Pollen detection methods

Abstract

This invention discloses a pollen detection method. The detection method includes: initiating detection, automatically verifying the detection link status, and generating a detection command after the detection link is normal; collecting aerosol particles in the atmosphere according to the detection command; performing optical detection on the collected aerosol particles to obtain multi-dimensional feature information of the aerosol particles; analyzing the multi-dimensional feature information based on a preset multi-source data fusion algorithm, outputting pollen type and concentration data in the aerosol particles, and generating information on other aerosols besides pollen. The technical solution of this application can effectively improve detection reliability, increase sample representativeness, increase detection dimensions, and improve detection accuracy and automation level.

Description

Technical Field

[0001] This invention relates to the field of environmental testing technology, specifically to a pollen detection method. Background Technology

[0002] With the acceleration of global climate change and urbanization, urban green areas are continuously expanding, and the composition of particles in the atmosphere is becoming increasingly complex. Pollen, as a major airborne allergen, has become a core trigger for respiratory diseases such as allergic rhinitis and asthma, seriously affecting public health. At the same time, other particles such as bacteria, fungal spores, and viruses also exist in the atmosphere, and their spread and diffusion are directly related to public health safety and ecological environment quality. Therefore, the need for accurate detection of pollen and other aerosols is becoming increasingly urgent.

[0003] Current pollen detection methods largely rely on manual sampling combined with laboratory analysis. First, there is a lack of automatic verification mechanisms for the detection process after it is initiated, making it prone to data distortion due to abnormalities in the process, thus affecting the reliability of the results. Second, the sampling process lacks standardized adaptability, making it difficult to meet the particle collection needs under different environments, resulting in insufficient sample representativeness. Third, the detection methods mostly focus on a single pollen target, only obtaining limited morphological or concentration information, and cannot simultaneously acquire relevant characteristics of other aerosols. In short, current solutions tend to have low accuracy in pollen species identification, large errors in concentration calculation, and low automation of the detection process, failing to meet the needs of real-time monitoring. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a pollen detection method that can effectively improve detection reliability, increase sample representativeness, add detection dimensions, and improve detection accuracy and automation.

[0005] This application provides a pollen detection method, the detection method comprising:

[0006] Initiate detection, automatically verify the status of the detection link, and generate a detection command after the detection link is normal.

[0007] Collect aerosol particles in the atmosphere according to the detection instructions;

[0008] Optical detection is performed on the collected aerosol particles to obtain multi-dimensional feature information of the aerosol particles;

[0009] The multi-dimensional feature information is analyzed based on a preset multi-source data fusion algorithm, and the pollen type and concentration data in the aerosol particles are output, and other aerosol information other than pollen is generated.

[0010] In one aspect, the steps for optically detecting the collected aerosol particles include:

[0011] Holographic imaging, polarized light, and fluorescence were used to detect the collected aerosol particles, respectively.

[0012] In one aspect, the steps for detecting acquired aerosol particles using holographic imaging include:

[0013] An object beam and a reference beam are emitted toward the aerosol particles, and the object beam irradiates the aerosol particles to generate scattered light.

[0014] The scattered light is interfered with the reference light to form a hologram;

[0015] The hologram is digitally processed to obtain the three-dimensional morphology, surface texture, and size characteristics of the aerosol particles.

[0016] In one aspect, the step of detecting collected aerosol particles using polarized light includes:

[0017] Output linearly polarized light and adjust the polarization state of the linearly polarized light;

[0018] The polarized light is irradiated onto the aerosol particles, and polarized scattering signals at different scattering angles are collected.

[0019] The scattered signal is demodulated and the polarization degree, depolarization ratio, and scattering angle-polarization degree distribution curve are extracted.

[0020] In one aspect, the steps for detecting collected aerosol particles using fluorescence include:

[0021] Aerosol particles are irradiated with a laser of a preset wavelength to excite characteristic biomolecules within the particles to produce specific fluorescence, and the fluorescence signal is obtained by separating the specific fluorescence characteristics.

[0022] The fluorescence spectrum of the fluorescence signal was analyzed to extract the fluorescence peak wavelength, fluorescence intensity, and fluorescence lifetime.

[0023] In one aspect, the step of collecting aerosol particles in the atmosphere according to the detection command includes:

[0024] The collected atmospheric samples were pre-processed to remove impurities and excess water vapor;

[0025] Adjust the sampling flow rate according to the preset detection requirements, select the sampling mode, and transport the atmospheric sample to the detection area. The sampling mode includes continuous sampling and intermittent sampling.

[0026] In one aspect, the step of analyzing the multi-dimensional feature information based on a preset multi-source data fusion algorithm and outputting pollen species and concentration data in the aerosol particles includes:

[0027] Integrate multi-dimensional feature information obtained by optical detection, and construct a multi-dimensional particle feature spectrum based on the multi-dimensional feature information;

[0028] The multidimensional particle feature spectrum is compared with pre-stored pollen feature data to identify pollen species.

[0029] Based on the particle counting algorithm and sampling parameters, the corresponding pollen concentration data is calculated.

[0030] In one aspect, the detection method further includes:

[0031] The detection link status and environmental parameters are monitored in real time, and the sampling flow rate and detection parameters are adjusted according to the detection link status and environmental parameters.

[0032] In one aspect, the steps for generating aerosol information other than pollen include:

[0033] Based on the aforementioned optical detection, other aerosol morphological characterization data, biological attribute data, and spatial structure data are captured;

[0034] The morphological characterization data reflects the appearance and surface features of aerosols, the biological attribute data reflects the biological activity state and biological category attributes of aerosols, and the spatial structure data reflects the three-dimensional structural features of aerosols, which can be used for public health prevention and control, ecological environment assessment and scientific research analysis.

[0035] In one aspect, following the step of generating aerosol information other than pollen, the following are included:

[0036] The pollen species and concentration data, as well as the multi-dimensional feature information of other aerosols, are encrypted.

[0037] Storing and transmitting encrypted data;

[0038] Based on the transmitted data, the detection results are displayed in real time;

[0039] Based on the stored data, historical data can be backtracked to query detection data for a specified time period;

[0040] Based on the stored and transmitted data and related influencing factors, trend prediction is carried out to anticipate the changing trends of pollen and other aerosols.

[0041] Based on pollen concentration and the biosafety characteristics of other aerosols, targeted protection recommendations are generated.

[0042] The beneficial effects of this invention are as follows: By automatically verifying the detection link status during the detection startup phase and generating detection commands when the link is normal, the distortion of detection data caused by link anomalies is reduced from the source, improving the reliability of detection results; standardized collection of aerosol particles in the atmosphere according to detection commands ensures the standardization and representativeness of sample collection, unifying standards; optical detection of aerosol particles to obtain multi-dimensional feature information, combined with analysis using a preset multi-source data fusion algorithm, increases the detection dimensions, improving the accuracy of pollen species identification and concentration calculation, and enabling the generation of other aerosol information besides pollen, meeting the needs of diverse application scenarios such as public health prevention and control, and ecological environment assessment. Simultaneously, the entire process requires no manual intervention, improving the automation level of detection, effectively increasing detection reliability, sample representativeness, and detection dimensions, thereby improving detection accuracy and automation. Attached Figure Description

[0043] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0044] Figure 1 This is a schematic diagram of the process steps for the pollen detection method of this application. Detailed Implementation

[0045] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. These embodiments are merely illustrative of the technical solution of the present invention and are therefore intended to limit the scope of protection of the present invention.

[0046] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application should have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0047] like Figure 1 As shown, this application provides a pollen detection method, the detection method including:

[0048] Step S10: Initiate the detection process and automatically verify the status of the detection links. Once the detection links are confirmed to be normal, a detection command is generated. After the detection process begins, the status of each link related to the detection is automatically verified. These links include signal transmission links, device collaboration links, and data interaction links. A built-in self-test program checks the connection stability, functional effectiveness, and parameter matching of each link one by one. After confirming that all detection links are in normal working condition and there are no abnormalities affecting the detection results, standardized detection commands are automatically generated. These commands clearly define the core execution basis for subsequent sampling, detection, and analysis. This approach can proactively identify potential link failure risks and reduce detection deviations caused by link anomalies.

[0049] Step S20: Collect aerosol particles from the atmosphere according to the detection command. Upon receiving the detection command, the aerosol particle collection process will be initiated according to the preset requirements of the command. The collection process will capture various types of aerosol particles in the atmospheric environment, including pollen, bacteria, fungal spores, viruses, and other non-biological particles. The collection process will follow standardized operating procedures to ensure the acquisition of representative atmospheric samples and reduce the distortion of subsequent detection results due to sampling deviations.

[0050] Step S30 involves performing optical detection on the collected aerosol particles to obtain multi-dimensional feature information. This multi-dimensional feature information includes morphological features, structural features, and biological attribute features of the particles, such as their appearance, surface texture, three-dimensional structure, polarization characteristics, and fluorescence properties. Optical detection can comprehensively uncover the essential characteristics of aerosol particles from different perspectives, and the obtained multi-dimensional feature information provides rich and valuable basic data for subsequent data analysis, helping to identify the category and attributes of aerosol particles.

[0051] Step S40: Based on a preset multi-source data fusion algorithm, multi-dimensional feature information is analyzed to output pollen type and concentration data in aerosol particles, and information on other aerosols besides pollen is generated. The preset multi-source data fusion algorithm comprehensively analyzes and deeply mines the multi-dimensional feature information of aerosol particles acquired by optical detection. The multi-source data fusion algorithm can integrate feature data from different dimensions, fully leverage the complementary advantages of various data types, filter out pollen-related feature information, accurately identify the specific pollen type, and calculate the corresponding pollen concentration data. Simultaneously, during the analysis process, relevant feature information of other aerosol particles besides pollen can also be extracted, forming a comprehensive detection result.

[0052] Multi-source data fusion algorithms include feature-layer weighted fusion algorithms, decision-layer voting fusion algorithms, or Bayesian inference fusion algorithms.

[0053] In this embodiment, the detection link status is automatically verified during the detection startup phase. Detection instructions are generated when the link is normal, reducing the data distortion caused by link anomalies from the source and improving the reliability of the detection results. Aerosol particles in the atmosphere are collected in a standardized manner according to the detection instructions, ensuring the standardization and representativeness of sample collection and unifying standards. Optical detection of aerosol particles is used to obtain multi-dimensional feature information, which is then analyzed in combination with a preset multi-source data fusion algorithm to improve the detection dimensions. This not only improves the accuracy of pollen species identification and concentration calculation, but also enables the generation of information on other aerosols besides pollen, meeting the needs of diverse application scenarios such as public health prevention and control and ecological environment assessment. At the same time, the entire process does not require manual intervention, improving the automation level of detection, effectively improving detection reliability, sample representativeness, increasing detection dimensions, and improving detection accuracy and automation.

[0054] In one embodiment of this application, the step of performing optical detection on the collected aerosol particles includes:

[0055] In step S300, holographic imaging, polarized light, and fluorescence are used to detect the collected aerosol particles. These three techniques—holographic imaging, polarized light, and fluorescence—are employed to detect the collected aerosol particles, uncovering their core features from different dimensions. The three detection techniques complement and synergistically work together to comprehensively capture the multi-dimensional features of aerosol particles, thus improving the overall comprehensiveness of the detection results.

[0056] In one embodiment of this application, the step of detecting collected aerosol particles using holographic imaging includes:

[0057] In step S310, an object beam and a reference beam are emitted towards the aerosol particles. The object beam illuminates the aerosol particles and generates scattered light. In the holographic image detection stage, the object beam and reference beam are simultaneously emitted towards the acquired aerosol particles. Both the object beam and the reference beam can be lasers. The object beam acts directly on the surface of the aerosol particles, and after interacting with the particles, it generates scattered light. The scattered light carries basic information such as the spatial position and morphological contour of the aerosol particles. The reference beam maintains its original propagation path and does not come into contact with the aerosol particles, providing a reference for subsequent interference.

[0058] In step S311, the scattered light is interfered with the reference light to form a hologram. The scattered light carrying aerosol particle information is superimposed with the reference light, which serves as a reference, and interference occurs between the two. Due to the difference in propagation characteristics between the scattered light and the reference light, a hologram containing three-dimensional spatial information of the aerosol particles is formed during the interference process. This information covers key details such as the particle's depth and contour. The formation of the hologram enables effective recording of aerosol particle information, and compared with traditional two-dimensional imaging methods, it can more comprehensively preserve the spatial features of the particles.

[0059] Step S312 involves digitally processing the hologram to obtain the three-dimensional morphology, surface texture, and size characteristics of the aerosol particles. A digital processing algorithm is used to analyze and process the hologram. By analyzing and restoring the interference fringes in the hologram, the three-dimensional spatial information of the aerosol particles contained in the hologram is transformed into intuitive and quantifiable feature data. Ultimately, the three-dimensional morphology, surface texture, and size characteristics of the aerosol particles are obtained. The three-dimensional morphology clearly shows the three-dimensional structure of the particles, the surface texture reflects the fine lines and unevenness of the particle surface, and the size characteristics provide an accurate basis for quantifying the particle size. These feature data help improve the reliability of the overall detection and analysis.

[0060] The digital processing of holograms mainly involves capturing holograms using high-resolution CCD / CMOS image sensors, converting the optical pattern into a digital image in the form of a pixel array, and preserving the grayscale distribution and spatial position information of the interference fringes. Noise reduction algorithms (such as median filtering and Gaussian filtering) are used to remove image noise, and contrast enhancement and histogram equalization techniques are used to strengthen the difference between the interference fringes and the background, thereby improving image quality. Fourier transform and phase-shifting interferometry are employed to analyze the spatial frequency and phase distribution of the interference fringes, restoring the optical path difference information of the aerosol particles. Based on the phase data, combined with diffraction propagation algorithms (such as angular spectral propagation), the spatial distribution of the object light is inferred through numerical calculations, reconstructing a three-dimensional model of the aerosol particles. The reconstructed three-dimensional model is then digitally analyzed, using edge detection and morphological processing to extract particle size parameters (particle size, volume), surface texture parameters (roughness, texture density), and morphological parameters (sphericity, contour complexity), which are then converted into digital data.

[0061] In one embodiment of this application, the step of detecting collected aerosol particles using polarized light includes:

[0062] Step S320: Output linearly polarized light and adjust its polarization state. In the polarized light detection stage, linearly polarized light is first output as the detection light source. Linearly polarized light has a fixed vibration direction. The polarization state of this linearly polarized light is flexibly adjusted according to the detection requirements. By changing parameters such as polarization direction and polarization type, the polarized light can be adapted to the detection needs of aerosol particles with different properties. This adjustable polarized light output design can more effectively stimulate the polarization scattering characteristics of aerosol particles, helping to distinguish different types of particles.

[0063] Step S321: Polarized light is irradiated onto aerosol particles, and polarized scattering signals at different scattering angles are collected. Due to differences in the internal structure and refractive index distribution of different aerosol particles, the interaction between polarized light and the particles will generate specific polarized scattering signals at different scattering angles. These polarized scattering signals at different angles are collected using a multi-channel acquisition device to ensure that no key signal information is missed.

[0064] Step S322 involves demodulating the scattered signal and extracting the degree of polarization, depolarization ratio, and scattering angle-degree of polarization distribution curve. The collected polarized scattering signal is demodulated to remove noise interference and redundant information, restoring the signal's true characteristics. Based on this, key feature parameters such as the degree of polarization, depolarization ratio, and scattering angle-degree of polarization distribution curve are extracted from the demodulated signal. The degree of polarization reflects the particle's ability to retain polarized light, the depolarization ratio reflects the anisotropy of the particle's internal structure, and the scattering angle-degree of polarization distribution curve visually presents the changing patterns of polarization characteristics under different scattering angles. These feature parameters accurately reflect the differences in the microstructure and material properties of aerosol particles, helping to improve the accuracy of detection and analysis.

[0065] In one embodiment of this application, the step of detecting collected aerosol particles using fluorescence includes:

[0066] Step S330 involves irradiating aerosol particles with a laser of a preset wavelength to excite specific fluorescence from characteristic biomolecules within the particles. The specific fluorescence is then separated to obtain the fluorescence signal. The selected preset wavelength laser precisely matches the excitation requirements of characteristic biomolecules within the particles, such as nucleic acids, proteins, and flavonoids—common biomolecules found in bioaerosols. When the laser interacts with these characteristic biomolecules, it excites them to produce specific fluorescence. Different types of biomolecules or aerosol particles with different properties produce fluorescence with variations in wavelength and intensity. An optical separation device is used to perform characteristic separation of the generated specific fluorescence, eliminating interference factors such as ambient stray light and non-target molecule fluorescence, thereby obtaining a high-purity fluorescence signal that reflects the characteristics of the target particles.

[0067] Step S331 involves analyzing the fluorescence spectrum of the fluorescence signal and extracting the fluorescence peak wavelength, fluorescence intensity, and fluorescence lifetime. The separated fluorescence signal is analyzed using fluorescence spectra. Spectral analysis techniques are used to extract core information from the fluorescence signal, transforming it into an intuitive fluorescence spectrum curve. Based on the fluorescence spectrum curve, three key characteristic parameters—fluorescence peak wavelength, fluorescence intensity, and fluorescence lifetime—are further extracted. The fluorescence peak wavelength reflects the inherent properties of the characteristic biomolecules after excitation, and different biomolecules have significantly differentiable fluorescence peak wavelengths. Fluorescence intensity is related to the content and activity of characteristic biomolecules within the particle, indirectly reflecting the biological activity state of the particle. Fluorescence lifetime reflects the duration of the fluorescent molecule in the excited state, influenced by factors such as molecular structure and environment, and can help distinguish different types of aerosol particles. These characteristic parameters provide important support for the identification of aerosol particles from a biological attribute perspective, helping to accurately distinguish pollen from other bioaerosols and non-bioaerosols, and providing data reference for assessing the biosafety of aerosol particles.

[0068] In one embodiment of this application, the step of collecting aerosol particles in the atmosphere according to a detection command includes:

[0069] Step S210 involves preprocessing the collected atmospheric sample to remove impurities and excess moisture. Impurities include non-target substances such as excessively large dust particles and suspended particulate matter that may interfere with the detection results. Excess moisture refers to gaseous water exceeding the suitable humidity range for detection. The preprocessing optimizes the purity of the atmospheric sample, reduces the interference of impurities and moisture on subsequent optical detection, and allows aerosol particles to more clearly exhibit their characteristics.

[0070] Step S220: Adjust the sampling flow rate according to preset detection requirements, select a sampling mode, and deliver the atmospheric sample to the detection area. The sampling modes include continuous sampling and intermittent sampling. Select the appropriate sampling mode from the two options. Continuous sampling enables uninterrupted capture of atmospheric samples, suitable for scenarios requiring continuous monitoring of aerosol particle changes; intermittent sampling allows sample collection at preset time intervals, suitable for detection needs during specific time periods or low-power operation scenarios. After completing the flow rate adjustment and mode selection, the pre-processed atmospheric sample is delivered to the designated detection area.

[0071] For example, the sample is delivered to the designated detection area via a dedicated gas delivery system. This system uses low-adsorption, interference-resistant tubing materials to prevent aerosol particles in the sample from being adsorbed by the tubing walls or undergoing property changes. It is also equipped with a flow stabilization control module to ensure that the sample delivery rate is precisely matched to the detection requirements, preventing uneven particle distribution due to flow rate fluctuations. During delivery, the sample smoothly enters the core reaction chamber of the detection area along a preset path. The reaction chamber provides a sealed and stable detection environment for optical detection (holographic imaging, polarized light, fluorescence detection), ensuring that aerosol particles are evenly distributed within the detection area and remain there for a sufficient time, allowing various optical detection technologies to fully capture particle characteristics.

[0072] In one embodiment of this application, the step of analyzing multi-dimensional feature information based on a preset multi-source data fusion algorithm and outputting pollen species and concentration data in aerosol particles includes:

[0073] Step S410 integrates the multi-dimensional feature information acquired through optical detection and constructs a multi-dimensional particle feature spectrum based on this information. The multi-dimensional feature information acquired during the optical detection process is systematically integrated. This information encompasses the three-dimensional morphology, surface texture, and size features obtained from holographic imaging; parameters such as polarization degree and depolarization ratio obtained from polarized light detection; and biological attribute features such as fluorescence peak wavelength and fluorescence intensity extracted from fluorescence detection. Through standardized data integration methods, feature information of different dimensions and types is categorized, summarized, and correlated. A comprehensive multi-dimensional particle feature spectrum reflecting the properties of aerosol particles is constructed based on unified data standards. This feature spectrum centrally presents information about the physical structure, microscopic properties, and biological characteristics of aerosol particles.

[0074] For example, following the logic of data standardization, classification association, and structured modeling, heterogeneous data from holographic images, polarized light, and fluorescence detection are transformed into a unified feature set. The specific process is as follows: First, various feature data are standardized by converting physical features such as three-dimensional morphology and size to a unified dimension, and normalizing and calibrating parameters such as polarization degree and fluorescence intensity to eliminate dimensional differences and error interference from different detection technologies. Then, an association index is established for each individual particle, assigning a unique identifier to each aerosol particle and binding its corresponding physical morphological features, optical properties, and biological characteristics to ensure the correspondence of multi-dimensional information for a single particle. Finally, a structured feature matrix is ​​constructed, with particles as rows and feature types as columns, forming a multi-dimensional feature matrix containing all standardized feature parameters for each particle, while also labeling the source of the features, forming a clear and well-organized data structure.

[0075] Step S420 involves comparing the multi-dimensional particle feature spectrum with pre-stored pollen feature data to identify pollen species. The pre-stored pollen feature data covers standard feature information of various common pollens in terms of physical morphology, optical properties, and biological characteristics, forming a pollen feature database. During the comparison process, key feature parameters in the feature spectrum and standard data are matched one by one to analyze the degree of fit and differences between the two, and a comprehensive judgment is made based on feature weight allocation rules. Based on the comparison results, pollen types that highly match the target particle features are selected, thus completing the pollen species identification.

[0076] The feature weighting allocation rule combines the essential differences between pollen and other aerosols to assign differentiated importance weights to multi-dimensional feature parameters, and then achieves the determination through weighted calculation. The specific process is as follows: First, the weight allocation basis is determined based on the unique attributes of pollen. Pollen differs significantly from other aerosols in three-dimensional morphology and specific fluorescence characteristics. These features contribute more to pollen identification and are assigned higher weights (e.g., 0.3-0.4). Features with strong generality, such as polarization parameters, are assigned lower weights (e.g., 0.1-0.25), with a total weight of 1. The multi-dimensional feature parameters of a single particle are compared with the standard parameters of a pre-stored pollen feature library to calculate the similarity score of each feature. Then, the weighted total score is calculated by multiplying the feature similarity score by the corresponding weight. If the weighted total score is higher than the threshold (e.g., 0.75), the particle is determined to be the target pollen, and the most suitable pollen species is matched; if the total score is lower than the threshold, it is determined to be other aerosols, and its feature data is stored separately.

[0077] Step S430: Based on the particle counting algorithm and sampling parameters, the corresponding pollen concentration data is calculated. The particle counting algorithm is used to count pollen grains within the detection area, statistically analyzing the number of grains per unit volume that match the characteristics of the target pollen species. Simultaneously, key sampling parameters from the sampling process are considered, including sampling flow rate, sampling time, and sampling volume. By establishing a correlation calculation model between the counting results and the sampling parameters, the particle count data and sampling parameters are quantified and calculated. Taking into account factors such as sample representativeness and detection efficiency during the sampling process, the corresponding pollen concentration data is calculated. This calculation method integrates actual particle count information and sampling operation parameters, reflecting the actual distribution density of pollen in the atmospheric environment.

[0078] In one embodiment of this application, the detection method further includes:

[0079] Step S50 involves real-time monitoring of the detection link status and environmental parameters, adjusting the sampling flow rate and detection parameters based on these parameters. Throughout the entire detection process, the detection link status and environmental parameters are continuously monitored in real time. The detection link status includes key operational indicators such as signal transmission stability and the collaborative working status of each detection module. Environmental parameters cover external conditions that may affect the detection results, such as temperature, humidity, and air pressure in the detection area. By capturing this data in real time, dynamic changes during the detection process can be promptly identified. When fluctuations in the detection link status or deviations in environmental parameters from the appropriate range are detected, the sampling flow rate and related detection parameters are automatically and flexibly adjusted according to preset adjustment rules. This dynamic adjustment mechanism ensures that the sampling and detection process always adapts to the current operating status and environmental conditions, reducing the adverse effects of external factors and equipment operational fluctuations on the detection results, ensuring the stability and continuity of the detection process, and further improving the accuracy of the detection data.

[0080] The preset adjustment rules are a dynamic adaptation system based on the correlation between the detection link status, environmental parameters, sampling flow rate, and detection parameters. This system ensures that the sampling and detection processes always match the current operating conditions, guaranteeing detection stability and data accuracy. The adjustment rules are formulated with a problem-and-response focus. For example, when signal transmission fluctuates in the detection link, the sampling flow rate is reduced, while the signal integration time of optical detection is increased to reduce insufficient feature capture caused by particles flowing too quickly through the detection area. When humidity increases in the detection area, the excitation light intensity of fluorescence detection is appropriately increased to compensate for the attenuation effect of water vapor on the light signal. When there are abnormal fluctuations in air pressure, the sampling pump power and the polarization state adjustment step size of polarized light detection are adjusted simultaneously to ensure that particle sampling efficiency matches the sensitivity of optical detection. General adjustment rules are triggered by quantized thresholds (e.g., initiating light intensity adjustment when humidity exceeds 60% or flow rate adjustment when signal transmission stability is below 95%), and can be executed automatically without manual intervention. This achieves dynamic optimization of sampling and detection parameters, reducing the adverse effects of external environmental and equipment operation fluctuations on the detection results.

[0081] In one embodiment of this application, the step of generating aerosol information other than pollen includes:

[0082] Step S401 involves capturing morphological characterization data, biological attribute data, and spatial structure data of other aerosols based on optical detection. Morphological characterization data reflects the appearance and surface features of aerosols; biological attribute data reflects the biological activity state and biological category attributes of aerosols; and spatial structure data reflects the three-dimensional structural features of aerosols, for use in public health control, ecological environment assessment, and scientific research analysis. During the analysis of multi-dimensional feature information based on a preset multi-source data fusion algorithm, key feature data of other aerosols besides pollen are simultaneously captured based on the data obtained in the optical detection step. Specifically, this includes morphological characterization data, biological attribute data, and spatial structure data. Morphological characterization data can intuitively reflect the appearance and surface features of other aerosols, such as the geometric contour of particles, surface smoothness, or texture distribution. Biological attribute data can effectively reflect the biological activity state of other aerosols and their biological category attributes, such as whether they are active bacteria, fungal spores, or abiotic particles. Spatial structure data clearly presents the three-dimensional structural features of other aerosols, including the three-dimensional configuration of particles and the distribution of internal pores. These multidimensional data can provide relevant references for potential pathogenic microorganisms in public health prevention and control, supplement basic data on atmospheric aerosol community structure for ecological environment assessment, and also provide rich sample characteristic evidence for scientific research analysis, expanding the application scenarios of detection results.

[0083] In one embodiment of this application, after the step of generating aerosol information other than pollen, the following steps are included:

[0084] Step S60 involves encrypting pollen species and concentration data, as well as multi-dimensional characteristic information of other aerosols. This encryption process is applied to the pollen species and concentration data acquired during the detection process, along with the multi-dimensional characteristic information of other aerosols. By employing encryption algorithms, the data is encoded and protected to prevent leakage, tampering, or unauthorized access during subsequent storage and transmission. Encryption ensures the security and integrity of the detection data, guaranteeing that sensitive information and core detection results are only accessible to authorized entities. To balance encryption efficiency and transmission security, a hybrid encryption method combining asymmetric and symmetric encryption is typically used. For example, AES / SM4 (symmetric algorithm) is used to quickly encrypt the original pollen detection data and other aerosol characteristic information; RSA / SM2 (asymmetric algorithm) is used to encrypt the key for the symmetric algorithm, rather than directly encrypting the data. The ciphertext data and the encrypted key are transmitted / stored together. The authorized party first decrypts the data using their private key to obtain the symmetric key, and then uses this key to decrypt the original detection data.

[0085] Step S61: Store and transmit the encrypted data. After data encryption, all encrypted detection data is categorized and stored using a stable and reliable storage method to ensure long-term data retention and easy retrieval. Simultaneously, the encrypted data is accurately transmitted to the designated terminal device or data management platform according to a preset transmission protocol and path. The storage and transmission process adheres to data security standards, ensuring the stability and confidentiality of data during its flow, and achieving secure retention and sharing of detection data.

[0086] Step S62: Based on the transmitted data, display the detection results in real time. Based on the encrypted data transmitted to the terminal device or platform, present the detection results in real time through visualization. Use intuitive and easy-to-understand display formats, such as data charts, numerical lists, and status indicators, to present the specific types and concentrations of pollen, as well as multi-dimensional characteristic information of other aerosols. Real-time display allows users to quickly grasp the current status of pollen and other aerosols in the atmospheric environment, meeting the needs of real-time monitoring and immediate decision-making.

[0087] Step S63: Based on the stored data, historical data backtracking is performed to query detection data for a specified time period. Users can set a specified time range or period according to their actual needs, and retrieve historical detection records such as pollen species and concentration data, and other aerosol characteristic information from the stored database for that period. Historical data backtracking helps users trace the atmospheric environmental conditions at different times, analyze the changing patterns of pollen and other aerosols, and provide complete historical data support for related research and evaluation work.

[0088] Step S64 involves trend prediction based on stored and transmitted data and relevant influencing factors to forecast the changing trends of pollen and other aerosols. This includes integrating stored and transmitted detection data with relevant factors affecting the distribution and changes of pollen and other aerosols, such as seasonal changes, meteorological conditions, environmental management measures, and vegetation growth, for comprehensive analysis. By mining the correlation patterns between data, the changes in pollen types and concentrations, as well as the distribution and characteristic changes of other aerosols, can be predicted over a future period.

[0089] Step S65: Generate targeted protection recommendations based on pollen concentration and the biosafety characteristics of other aerosols. Based on the detected pollen concentration data, combined with the biosafety characteristics of other aerosols, such as the presence of active pathogens and their allergenicity, comprehensively assess the potential impact of the atmospheric environment on human health. For example, generate targeted protection recommendations for individuals with allergies, those with weakened immune systems, and the general public, including suggestions on choosing the right time to go out, wearing protective equipment, and environmental disinfection measures.

[0090] Furthermore, in this application, the multi-source data fusion algorithm adds a dynamic weight adaptive mechanism based on feature weight allocation. The feature weights are automatically adjusted according to environmental parameters and particle type during the detection process. In high humidity environments, the weight of fluorescence features is automatically increased, while the weight of polarization features is correspondingly decreased. When pollen from closely related species is detected, the weight of three-dimensional morphological texture features is increased, ensuring the relevance of core identification features under different working conditions.

[0091] A multi-model collaborative error correction module was added, integrating the morphological matching model, fluorescence spectroscopy verification model, and polarization attribute filtering model, and establishing a confidence level cross-validation rule. When the confidence level of a single model is lower than the set standard, the other two models are triggered for secondary verification. Through the collaborative logic of selecting two out of the three, false judgments are eliminated, improving the accuracy of pollen species identification.

[0092] A feature redundancy removal process has been added, employing principal component analysis and mutual information entropy calculation to automatically select the core features that contribute most to pollen identification and remove irrelevant feature parameters. This reduces the amount of data processing while improving recognition efficiency, adapting to the computing power requirements of portable detection devices.

[0093] A humidity-intensity mapping table is established for fluorescence detection to acquire humidity data of the detection area in real time. When the humidity exceeds a preset threshold, the excitation light power is automatically adjusted according to the humidity value, and a water vapor scattering correction algorithm is activated to counteract the attenuation effect of water vapor on the fluorescence signal and reduce detection errors in high humidity environments.

[0094] The sample pretreatment process employs a combination of two-stage filtration and particle size prediction. The first-stage filtration removes large particles, while the second-stage filtration uses a dynamic electric field to adsorb fine dust. Combined with holographic image pre-scanning, non-pollen particles are pre-identified and marked, and their feature weights are automatically reduced during subsequent detection to avoid interference from impurities in pollen identification.

[0095] A new temperature and pressure linkage calibration feature has been added, incorporating data from temperature and pressure sensors and establishing calibration formulas for characteristic parameters. When temperature changes reach a set range or pressure fluctuations occur, characteristic parameters such as polarization degree and fluorescence lifetime are automatically corrected to ensure the stability of test results under different environmental conditions.

[0096] By combining biological attribute data obtained from fluorescence detection, an allergenicity feature library was constructed. By extracting the fluorescence characteristics corresponding to allergenic proteins in pollen, pollen was classified into three levels: highly allergenic, moderately allergenic, and lowly allergenic. Simultaneously, this was correlated with an allergy incidence probability model, making the detection results more valuable for health guidance.

[0097] The trend prediction function has been optimized by integrating historical monitoring data with real-time meteorological data and employing a time-series prediction algorithm. By learning the variation patterns of pollen concentration with meteorological conditions, the pollen concentration change curve for the next 24 to 72 hours can be predicted, providing forward-looking data support for public health prevention and control.

[0098] To address the challenge of identifying pollen morphologies from closely related species, a specific fluorescent labeling and microtexture enhancement technique was employed. By exciting specific components of the pollen cell wall with a laser of a preset wavelength to generate a unique fluorescent label, and simultaneously utilizing holographic super-resolution processing, the subtle differences in pollen surface textures were amplified, enabling precise differentiation of pollen from closely related species.

[0099] Employing a miniaturized optical detection unit design, the holographic imaging polarization light and fluorescence detection modules are integrated into a compact housing. Combined with a low-power CCD sensor and a compact laser light source, along with a lightweight algorithm design, this achieves portability of the detection equipment, meeting the needs of real-time on-site detection.

[0100] An automatic air path cleaning function has been added. During the testing interval, the pipeline is purged by reverse airflow, in conjunction with the ultraviolet sterilization module. This avoids cross-contamination caused by pollen residue, extends the equipment's lifespan, and ensures long-term testing accuracy.

[0101] A wireless ad-hoc network data sharing module has been added, enabling multiple detection devices to connect via the wireless ad-hoc network. Distributed monitoring of pollen concentration within the region is conducted, and data fusion is used to generate a regional pollen distribution heat map, providing more comprehensive data support for ecological environment assessment.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for detecting pollen, characterized in that, The detection method includes: Initiate detection, automatically verify the status of the detection link, and generate a detection command after the detection link is normal. Collect aerosol particles in the atmosphere according to the detection instructions; Optical detection is performed on the collected aerosol particles to obtain multi-dimensional feature information of the aerosol particles; The multi-dimensional feature information is analyzed based on a preset multi-source data fusion algorithm, and the pollen type and concentration data in the aerosol particles are output, and other aerosol information other than pollen is generated.

2. The detection method according to claim 1, characterized in that, The steps for optical detection of the collected aerosol particles include: Holographic imaging, polarized light, and fluorescence were used to detect the collected aerosol particles, respectively.

3. The detection method according to claim 2, characterized in that, The steps for detecting collected aerosol particles using holographic imaging include: An object beam and a reference beam are emitted toward the aerosol particles, and the object beam irradiates the aerosol particles to generate scattered light. The scattered light is interfered with the reference light to form a hologram; The hologram is digitally processed to obtain the three-dimensional morphology, surface texture, and size characteristics of the aerosol particles.

4. The detection method according to claim 2, characterized in that, The steps for detecting collected aerosol particles using polarized light include: Output linearly polarized light and adjust the polarization state of the linearly polarized light; The polarized light is irradiated onto the aerosol particles, and polarized scattering signals at different scattering angles are collected. The scattered signal is demodulated and the polarization degree, depolarization ratio, and scattering angle-polarization degree distribution curve are extracted.

5. The detection method according to claim 2, characterized in that, The steps for detecting collected aerosol particles using fluorescence include: Aerosol particles are irradiated with a laser of a preset wavelength to excite characteristic biomolecules within the particles to produce specific fluorescence, and the fluorescence signal is obtained by separating the specific fluorescence characteristics. The fluorescence spectrum of the fluorescence signal was analyzed to extract the fluorescence peak wavelength, fluorescence intensity, and fluorescence lifetime.

6. The detection method according to any one of claims 1 to 5, characterized in that, The steps for collecting aerosol particles in the atmosphere according to the detection command include: The collected atmospheric samples were pre-processed to remove impurities and excess water vapor; Adjust the sampling flow rate according to the preset detection requirements, select the sampling mode, and transport the atmospheric sample to the detection area. The sampling mode includes continuous sampling and intermittent sampling.

7. The detection method according to any one of claims 1 to 5, characterized in that, The steps of analyzing the multi-dimensional feature information based on a preset multi-source data fusion algorithm and outputting pollen species and concentration data in the aerosol particles include: Integrate multi-dimensional feature information obtained by optical detection, and construct a multi-dimensional particle feature spectrum based on the multi-dimensional feature information; The multidimensional particle feature spectrum is compared with pre-stored pollen feature data to identify pollen species. Based on the particle counting algorithm and sampling parameters, the corresponding pollen concentration data is calculated.

8. The detection method according to any one of claims 1 to 5, characterized in that, The detection method further includes: The detection link status and environmental parameters are monitored in real time, and the sampling flow rate and detection parameters are adjusted according to the detection link status and environmental parameters.

9. The detection method according to claim 1, characterized in that, The steps for generating aerosol information other than pollen include: Based on the aforementioned optical detection, other aerosol morphological characterization data, biological attribute data, and spatial structure data are captured; The morphological characterization data reflects the appearance and surface features of aerosols, the biological attribute data reflects the biological activity state and biological category attributes of aerosols, and the spatial structure data reflects the three-dimensional structural features of aerosols, which can be used for public health prevention and control, ecological environment assessment and scientific research analysis.

10. The detection method according to claim 1, characterized in that, Following the steps of generating aerosol information other than pollen, the following steps are included: The pollen species and concentration data, as well as the multi-dimensional feature information of other aerosols, are encrypted. Storing and transmitting encrypted data; Based on the transmitted data, the detection results are displayed in real time; Based on the stored data, historical data can be backtracked to query detection data for a specified time period; Based on the stored and transmitted data and related influencing factors, trend prediction is carried out to anticipate the changing trends of pollen and other aerosols. Based on pollen concentration and the biosafety characteristics of other aerosols, targeted protection recommendations are generated.

Images