baseline is selected as the spatial merging region, where k is the line type parameter; Spatial dimension merging combines the spatial dimensions of optical fibers into a single row by averaging multiple rows.

2. The method for combining signals from an optical imaging spectrometer using optical fiber according to claim 1, characterized in that, The illumination and data acquisition processes specifically include: The fiber optic inlet is uniformly illuminated using a light source. The choice of light source is related to the band parameters of the spectrometer. For fiber optic spectrometers in the ultraviolet band, any of the following can be used: sunlight, mercury lamp, or xenon lamp. The intensity of the light source is adjusted by adjusting the detector integration time to ensure that the spectrometer detector imaging does not result in overexposure. After the light source illuminates the fiber optic inlet, the light passes through the fiber, slit, and beam splitter, forming a two-dimensional light spot on the detector and generating level 0 data.

3. The method for combining signals from an optical imaging spectrometer using optical fiber according to claim 1, characterized in that, In the selection of spatial dimension data and peak identification, when applying the AMPD algorithm to level 0 data, the light intensity value is negatively evaluated before calculation. The cladding between adjacent optical fibers causes a decrease in light intensity, which is a trough. After taking the negative value, the boundary between the two optical fibers is found.

4. The method for combining signals from an optical imaging spectrometer using optical fiber according to claim 1, characterized in that, In the super-Gaussian fitting and baseline selection process, after distinguishing different optical fibers, the spatial dimension data of a single optical fiber is obtained, and super-Gaussian fitting is then performed on it. The expression for the super-Gaussian is: The parameters ω and k determine the shape and width of the supergaussian. When k=2, the line type is a standard Gaussian; When k < 2, the line shape gradually becomes sharper in the middle; When k>2, the line gradually flattens out in the middle; A baseline is selected as the reference for merging spatial dimension data; baseline selection criteria. The baseline expression is similar to the half-width at half-height expression, but its value is affected by the linetype parameter k. When the value of k is greater than 2, the baseline is smaller than the half-width at half-height, and the selected spatial dimension is smaller than the selected spatial dimension at half-width; when the value of k is less than 2, the opposite is true.

5. The method for combining signals from an optical imaging spectrometer using optical fiber according to claim 1, characterized in that, Spatial dimension merging specifically involves, after fitting the baseline, intersecting the baseline with the fitted curve and solving for the portion above the baseline, which is then used as the selection of the spatial dimension to be merged. Finally, the spatial dimensions corresponding to the optical fibers are merged by averaging multiple rows.

Images