Optical imaging spectrometer signal merging method applying optical fiber

Abstract

The invention relates to the technical field of optical imaging spectrometer signal merging, in particular to an optical imaging spectrometer signal merging method applying optical fibers. According to the technical scheme, the method comprises the following steps that light source illumination and data acquisition are conducted, a light source is used for evenly illuminating an optical fiber inlet, and a fiber spectrometer detector generates 0-level data; and spatial dimension data selection and peak value identification: selecting a column of illuminated spatial dimension data from the 0-level data, performing peak value identification by adopting an AMPD algorithm, and distinguishing spatial dimensions corresponding to different optical fibers. The light source is reasonably selected and the intensity is adjusted, so that the data acquisition accuracy is ensured; the AMPD algorithm is adopted to efficiently distinguish optical fibers, and the noise immunity is high; the super Gaussian fitting can flexibly adapt to different transmission conditions; the new baseline selection standard can automatically adjust the merging range, improve the signal quality and accuracy, directly determine the merging region from the 0-level data, improve the processing efficiency, reduce the noise, and provide reliable support for spectral analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of optical imaging spectrometer signal merging, and in particular to a method for merging optical imaging spectrometer signals using optical fibers. Background Art

[0002] An optical imaging spectrometer using optical fiber (hereinafter referred to as fiber spectrometer) is a spectral analysis instrument that uses optical fiber as a light transmission medium to transmit an external light source into the spectrometer for analyzing the intensity distribution of light of different wavelengths. The core diameter of the optical fiber outlet matches the slit width of the spectrometer, and through precise optical axis alignment and divergence angle control, it is ensured that the optical fiber output light can effectively enter the slit. When there are multiple optical fibers, since the cladding thickness of the optical fiber plays a physical isolation role in the optical fiber array, the light entering the slit of the spectrometer through the optical fiber is not a continuous beam, but is composed of multiple segmented beams. In an ideal state, the CCD image plane of the spectrometer presents a segmented, uniform, and clearly identifiable light spot in the spatial dimension.

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

[0004] Fiber optic spectrometer is an optical analysis instrument based on fiber optic transmission technology. It is used to guide optical signals into the spectrometer through optical fibers to analyze and measure information such as the wavelength and intensity of the light. Fiber optic spectrometer is currently the most common spectrometer. The flexibility of optical fiber allows it to be easily integrated into various measurement scenarios to adapt to different experimental environments and application requirements; compared with traditional spectrometers, fiber optic spectrometers are usually 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, medicine and biology, astronomy and remote sensing, chemical analysis and experimental research and other fields.

[0005] The fiber optic spectrometer consists of a fiber optic transmission system, a slit, a spectroscopic element, a detector, and a data processing system. The fiber optic transmission system is usually composed of a special optical lens, an optical fiber, and a fiber optic interface. The optical fiber is responsible for collecting and transmitting optical signals. The optical fiber usually includes a core and a cladding. The core is used for light transmission, and the cladding is used for isolation and reflection. The input end of the fiber optic interface collects light from the light source, and the output end transmits the light to the slit of the spectrometer. When the light beam output by the optical fiber enters the spectrometer, it is spatially filtered through the slit to limit the beam width to improve the spectral resolution. The spectroscopic element is usually a grating, which decomposes the incoming white light into light of different wavelengths to form a spectrum. The detector is usually a CCD or CMOS sensor, which is responsible for detecting light of different wavelengths after dispersion and converting it into electrical signals. The data processing system converts the electrical signal output by the detector into a digital signal to process and analyze the spectral data.

[0006] After the optical signal transmitted by the optical fiber passes through the slit of the spectrometer, a two-dimensional spot image will be 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 transmitted in the optical fiber array (that is, the optical signals transmitted at different positions of the optical fiber array).

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

[0008] The output end of the optical fiber transmits light to the spectroscopic system of the spectrometer through a slit. The slit plays a role in limiting the light flux and improving the spectral resolution in this process. The matching between the output end of the optical fiber and the slit is divided into: 1. Matching of the core diameter of the optical fiber and the width of the slit: The core diameter of the optical fiber determines the size of the light beam transmitted by the optical fiber, 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 optical fiber is usually divergent and needs to be collimated by a collimating lens to form a parallel light beam passing through the slit. After the collimated light beam passes through the slit, it enters the spectroscopic system. The degree of collimation will affect the transmission efficiency and imaging quality of the light.

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

[0010] The slit, beam splitter and detector in the spectrometer jointly 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 factors of the spectrometer itself, the coupling between the optical fiber and the slit may be poor due to the following factors, resulting in low efficiency of optical signal transmission and unsatisfactory light spots formed on the CCD image plane. 1. The fiber core diameter does not match the slit width: If the fiber core diameter is too small and the slit is too wide, the light beam entering the spectrometer will diverge or be insufficient, resulting in uneven shape and intensity distribution of the light spot on the CCD; conversely, if the fiber core diameter is too large, part of the light beam cannot enter the slit, resulting in signal loss. This causes the light spot on the CCD image plane to be blurred, incomplete or weakened. 2. Alignment error between the fiber outlet and the slit: If the fiber outlet is not accurately aligned with the slit, the light beam will deviate from the predetermined optical path, resulting in part of the light failing to pass through the slit into the spectrometer. This causes the light spot on the CCD to deviate or partially disappear, resulting in an incomplete beam. 3. Uneven arrangement of the fiber array: If multiple optical fibers are arranged unevenly or incorrectly, the light beam may be partially blocked or deviated when entering the slit. This improper arrangement will result in an irregular light spot array on the CCD, and even cause differences in the intensity distribution between the light spots. This causes the light spot array to be irregular or the spacing to be uneven, with some spots being stronger and some spots being weaker. 4. The fiber divergence angle does not match the slit acceptance angle: If the divergence angle is too large, the light beam entering the spectrometer slit will diverge excessively, 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. This causes the intensity of the light spot on the CCD to weaken, and the light spot to diffuse or be too concentrated. 5. Quality issues with the fiber outlet end face: If the fiber outlet end face is scratched, contaminated, or damaged, scattering and diffraction will occur during light transmission, resulting in an irregular light beam and affecting the coupling efficiency. This causes the shape of the light spot on the CCD image plane to be irregular, the edge of the light spot to be blurred or noisy, and the overall intensity of the light spot array to be reduced. 6. Micro-bending or stretching of the optical fiber: If the optical fiber is subjected to mechanical stress (such as excessive bending or stretching) during transmission, it may cause the optical path inside the optical fiber to change, resulting in transmission loss, changes in the beam divergence angle, and ultimately affecting coupling. The intensity of the light spot on the CCD is reduced or uneven, the shape of the light spot becomes irregular, and some optical fibers even do not output light. 7. Problems with the optical fiber fixing and alignment device: If the fixing device between the optical fiber and the slit is loose or the alignment mechanism is unstable, it may cause a slight displacement or vibration of the optical fiber relative to the slit, thereby affecting the coupling effect. The light spot position is unstable, and the shape and intensity of the light spot change over time. 8. Optical fiber cladding reflection problem: If the cladding material and structure of the optical fiber are not ideal, or the optical fiber is not in proper contact with the external environment (such as damage or contamination of the cladding), the reflection efficiency in the optical fiber may be reduced, resulting in loss during the light beam transmission process. The intensity of the light spot on the CCD is generally low, and some optical fibers may even not output light at all.

[0012] like Figure 1The figure shows the curve of a single light beam in the spatial dimension generated by a certain ultraviolet fiber spectrometer after the fiber entrance is evenly illuminated by sunlight and transmitted to the two-dimensional CCD image plane. (a) The output of the fiber is relatively smooth and uniform, and there is no obvious fluctuation or unevenness in the light intensity distribution. Such an output means that the coupling of the fiber is good, the light is evenly distributed in the spatial dimension, and there is no obvious light loss or interference. (b) The light intensity fluctuates with the change of spatial position, but it still has good coherence overall. (c) The fluctuation becomes larger, and the light intensity at different positions shows obvious differences, indicating that there is more significant poor coupling in the fiber or that the surface defects of the fiber cause different transmission efficiencies of the light beam in different areas. (d)(e)(f) As the coupling effect deteriorates, the spatial distribution of the light intensity gradually presents a Gaussian distribution. The light intensity distribution has a central peak, but there is attenuation on both sides, indicating that the coupling of the fiber is gradually concentrated in a certain area, which may be due to the inconsistency between the fiber core and the cladding or the incomplete alignment of the geometric center of the fiber. The light intensity is concentrated in a narrow area, while the light intensity on both sides decays rapidly, which may reflect that the fiber has more serious non-uniformity during transmission, or even partial light loss. (g) Spatial distribution of light intensity of three consecutive different optical fibers. It can be clearly seen that the light intensity of different optical fibers varies greatly.

[0013] The light in the optical fiber enters the slit and forms a two-dimensional image on the detector after passing through the spectrometer. For most spectral analyses, wavelength information is the key research object, while the spatial distribution of the optical fiber does not carry information useful for spectral analysis. Therefore, by merging the CCD image in the spatial dimension, the spatial information related to the position of the optical fiber can be compressed, thereby focusing on the wavelength information of the spectrum. 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 strength. The noise is usually randomly distributed. After merging, the proportion of noise relative to the signal decreases, reducing the random noise of the measurement, thereby enhancing the signal quality.

[0014] For the spectrometer, the one-dimensional spatial scale of the slit corresponds to the spatial dimension in the two-dimensional detector image. When there is no optical fiber, the spatial dimension is continuous and is often used for merging. When optical 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 based on the dark lines 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 optical fiber at a single wavelength, and select the portion above the half-height width as the area for merging the spatial dimension;

[0015] In summary, the present application proposes a method for combining signals of an optical imaging spectrometer using optical fiber. Summary of the invention

[0016] The purpose of the present invention is to address the problems in the optical imaging spectrometer using optical fiber in the background technology. Due to the problems of optical fiber manufacturing and assembly process, there are various factors in the coupling between optical fiber and slit, which leads to poor coupling effect, resulting in non-uniform light spot on the CCD image plane, affecting the light signal transmission efficiency and data quality. The existing spatial dimension merging method has shortcomings and it is difficult to effectively deal with the problems of non-uniform light spot and noise. A method for merging signals of an optical imaging spectrometer using optical fiber is proposed.

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

[0018] Light source illumination and data collection: Use the light source to evenly illuminate the fiber entrance, and the fiber spectrometer detector generates level 0 data;

[0019] Spatial dimension data selection and peak recognition: select a column of spatial dimension data from the 0-level data, use the AMPD algorithm for peak recognition, and distinguish the spatial dimensions corresponding to different optical fibers;

[0020] Super Gaussian fitting and baseline selection: Super Gaussian fitting is performed on single fiber data, and the optimized baseline is selected. Select the part above the baseline as the area for spatial merging;

[0021] Merge spatial dimensions: merge the spatial dimensions of optical fibers into one row, and the merging method is to average multiple rows.

[0022] Optionally, the light source illumination and data collection specifically include:

[0023] Use a light source to evenly illuminate the fiber entrance. The choice of light source is related to the band parameters of the spectrometer. The fiber optic spectrometer in the ultraviolet band can be any one of sunlight, mercury lamp, and xenon lamp. The light source intensity is adjusted by adjusting the detector integration time to ensure that the spectrometer detector imaging is not overexposed.

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

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

[0026] Optionally, in the super-Gaussian fitting and baseline selection, after different optical fibers are distinguished, the data of a single optical fiber in the spatial dimension is obtained, and a super-Gaussian fitting is performed on it.

[0027] The expression of super Gauss is

[0028] The parameters k and k determine the shape and width of the super-Gaussian.

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

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

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

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

[0033] Optionally, the spatial dimension merging specifically includes, after fitting the baseline, intersecting the baseline with the fitting curve and solving it, obtaining the portion above the baseline as the choice of the merged spatial dimension, and finally merging the spatial dimensions corresponding to the optical fiber by averaging multiple rows.

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

[0035] Select a suitable light source and adjust the intensity to ensure that the detector imaging is not overexposed and obtain accurate and clear optical signal data. Evenly illuminate the fiber entrance to ensure that the brightness of different fibers is average and avoid that some fibers are not illuminated, providing stable and reliable raw data for subsequent processing.

[0036] The AMPD algorithm is used to distinguish different optical fibers. It has no hyperparameters and does not require parameter adjustment. It has good adaptability, strong noise resistance, low requirements on periodicity, wide applicability, and can accurately identify optical fiber boundaries.

[0037] Adapting to different optical fiber transmission conditions, it can present different line types according to the actual situation and accurately describe various transmission conditions. Customizable fitting can be achieved through parameter adjustment to improve adaptability and fitting accuracy to different conditions.

[0038] The new baseline selection standard 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 directly obtained from the level 0 data, which improves data processing efficiency, focuses on wavelength information, reduces noise, and enhances signal quality.

[0039] The present invention ensures accurate data collection by reasonably selecting the light source and adjusting the intensity; adopts the AMPD algorithm to efficiently distinguish optical fibers with strong noise resistance; super-Gaussian fitting can flexibly adapt to different transmission conditions; the new baseline selection standard can automatically adjust the merging range to improve signal quality and accuracy, and can also directly determine the merging area from level 0 data, improve processing efficiency, reduce noise, and provide reliable support for spectral analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of a method for combining signals of an optical imaging spectrometer using optical fiber;

[0041] Figure 2 A flow chart of a method for combining signals of an optical imaging spectrometer using optical fiber;

[0042] Figure 3 This is the sunlight imaging image of the UV fiber optic spectrometer;

[0043] Figure 4 This is the sunlight imaging image of the spectrometer;

[0044] Figure 5 This is a diagram showing the results of distinguishing a single optical fiber;

[0045] Figure 6 This is the peak finding result of AMPD;

[0046] Figure 7 is the super-Gaussian line shape corresponding to different k values;

[0047] Figure 8 This is a super-Gaussian fitting diagram for a single optical fiber;

[0048] Fig. 9 is the half-height width diagram after super-Gaussian fitting;

[0049] Fig.10 This is a trend diagram of the baseline and half-height width as a function of the k value proposed in this embodiment;

[0050] Fig.11 The results of the baseline fitting proposed in this example are Figure 1 ;

[0051] Fig.12 This is a diagram showing the results of distinguishing a single optical fiber;

[0052] Fig.13 The results of the baseline fitting proposed in this example are Figure 2 . DETAILED DESCRIPTION

[0053] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0054] Example

[0055] like Figure 2 As shown, the present embodiment proposes a method for combining signals of an optical imaging spectrometer using optical fibers, according to Figure 2 The process and details of the technical solution are explained in detail through the actual application cases.

[0056] The fiber entrance is usually connected to a special optical lens, which is evenly illuminated by a light source. The choice of light source is related to the band parameters of the spectrometer. Taking the fiber spectrometer in the ultraviolet band as an example, the light source can be sunlight, mercury lamp, or xenon lamp. The intensity of the light source should ensure that the detector imaging of the spectrometer is not overexposed, which can be adjusted by adjusting the detector integration time. The light source evenly illuminates the fiber entrance to ensure that the brightness between different optical fibers is averaged to avoid some optical fibers not being illuminated.

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

[0058] like Figure 3 , the image is the imaging of the two-dimensional CCD detector when the ultraviolet fiber spectrometer uses the sun as the light source. The fiber spectrometer has 49 optical fibers. In the image, the horizontal axis is the spectral dimension, such as Figure 3 The blue curve in the figure is the Fraunhofer line from 300nm to 400nm; the vertical axis is the spatial dimension, and adjacent optical fibers are separated in the spatial dimension due to the cladding. As an illustration, Figure 4 This is an image of the detector being uniformly illuminated in spatial dimensions when there is no optical fiber.

[0059] like Figure 4 As shown, the CCD is evenly illuminated in the spatial dimension (vertical axis direction). According to the 0-level data of the CCD, a group ( Figure 3 The data with illuminated spatial dimension (a column in the middle) is used to perform periodic peak search on the data to distinguish different optical fibers. This embodiment adopts the AMPD algorithm, which is applied to periodic or quasi-periodic signals. The algorithm itself has (almost) no hyperparameters and no parameter adjustment is required. It has good adaptability to the signal. The only assumption is that the signal is periodic or quasi-periodic.

[0060] It has strong noise immunity and, as we will see later, does not have very high requirements for periodicity.

[0061] Figure 5 The results of applying the AMPD algorithm to distinguish a single fiber are shown in Figure 1. Adjacent fibers are drawn in different colors. The spectral data in the spatial dimension is a column selected from the level 0 data, such as Figure 3 The red curve in .

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

[0063] Figure 6 The blue curve in the middle is the result of taking a negative value for the spatial dimension data, and the red mark is the peak value calculated by the AMPD algorithm. Different optical fibers can be distinguished based on the position of the peak value, and then the Figure 5 result.

[0064] After different optical fibers are distinguished, the data of a single optical fiber in the spatial dimension can be obtained and a super-Gaussian fitting can be performed on it.

[0065] The principles and methods of super-Gaussian fitting are briefly explained here.

[0066] The expression of super Gauss is:

[0067]

[0068] The parameters ω and k determine the shape and width of the super Gaussian. When k = 2, the line shape is a standard Gaussian. Figure 7 The super Gaussian line shape corresponding to different k values. When k < 2, the line shape gradually becomes sharp in the middle, and when k > 2, the line shape gradually becomes flat in the middle.

[0069] Figure 8 The results of super-Gaussian fitting are performed for all single-fiber spatial dimension data. The fitting results show that the super-Gaussian line shape better describes the effect of fiber transmission under different conditions.

[0070] Next, we need to select the baseline as a reference for merging spatial dimension data. Generally, in the application of spectral resolution, Gaussian and super Gaussian fitting often use half-width as a reference for resolution. The half-width expression of super Gaussian is:

[0071]

[0072] For comparison, the super-Gaussian half-width is selected as the baseline, and the following is obtained: Fig. 9 result.

[0073] Fig. 9 The blue line in the middle is the half-height width after super-Gaussian fitting. If the area above the blue line is selected as the part of the spatial merger, the part with weaker light intensity at the edge of the optical fiber transmission will also be included. When the optical fiber transmission state is good, this inclusion is unnecessary and will lead to a decrease in the overall intensity after the merger.

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

[0075]

[0076] The baseline expression is similar to the half-height width expression, but the value is affected by the line type parameter k. Fig.10 The above formula is compared with the result of the half-height width changing with the k value, and ω=1 is set.

[0077] like Fig.10 As shown in the figure, it can be clearly seen that when the k value is greater than 2, the baseline proposed by this 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 k value is less than 2, the opposite is true.

[0078] Fig.11 The blue line in the figure shows the result of using the new baseline, which selects the flatter part of the line shape for most fibers.

[0079] like Fig.12 and Fig.13 A set of fitting results for poor fiber transmission conditions (k value is less than 2) are shown. After fitting the baseline, the baseline and the fitting curve can be intersected and solved to obtain the part above the baseline, which can be used as the choice of the merged spatial dimension. Some of the results are shown in Table 1:

[0080] Table 1 Some optical fibers correspond to the spatial dimensions in CCD

[0081]

[0082]

[0083] After obtaining the spatial dimensions corresponding to the optical fiber, they are merged by averaging multiple rows.

[0084] This application selects a suitable light source and adjusts the intensity to ensure that the detector imaging is not overexposed. This operation can ensure that accurate and clear optical signal data is obtained. By evenly illuminating the fiber entrance, the brightness between different fibers is averaged, and the situation where some fibers are not illuminated is avoided, thereby providing a more stable and reliable raw data basis for subsequent signal processing.

[0085] Among them, the use of the AMPD algorithm for periodic peak finding to distinguish different optical fibers has many advantages. First, the algorithm has almost no hyperparameters and does not require parameter adjustment. It has good adaptability to the signal, which greatly reduces the complexity of the operation and the technical threshold. Secondly, it has strong anti-noise ability. Even in the presence of certain noise interference, it can accurately identify the boundaries between different optical fibers, improving the accuracy of optical fiber distinction. The characteristic of not requiring high periodicity also makes the algorithm more widely applicable in various practical application scenarios.

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

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

[0088] It is worth noting that the proposed new baseline selection criteria It is of great significance. It can automatically adjust the scope of spatial dimension merging according to the changes in linear parameters, and fully consider the signal characteristics under different optical fiber transmission conditions. When the optical fiber transmission condition is good, the baseline selects a narrower spatial dimension, which makes the strong and uniform part of the optical signal more concentrated, improving the quality and accuracy of the signal; when the optical fiber transmission condition is poor, the baseline is wider, and more spatial dimension data is selected to reduce errors, ensuring that useful signals can be obtained as much as possible under unfavorable conditions.

[0089] Furthermore, the baseline and the fitting curve are intersected to obtain the merged spatial dimension selection, and then the merging method of averaging multiple rows 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 optical fiber from the 0-level data of different types of fiber optic spectrometers, which improves the efficiency and automation of data processing.

[0090] The above specific embodiments are only several optional embodiments of the present invention. Based on the technical solutions of the present invention and the relevant inspirations 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 of an optical imaging spectrometer using optical fiber, characterized in that: The method comprises the following steps: Light source illumination and data collection: Use the light source to evenly illuminate the fiber entrance, and the fiber spectrometer detector generates level 0 data; Spatial dimension data selection and peak recognition: select a column of spatial dimension data from the 0-level data, use the AMPD algorithm for peak recognition, and distinguish the spatial dimensions corresponding to different optical fibers; Super Gaussian fitting and baseline selection: Super Gaussian fitting is performed on single fiber data, and the optimized baseline is selected. Select the part above the baseline as the area for spatial merging; Merge spatial dimensions: merge the spatial dimensions of optical fibers into one row, and the merging method is to average multiple rows.

2. The method for combining signals of an optical imaging spectrometer using optical fiber according to claim 1, characterized in that: The light source illumination and data collection specifically include: Use a light source to evenly illuminate the fiber entrance. The choice of light source is related to the band parameters of the spectrometer. The fiber optic spectrometer in the ultraviolet band can be any one of sunlight, mercury lamp, and xenon lamp. The light source intensity is adjusted by adjusting the detector integration time to ensure that the spectrometer detector imaging is not overexposed. After the light source illuminates the optical fiber entrance, the light passes through the optical fiber, slit, and beam splitter, forming a two-dimensional light spot on the detector, generating level 0 data.

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

4. The method for combining signals of an optical imaging spectrometer using optical fiber according to claim 1, characterized in that: In the super-Gaussian fitting and baseline selection, after different optical fibers are distinguished, the data of a single optical fiber in the spatial dimension is obtained, and a super-Gaussian fitting is performed on it. The expression of super Gauss is The parameters ω and k determine the shape and width of the super-Gaussian. When k = 2, the line type is standard Gaussian; When k < 2, the line gradually peaks in the middle; When k>2, the line gradually flattens in the middle; Select the baseline as a reference for merging spatial dimension data. 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 k value is greater than 2, the baseline is smaller than the half-width, and the selected spatial dimension is smaller than the half-width; when the k value is less than 2, the opposite is true.

5. The method for combining signals of an optical imaging spectrometer using optical fiber according to claim 1, characterized in that: The spatial dimension merging specifically includes, after fitting the baseline, intersecting the baseline with the fitting curve and solving it, obtaining the portion above the baseline as the choice of the merged spatial dimension, and finally merging the spatial dimensions corresponding to the optical fiber by averaging multiple rows.

Images