Silkworm cocoon surface pollution detection method

Abstract

The invention is applicable to the technical field of silkworm cocoon quality detection, and provides a silkworm cocoon surface pollution detection method, which is used for performing non-destructive detection on silkworm cocoon surface pollution by combining a high-sensitivity confocal Raman spectrum analysis system and a self-adaptive spectrum analysis technology based on a Raman spectrum analysis technology. According to the self-adaptive spectrum analysis technology, firstly, a probe on a scanning platform is used for curved surface reconstruction based on pre-scanning point cloud data based on an irradiation sampling surface of a sample; screening out pre-scanning points which do not contain microbial spores, and generating a three-dimensional data model which is highly matched with the distribution of microorganisms on the silkworm cocoons; corresponding to the reconstructed three-dimensional coordinates of the remaining multiple points, analyzing the corresponding Raman spectrograms one by one by utilizing Raman detection to obtain an identification result. Therefore, the method can significantly reduce the operation complexity and accelerate the overall detection rate.

Description

Technical Field

[0001] The invention is applicable to the technical field of silk cocoon quality detection and provides a method for detecting surface contamination of silk cocoons. Background Art

[0002] As an important raw material for the silk industry, the quality of silk cocoons directly affects the quality and market competitiveness of silk products. They also have high economic value in other fields such as medicine and food. However, microbial contamination is an issue that cannot be ignored during the collection and storage of silk cocoons. The growth and reproduction of microorganisms such as Aspergillus and Beauveria bassiana not only affect the quality of silk cocoons, but may also lead to the degradation of silk fibers, reduce the strength and toughness of silk, and further affect its durability and glossiness.

[0003] At present, there are several main methods for detecting microbial contamination on the surface of silk cocoons. The plate culture method is the most traditional method, which determines the degree of contamination by inoculating samples onto the culture medium and observing the growth of microorganisms. However, this method has a long detection cycle, usually 3-7 days, which cannot meet the needs of immediate processing of fresh cocoons and has low efficiency. The PCR detection method detects microorganisms by amplifying specific DNA fragments, with high sensitivity and specificity. However, this method requires destroying the cocoon shell to extract DNA, which not only causes economic losses, but may also affect the subsequent use of the cocoon, especially for those used in silk production. The loss is more significant. Traditional Raman detection uses Raman scattering spectroscopy to analyze the molecular structure of substances, but due to the strong fluorescence interference of sericin on the surface of the cocoon shell, the signal-to-noise ratio is low (less than 5:1), making it difficult to accurately identify low-concentration spores. Near-infrared imaging technology obtains image information by detecting the spectral characteristics of objects in the near-infrared band, but its spatial resolution is low, usually greater than 50μm, and cannot effectively detect the presence of microscopic spores.

[0004] Although these detection methods can detect microbial contamination on the surface of cocoons to a certain extent, they still have obvious limitations. In particular, in terms of detection speed, they generally have a long detection cycle and cannot meet the needs of real-time detection. In addition, the existing methods are still insufficient in solving problems such as signal-to-noise ratio and accurate identification of low-concentration spores, and cannot provide efficient, non-destructive and accurate detection results. Therefore, it is urgent to develop more efficient and accurate technologies to solve these problems in order to meet the needs of the silk industry and related fields for cocoon quality control. Summary of the invention

[0005] In view of the above-mentioned defects, the purpose of the present invention is to provide a method for detecting surface contamination of silkworm cocoons, the purpose of which is to solve the problems raised in the background technology. Based on Raman spectroscopy analysis technology, a highly sensitive confocal Raman spectroscopy analysis system is combined with an adaptive spectral analysis technology to perform non-destructive detection of surface contamination of silkworm cocoons; Adaptive spectral analysis technology first uses the probe on the scanning platform to reconstruct the surface based on the pre-scanned point cloud data based on the irradiated sampling surface of the sample; pre-scanned points that do not contain microbial spores are screened out to generate a three-dimensional data model that is highly consistent with the distribution of microorganisms on the cocoon; The three-dimensional coordinates of the remaining multiple points after reconstruction are corresponding to each other, and the corresponding Raman spectra are analyzed one by one by using Raman detection to obtain the identification results.

[0006] Furthermore, the surface reconstruction based on the pre-scanned point cloud data and the generation of a three-dimensional model that is highly consistent with the cocoon surface include the following steps: S1, pre-scanning point cloud data acquisition, the point cloud data includes the location information of the surface sampling points and the RGB values ​​included in the labels, and the acquired point cloud data is processed to form a point cloud data set; S2, exclude the point cloud data that meets the noise condition in the point cloud data set through conditional filtering; subsequently, filter the point cloud data based on HSV specific color conditions to obtain the local features of the point cloud data; S3. The system adaptively reconstructs and generates a three-dimensional data model that is highly consistent with the distribution of microorganisms on the cocoon based on the local features of the point cloud data.

[0007] Furthermore, the point cloud data screening method based on the HSV specific color condition includes the following steps: S2.1. Convert the RGB values ​​in the point cloud data into HSV values; S2.2. Define the H value based on the color characteristics.

[0008] S2.3. Traverse the point cloud data, filter out the point cloud data with a value greater than H and retain it, and filter out the point cloud data with a value less than H.

[0009] Furthermore, the three-dimensional data model includes three-dimensional coordinates of the remaining multiple points on the cocoon that are highly consistent with the distribution of microorganisms based on the sample scanning surface, and the label includes color information.

[0010] Furthermore, the H value adopts the minimum H value of the sample point cloud data containing only one microorganism / spore.

[0011] Furthermore, the Raman detection process includes the following steps: S4. The system analyzes the characteristic Raman peaks of the average Raman spectrum of the microorganisms according to the characteristic Raman peaks, and makes a preliminary determination of the biochemical components; S5. Identify the fingerprints of different types of microorganisms based on the position and intensity of the characteristic Raman peaks and determine the types of microorganisms; S6. Through the analysis of the intensity of the spectral characteristic peaks, the content of the corresponding types of microorganisms is quantitatively analyzed, and the density of the microorganisms is calculated based on the total detection area.

[0012] Furthermore, in step S4, a recognition model is constructed using statistical analysis and deep learning data processing methods to automatically find peaks and extract key features.

[0013] Furthermore, the adaptive spectral analysis technology of the cocoon shell is performed using a three-dimensional automatic scanning platform driven by piezoelectric ceramics with a displacement accuracy of micrometer level.

[0014] Furthermore, during the system detection process, notch filtering technology is used to reduce background noise and improve the signal-to-noise ratio; at the same time, a spatial light modulator is used to modulate the phase, amplitude or polarization state of light to achieve precise control of light.

[0015] Furthermore, the confocal Raman spectroscopy analysis system uses a confocal micro-Raman spectrometer to emit an excitation light source, and the excitation light source uses a near-infrared laser source with a wavelength of 785nm, and cooperates with a long working distance objective lens with a working distance WD=25mm and a numerical aperture NA=0.6 to detect the sample. Therefore, the beneficial effects of this method are: 1. High sensitivity and accuracy: The combination of confocal micro-Raman spectrometer and dual-path fluorescence suppression module significantly improves the signal-to-noise ratio and sensitivity of detection, and can detect microbial spores as low as 0.5μm, ensuring the accuracy of the test results.

[0016] 2. Non-destructive testing: Raman spectroscopy is a non-contact, non-destructive analysis method that does not require destructive treatment of samples. It is suitable for practical application scenarios such as cocoon station collection and drying and warehouse management.

[0017] 3. Fast and automated: Combining 3D scanning and adaptive spectral analysis technology, it can quickly generate a 3D model that is highly consistent with the surface of the cocoon, and screen out sampling points containing microbial spores based on pre-scanned point cloud data, significantly reducing computational complexity and speeding up the overall detection rate.

[0018] 4. Intelligent identification: Through characteristic Raman peak extraction and machine learning algorithms (such as support vector machines and convolutional neural networks), rapid and accurate identification of microbial species can be achieved. Combined with spectral characteristic peak intensity analysis, the density of microorganisms can be quantitatively calculated to provide pollution level judgment.

[0019] 5. Visual report: The system automatically generates a visual report to intuitively display the test results, including microbial species, density and contamination level, to facilitate quick decision-making by cocoon station managers.

[0020] 6. Wide scope of application: This method is suitable for screening of microbial contamination in cocoon station collection and drying, storage management and other links. It can effectively identify a variety of common microbial contaminants such as Aspergillus, Penicillium, Beauveria bassiana, etc., and provide comprehensive technical support for cocoon quality control. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 Schematic diagram of laser Raman spectrometer. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0023] The present invention conducts rapid detection of microbial contaminants (such as Aspergillus and Beauveria bassiana) on the surface of cocoon shells based on Raman spectroscopy analysis technology. By combining a highly sensitive confocal Raman spectroscopy analysis system with an adaptive spectral analysis technology, it can achieve non-destructive detection of 0.5-5μm pollutant spores. The method is suitable for screening for microbial contamination in cocoon station collection and drying, storage management and other links.

[0024] The following is a detailed description of the technical terms, technical principles and equipment types involved in the confocal Raman spectroscopy analysis system: Raman detection is a spectral analysis technology based on the Raman scattering effect, which was discovered by Indian physicist Raman in 1928. The Raman scattering effect refers to the inelastic collision between photons and molecules when monochromatic light (such as laser) is irradiated on a substance, resulting in a change in the frequency of the scattered light, namely the Raman shift. The Raman shift is related to the vibration and rotation energy levels of the molecules of the substance. Therefore, Raman detection can be applied to the fields of materials science, chemistry, biology, medicine, environmental monitoring, etc.

[0025] At present, Raman detection is usually used in food safety to detect adulteration, pesticide residues, microbial contamination and other detection items. Compared with ordinary detection methods, Raman detection has technical advantages such as insensitivity to water, non-destructive, no need for sample preparation, portability and real-time performance.

[0026] Since Raman spectroscopy is a non-contact, non-destructive analytical technique, it can be used for efficient, automated batch screening technology (HTS) and laboratory analysis. Typical applications include the analysis of liquids / powders on multi-well plates. The HTS Raman system integrates sample movement, focusing, and automation of data acquisition and analysis procedures, and can collect spectra for hundreds of samples in sequence.

[0027] During Raman spectroscopy, when the high-intensity incident light of the excitation laser is scattered by molecules, most of the scattered light has the same wavelength as the incident laser, and this scattering is called Rayleigh scattering. However, there is a very small part (about 1 / 109) of the scattered light whose wavelength is different from the incident light. The change in its wavelength is determined by the chemical structure of the doped substance. This part of the scattered light is called Raman scattering. The Raman scattered light is symmetrically distributed on both sides of the Rayleigh scattered light, but its intensity is much weaker than that of the Rayleigh scattered light, usually only 10^-6~10^-9 of the Rayleigh light intensity.

[0028] See attached Figure 1 The laser Raman spectrometer is mainly composed of five parts: light source, external light path system, sample cell, monochromator, signal processing and output system. The light source is a laser that emits an excitation laser, which is irradiated to several samples on the sample cell through the external light path system, and then the signal processing system analyzes and outputs the Raman spectrum.

[0029] A Raman spectrum is usually composed of a certain number of Raman peaks, each of which represents the wavelength position and intensity of the corresponding Raman scattered light. Each spectrum peak corresponds to a specific molecular bond vibration, which includes both a single chemical bond, such as CC, C=C, NO, CH, etc., and the vibration of a group composed of several chemical bonds, such as the breathing vibration of the benzene ring, the vibration of the long chain of the polymer, and the lattice vibration. In general, the Raman spectrum (including peak position and relative intensity) provides a unique chemical fingerprint of the substance, which can be used to identify the substance and distinguish it from other substances. By searching the Raman spectrum database to find matching results, unknown substances can be quickly identified. When other conditions remain unchanged, the intensity of the spectrum is proportional to the sample concentration. After determining the relationship between peak intensity and concentration (standard curve) through a sample of standard concentration, concentration analysis can be performed. For a mixture, the relative peak intensity can provide information on the relative concentrations of various components, while the absolute peak intensity can reflect the absolute concentration information.

[0030] The present invention uses a confocal micro-Raman spectrometer to emit an excitation light source, that is, the excitation laser beam is focused into a tiny spot with a diameter of 0.5μm~1.0μm through a microscope. The Raman signal generated by the irradiation of this spot returns to the spectrometer through the microscope, and then the spectral information is obtained. The equipment uses a near-infrared laser wavelength of 785nm, and cooperates with a long working distance objective lens (working distance WD=25mm, numerical aperture NA=0.6) to detect the sample.

[0031] In the detection process, a dual-path fluorescence suppression module is used, including notch filtering and spatial light modulation processes. The function of the dual-path fluorescence suppression module is to improve the detection sensitivity and accuracy. During the detection, the autofluorescence of the sample itself or other impurities may interfere with the target fluorescence signal. The dual-path fluorescence suppression module can effectively suppress these background fluorescence, making the target fluorescence signal more prominent, thereby improving the signal-to-noise ratio and accuracy of the detection.

[0032] Notch filtering is achieved by installing a notch filter in the receiving optical path, that is, installed at the Raman probe; notch filtering is an optical filtering technology used to selectively reject light of a specific wavelength while transmitting light of other wavelengths. In fluorescence detection, the notch filter can effectively block the reflection and scattering of the excitation light, thereby reducing background noise and improving the signal-to-noise ratio. Specifically, the notch filter is located at the Raman probe, based on the principle of phase cancellation interference, by introducing an anti-phase wave with the same frequency as the excitation light, so that the reflection and scattering of the excitation light cancel each other in the filter, thereby effectively removing the interference of the excitation light and improving the signal-to-noise ratio of the Raman signal.

[0033] A spatial light modulator is a device that can modulate the light field and can be used to change the phase, amplitude or polarization state of light. In a dual-path fluorescence suppression module, the spatial light modulator can be used to dynamically adjust the light path to achieve precise control of light. For example, the HDSLM80R Plus series spatial light modulator can perform pure phase modulation on the light field, with a maximum refresh rate of 60Hz, a pixel size of 8μm, a data bit depth of 8 / 10bit, a fill factor of >95%, and is suitable for wavelengths of 420~1100nm.

[0034] Based on the above detection principles and related equipment, the present invention uses the above technology to further perform adaptive spectral analysis of cocoon shells in the following manner: The adaptive spectral analysis technology of the cocoon shell uses a three-dimensional automatic scanning platform driven by piezoelectric ceramics. The piezoelectric ceramic material will undergo mechanical deformation under the action of an external electric field. By controlling the size and frequency of the applied voltage, high-precision displacement control can be achieved.

[0035] The present invention uses a piezoelectric ceramic-driven mobile platform with a displacement accuracy of micrometer level. Driven by the piezoelectric ceramic-driven three-dimensional automatic scanning platform, the probe performs surface reconstruction based on the pre-scanned point cloud data based on the irradiated sampling surface of the sample; Specifically, the surface reconstruction adaptively adjusts the parameters of the surface reconstruction according to the local features of the point cloud data to generate a 3D model that is highly consistent with the cocoon surface. It includes the following three steps: S1. Pre-scan point cloud data acquisition: Use the CloudCompare software built into the 3D scanner to obtain the point cloud data of the cocoon surface; the point cloud data is stored in the ply file format, which contains the 3D coordinates of multiple points based on the sample scanning surface, that is, the location information of the surface sampling points. At the same time, the label contains color information (RGB value), normal information (surface orientation) and other information. The acquired point cloud data are aligned and integrated to form a complete point cloud data set; S2. Point cloud data processing: Preprocess the acquired point cloud data set, including denoising, filtering, feature extraction, etc. Conditional filtering is used to exclude point cloud data that meet the noise conditions in the point cloud data set; subsequent filtering is performed by filtering point cloud data based on HSV specific color conditions, that is, point cloud data is filtered by a filtering method based on the HSV (hue, saturation, brightness) color space.

[0036] The point cloud data screening method based on HSV specific color conditions includes the following steps: S2.1. Use the RGB to HSV conversion formula to convert the RGB values ​​in the point cloud data into HSV values.

[0037] S2.2. Define the H value based on the color characteristics (use the minimum H value of the sample point cloud data with only one microorganism / spore).

[0038] S2.3. Filter the point cloud data that meets the conditions: traverse the point cloud data, filter out the point cloud data with a value greater than H and retain it, and filter out the point cloud data with a value less than H.

[0039] S3, Adaptive surface reconstruction: Adaptively reconstruct and generate a three-dimensional data model that is highly consistent with the distribution of microorganisms on the cocoon based on the local features of the point cloud data (i.e., the local point cloud data features remaining after filtering out some point cloud data based on the point cloud data with specific HSV color conditions). The three-dimensional data model contains the three-dimensional coordinates of the remaining multiple points that are highly consistent with the distribution of microorganisms on the cocoon based on the sample scanning surface, i.e., the position information of the subsequent spectral analysis, and the label contains color information, normal information, etc. Therefore, the system uses the RGB evaluation threshold to screen out pre-scan points that do not contain microbial spores, and the screened out pre-scan points do not undergo subsequent spectral analysis steps; that is, the pre-scan points screened out based on the point cloud data under HSV specific color conditions can significantly reduce the computational complexity, thereby saving computing resources and speeding up the spectral analysis rate of the overall sample.

[0040] Then, the system uses Raman detection to analyze the corresponding Raman spectra one by one according to the three-dimensional coordinates of the remaining multiple points after reconstruction. Then, the total detection area is combined with the number of microorganisms analyzed by Raman spectra to obtain the type and density information of microorganisms.

[0041] The following steps are involved: S4. Data processing and analysis of Raman spectra: The system analyzes the characteristic Raman peaks of the average Raman spectrum of microorganisms based on the characteristic Raman peaks, and makes a preliminary determination of the biochemical components.

[0042] During the process, the recognition model is constructed with the help of statistical analysis and deep learning data processing methods to achieve rapid and accurate recognition of microbial Raman spectra. Specifically, the data processing method is to use the automatic peak search function of the software, set the coordinate position event corresponding to the peak, the height and width parameters of the peak, automatically identify the characteristic Raman peak, and extract key features; use machine learning methods such as support vector machine (SVM) and convolutional neural network (CNN) to train Raman spectrum data, improve the accuracy and efficiency of recognition, and automatically identify characteristic Raman peaks.

[0043] S5. Identification of microbial species: Raman spectroscopy can reflect the differences between different types of microorganisms and provide panoramic fingerprint information of bacteria. According to the position and intensity of the characteristic Raman peak, the fingerprints of different types of microorganisms are identified, that is, the collected Raman spectral data is matched with the fingerprints in the standard spore database to determine the type of microorganism.

[0044] S6. Microbial density calculation: Through the intensity analysis of the spectral characteristic peaks, the corresponding types of microorganisms are quantitatively analyzed. That is, the intensity of the specific characteristic peak is measured, which is proportional to the content of the microorganisms. The data of multiple measurement points are statistically analyzed, and the density of the microorganisms is calculated in combination with the total detection area.

[0045] S7. Output the pollution type according to the identification results, such as Aspergillus, Penicillium, Beauveria bassiana, etc. According to the density of microorganisms, the pollution degree is divided into 1-5 levels, with 1 indicating light pollution and 5 indicating heavy pollution.

[0046] Through the above steps, it is finally possible to use Raman detection technology to analyze the microorganisms on the surface of silk cocoons point by point, obtain the type and density information of the microorganisms, and thus realize the accurate detection of microbial contamination on the surface of silk cocoons.

[0047] In the actual testing process, multiple groups of fresh cocoons are randomly selected for testing, and cocoon samples are cut from multiple locations of each group of fresh cocoons for sample preparation. After sample preparation, the hardware parameters are adjusted. Laser power: 150mW (to ensure no damage to the cocoon layer); Scanning mode: spiral path scanning (2mm diameter area, 15s / cocoon); Spectrum acquisition: integration time 200ms, 3 accumulations; The system performs Raman detection on the sample according to the above steps S1-S6; The processing flow of Raman spectra in the system is roughly as follows: Original spectrum → fluorescence background subtraction → denoising → characteristic peak extraction → spore identification → pollution level determination → visual report generation.

[0048] The spore detection limit (number of spores / mm²) was obtained during the spore identification process.

[0049] Table 1 is a comparison of the advantages and disadvantages of the present invention and the qPCR method in the detection of microbial contamination on the surface of silkworm cocoons:

[0050] Therefore, the present invention is based on Raman spectroscopy analysis technology, combined with confocal micro-Raman spectrometer, the adaptive spectral analysis technology proposed in the present invention, and the machine learning algorithm to achieve rapid and non-destructive detection of microbial contamination on the surface of silk cocoons. Through the high-precision focusing of the confocal micro-Raman spectrometer and the optimization of the fluorescence suppression module, microbial spores with a diameter of 0.5-5μm can be effectively detected; combined with three-dimensional scanning and adaptive surface reconstruction technology, accurate sampling of the complex surface of silk cocoons can be achieved; through the extraction of characteristic Raman peaks and the recognition of machine learning models, the type and density of microorganisms can be quickly determined, and finally a visual report is generated, providing a scientific basis for the sanitation control of the cocoon station collection and drying, storage management and other links.

[0051] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, technicians familiar with the field may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.

Claims

1. A method for detecting surface contamination of silkworm cocoons, characterized in that: Methods Based on Raman spectroscopy, a highly sensitive confocal Raman spectroscopy system was combined with adaptive spectral analysis technology to conduct non-destructive detection of surface contamination of silkworm cocoons. Adaptive spectral analysis technology first uses the probe on the scanning platform to reconstruct the surface based on the pre-scanned point cloud data based on the irradiated sampling surface of the sample; pre-scanned points that do not contain microbial spores are screened out to generate a three-dimensional data model that is highly consistent with the distribution of microorganisms on the cocoon; The three-dimensional coordinates of the remaining multiple points after reconstruction are corresponding to each other, and the corresponding Raman spectra are analyzed one by one by using Raman detection to obtain the identification results.

2. The method for detecting surface contamination of silkworm cocoons according to claim 1, characterized in that: The surface reconstruction based on the pre-scanned point cloud data and the generation of a three-dimensional model that is highly consistent with the cocoon surface include the following steps: S1, pre-scanning point cloud data acquisition, the point cloud data includes the location information of the surface sampling points and the RGB values ​​included in the labels, and the acquired point cloud data is processed to form a point cloud data set; S2, exclude the point cloud data that meets the noise condition in the point cloud data set through conditional filtering; subsequently, filter the point cloud data based on HSV specific color conditions to obtain the local features of the point cloud data; S3. The system adaptively reconstructs and generates a three-dimensional data model that is highly consistent with the distribution of microorganisms on the cocoon based on the local features of the point cloud data.

3. The method for detecting surface contamination of silkworm cocoons according to claim 2, characterized in that: The point cloud data screening method based on HSV specific color conditions includes the following steps: S2.

1. Convert the RGB values ​​in the point cloud data into HSV values; S2.

2. Define H value according to color characteristics; S2.

3. Traverse the point cloud data, filter out the point cloud data with a value greater than H and retain it, and filter out the point cloud data with a value less than H.

4. The method for detecting surface contamination of silkworm cocoons according to claim 1, characterized in that: The three-dimensional data model includes the three-dimensional coordinates of the remaining multiple points on the silk cocoon that are highly consistent with the distribution of microorganisms based on the sample scanning surface, and the label includes color information.

5. The method for detecting surface contamination of silkworm cocoons according to claim 3, characterized in that: The H value adopts the minimum H value of the sample point cloud data containing only one microorganism / spore.

6. The method for detecting surface contamination of silkworm cocoons according to claim 1, characterized in that: The Raman detection process comprises the following steps: S4. The system analyzes the characteristic Raman peaks of the average Raman spectrum of the microorganisms according to the characteristic Raman peaks, and makes a preliminary determination of the biochemical components; S5. Identify the fingerprints of different types of microorganisms based on the position and intensity of the characteristic Raman peaks and determine the types of microorganisms; S6. Through the analysis of the intensity of the spectral characteristic peaks, the content of the corresponding types of microorganisms is quantitatively analyzed, and the density of the microorganisms is calculated based on the total detection area.

7. The method for detecting surface contamination of silkworm cocoons according to claim 6, characterized in that: In step S4, a recognition model is constructed using statistical analysis and deep learning data processing methods to automatically find peaks and extract key features.

8. The method for detecting surface contamination of silkworm cocoons according to claim 1, characterized in that: The adaptive spectral analysis technology of the cocoon shell is carried out using a three-dimensional automatic scanning platform driven by piezoelectric ceramics with a displacement accuracy of micrometer level.

9. The method for detecting surface contamination of silkworm cocoons according to claim 1, characterized in that: During the system detection process, notch filtering technology is used to reduce background noise and improve the signal-to-noise ratio; at the same time, a spatial light modulator is used to modulate the phase, amplitude or polarization state of light to achieve precise control of light.

10. The method for detecting surface contamination of silkworm cocoons according to claim 1, characterized in that: The confocal Raman spectroscopy analysis system uses a confocal micro-Raman spectrometer to emit an excitation light source, and the excitation light source uses a near-infrared laser source with a wavelength of 785nm, and cooperates with a long working distance objective lens with a working distance WD=25mm and a numerical aperture NA=0.6 to detect the sample.

Images