Hyperspectral monitoring system platform

Abstract

The invention discloses a hyperspectral monitoring system platform, which comprises a spectral imaging module used for acquiring high-resolution spectral data of a target area within a spectral range of 400-1000nm; the data acquisition module is configured to acquire data acquired by the spectral imaging module at a rate that the transmission frame number per second is greater than or equal to 90fps; and the image processing module is used for processing and analyzing the acquired spectral data in real time, and comprises a high-precision slit design and a non-uniformity correction technology. The imaging hyperspectral technology provided by the invention can provide high-resolution space imaging, and generally can reach sub-meter level or even higher resolution, so that the system can capture fine pollutant distribution conditions, such as specific positions and diffusion paths of oil spill or chemical leakage, in an inland river environment, and the system has the advantages of being simple in structure, low in cost and the like. Therefore, the imaging hyperspectral technology has obvious innovation in real-time monitoring and accident response, and can quickly and accurately position and track the pollution source.

Description

Technical Field

[0001] The invention relates to the field of environmental monitoring, and in particular to a hyperspectral monitoring system platform. Background Art

[0002] Cross-section monitoring methods generally refer to methods of regularly collecting samples from rivers or water bodies for analysis. When dealing with pollution incidents, these methods mainly monitor water quality by installing fixed monitoring stations at specific locations in the river or collecting samples at different time points. The main limitation of this method is that they are often intermittent and discrete monitoring, which makes it difficult to capture the dynamic changes and spatial distribution of pollutants in the water body.

[0003] At present, imaging hyperspectral remote sensing technology has been gradually applied in environmental data collection due to its advantage of "map and spectrum in one". Compared with the accidental monitoring of lake and reservoir water pollution, inland river environmental pollution accidents such as oil spills and hazardous chemical leakages are extremely difficult to track, identify and analyze pollutants using conventional cross-section monitoring methods due to the narrow river channels, fast flow rates, complex terrain and strong fluidity of pollutants. Summary of the invention

[0004] The purpose of the present invention is to provide a hyperspectral monitoring system platform to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A hyperspectral monitoring system platform, comprising:

[0007] Spectral imaging module, used to obtain high-resolution spectral data of the target area within the spectral range of 400-1000nm;

[0008] A data acquisition module, configured to acquire data acquired by the spectral imaging module at a rate of ≥90 fps;

[0009] Image processing module, used to process and analyze the acquired spectral data in real time, including high-precision slit design and inhomogeneity correction technology;

[0010] An identification module configured to identify any petroleum-based contaminants in real time through an artificial intelligence algorithm with an identification accuracy rate of not less than 85%;

[0011] The evolution trend analysis module is used to quantitatively describe the evolution trend of oil film thickness and oil film type in the monitoring area based on the spatial distribution inversion model.

[0012] Furthermore, the spectral imaging module includes a high-resolution spectral imaging sensor, an optical lens system, an optical fiber bundle and a spectrometer. The optical lens system focuses light onto the spectral imaging sensor and transmits spectral data to the spectrometer for processing through the optical fiber bundle.

[0013] Furthermore, the data acquisition module includes a high-speed data acquisition card, a processor unit and a memory. The data output by the spectral imaging module is acquired in real time by the high-speed data acquisition card, then transmitted to the processor unit for real-time processing, and stored in the memory.

[0014] Furthermore, the image processing module includes an image processor, a digital signal processor and an image stitching algorithm, and the data output by the data acquisition module is processed by the image processor for image enhancement, slit design and inhomogeneity correction.

[0015] Furthermore, the recognition module includes an artificial intelligence processor, a classification algorithm and a real-time recognition library. The data processed by the image processing module is input into the recognition module, and the real-time recognition algorithm is executed by the artificial intelligence processor.

[0016] Furthermore, the evolution trend analysis module includes an evolution trend analyzer, a spatial distribution inversion model and a data visualization module, and the data collected by the data collection module is input into the evolution trend analysis module for further analysis.

[0017] Furthermore, the spectral resolution of the spectral imaging module is ≤2nm.

[0018] Furthermore, the image processing module also includes a wide-field-of-view large-format image stitching technology, which can cover a wide area and provide continuous high-resolution images.

[0019] Furthermore, the recognition module can recognize oil films on the first and second types of clean water surfaces in real time, and the recognition accuracy is not less than 90%.

[0020] Furthermore, the data acquisition module is installed on a shore-based platform with a height not exceeding 10 meters, and its spatial resolution is 0.08m.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] 1. The imaging hyperspectral technology of the present invention can provide high-resolution spatial imaging, which can usually reach a resolution of sub-meter level or even higher. This feature enables the system to capture subtle distribution of pollutants in inland river environments, such as the specific location and diffusion path of oil spills or chemical leaks. Traditional monitoring methods often cannot provide such detailed spatial information. Therefore, imaging hyperspectral technology is significantly innovative in real-time monitoring and accident response, and can quickly and accurately locate and track pollution sources.

[0023] 2. The imaging hyperspectral technology of the present invention can not only obtain spectral data in multiple bands, but also process and analyze these data efficiently. By analyzing the reflectivity and absorption characteristics of pollutants in different bands, the system can accurately identify different types of pollutants, even mixed pollution in complex backgrounds. The multi-band data analysis capability brings new perspectives and depth to inland river environment monitoring, and improves the accuracy and comprehensiveness of monitoring.

[0024] 3. The imaging hyperspectral technology of the present invention can realize real-time data collection and processing, so that the monitoring system can quickly respond to sudden pollution incidents. The system can continuously acquire spectral images, quickly analyze the diffusion path and change trend of pollutants, and provide important data support for emergency response. The real-time monitoring and emergency response capabilities are innovative in inland river environmental monitoring and help to take effective measures in a timely manner to reduce environmental damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a structural schematic diagram of a hyperspectral monitoring system platform of the present invention;

[0026] Figure 2 This is a schematic diagram of the connection of a spectral imaging module of a hyperspectral monitoring system platform of the present invention;

[0027] Figure 3 A schematic diagram of a data acquisition module connection block of a hyperspectral monitoring system platform of the present invention;

[0028] Figure 4 A schematic diagram of the connection of an image processing module of a hyperspectral monitoring system platform of the present invention;

[0029] Figure 5 This is a schematic diagram of the connection structure of the identification module of a hyperspectral monitoring system platform of the present invention;

[0030] Figure 6 This is a schematic diagram of the connection structure of an evolution trend analysis module of a hyperspectral monitoring system platform of the present invention. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] like Figure 1-6 As shown, the present invention provides a technical solution:

[0033] A hyperspectral monitoring system platform, comprising:

[0034] Spectral imaging module, used to obtain high-resolution spectral data of the target area within the spectral range of 400-1000nm;

[0035] Preferably, the spectral imaging module comprises a high-resolution spectral imaging sensor, an optical lens system, an optical fiber bundle and a spectrometer, wherein the optical lens system focuses light onto the spectral imaging sensor and transmits spectral data to the spectrometer for processing via the optical fiber bundle;

[0036] Furthermore, the spectral resolution of the spectral imaging module is ≤2nm;

[0037] In this embodiment, the spectral resolution (\delta\lambda) of the spectrometer is calculated according to the wavelength (\lambda) and relevant parameters of the spectrometer. For the hyperspectral imaging system, the spectral resolution is expressed by the following formula:

[0038] [\delta\lambda=\frac{\lambda}{N}]

[0039] Where (\lambda) is the wavelength range and (N) is the optical resolution of the spectrometer.

[0040] Specifically, the optical lens system is fixed to the shore-based platform by a mechanical bracket to ensure the correctness of the optical axis, and the focal length is adjusted by a focusing device;

[0041] In this embodiment, the optical lens system adjusts the focal length through a focusing device, and uses the following formula to calculate the object distance corresponding to the focal length change:

[0042] [d = \frac{f\cdot s}{fs}]

[0043] Where (d) is the object distance, (f) is the focal length of the lens, and (s) is the distance from the object to the lens.

[0044] A data acquisition module, configured to acquire data acquired by the spectral imaging module at a rate of ≥90 fps;

[0045] Preferably, the data acquisition module includes a high-speed data acquisition card, a processor unit and a memory, and the data output by the spectral imaging module is acquired in real time by the high-speed data acquisition card, then transmitted to the processor unit for real-time processing, and stored in the memory;

[0046] Furthermore, the data acquisition module is installed on a shore-based platform with a height not exceeding 10 meters, and its spatial resolution is 0.08m;

[0047] Specifically, the data acquisition card is installed in the control unit on the shore-based platform, and the processor unit is responsible for receiving, processing and storing the collected spectral data;

[0048] In this embodiment, the data acquisition module performs data acquisition at a rate of not less than 90fps, and the frame rate (f_{\text{fps}}) is calculated by the following formula:

[0049] [f_{\text{fps}}=\frac{1}{\Delta t}]

[0050] Where (\Delta t) is the time interval between two frames.

[0051] Image processing module, used to process and analyze the acquired spectral data in real time, including high-precision slit design and inhomogeneity correction technology;

[0052] Preferably, the image processing module includes an image processor, a digital signal processor and an image stitching algorithm, and the data output by the data acquisition module is processed by the image processor for image enhancement, slit design and inhomogeneity correction;

[0053] Furthermore, the image processing module also includes a wide-field large-format image stitching technology that can cover a wide area and provide continuous high-resolution images;

[0054] In this embodiment, the steps for implementing the wide field of view large format image stitching technology are as follows:

[0055] Use multiple viewing angles or continuously collected images to cover the entire monitoring area. Images are obtained by moving the hyperspectral monitoring system or by multiple fixed-position observation points.

[0056] During the acquisition process, ensure that the positions and angles of different images can be accurately aligned to facilitate subsequent stitching processing. Through image registration and correction, ensure that the perspective and position information of each image are consistent;

[0057] Before image stitching, it is necessary to identify and determine the overlapping areas between the images. These areas usually have overlapping image pixels for subsequent fusion processing;

[0058] By weighted averaging or feature-based fusion methods, the image pixels in the overlapping area are merged into a single pixel value to ensure the overall visual continuity and color consistency of the stitched image;

[0059] Perspective distortion caused by different viewing angles needs to be corrected to ensure that the proportions and shapes of objects in the final stitched image are accurate;

[0060] If the lens used has distortion, such as radial distortion, it needs to be corrected through correction methods to ensure the accuracy of straight lines and edges in the image;

[0061] Since processing a large number of high-resolution images requires a lot of computing and storage resources, the image processing module uses parallel computing and high-speed storage technology to ensure real-time or fast processing of large-scale data;

[0062] When designing a hyperspectral monitoring system, the real-time and efficiency of the image processing module are crucial to the overall system performance. Therefore, hardware selection and software design need to cooperate with each other to achieve efficient image stitching and processing.

[0063] An identification module configured to identify any petroleum-based contaminants in real time through an artificial intelligence algorithm with an identification accuracy rate of not less than 85%;

[0064] Preferably, the recognition module includes an artificial intelligence processor, a classification algorithm and a real-time recognition library, the data processed by the image processing module is input into the recognition module, and the artificial intelligence processor executes the real-time recognition algorithm;

[0065] Furthermore, the identification module can identify oil films on the first and second types of clean water surfaces in real time, and the identification accuracy is not less than 90%;

[0066] In this embodiment, the recognition module realizes real-time recognition of any petroleum pollutants through artificial intelligence algorithms, and the recognition accuracy is not less than 85%. The recognition accuracy (Acc) is calculated by the following formula:

[0067] [Acc=\frac{\text{TP}+\text{TN}}{\text{TP}+\text{TN}+\text{FP}+

[0068] \text{FN}}\times 100%]

[0069] Among them, (\text{TP}) is a true positive example (correctly identified positive example), (\text{TN}) is a true negative example (correctly identified negative example), (\text{FP}) is a false positive example (incorrectly identified negative example), and (\text{FN}) is a false negative example (incorrectly identified positive example).

[0070] The evolution trend analysis module is used to quantitatively describe the evolution trend of oil film thickness and oil film type in the monitoring area based on the spatial distribution inversion model.

[0071] Preferably, the evolution trend analysis module includes an evolution trend analyzer, a spatial distribution inversion model and a data visualization module, and the data collected by the data acquisition module is input into the evolution trend analysis module for further analysis.

[0072] In this embodiment, the execution steps of the hyperspectral monitoring system platform are as follows:

[0073] Step S1, start the hyperspectral monitoring system and detect the connection status of the sensor and the acquisition card.

[0074] Step S2, calibrating the optical lens system, adjusting the focal length and exposure time.

[0075] Step S3, start data acquisition, and collect spectral data at a rate of ≥90 fps.

[0076] In step S4, the data is preprocessed and corrected, including slit design and inhomogeneity correction.

[0077] Step S5: The image processing algorithm processes the data in real time, including spectral feature extraction and oil film thickness measurement.

[0078] Step S6, the real-time recognition module analyzes the processed data to identify petroleum pollutants such as diesel.

[0079] Step S7: The evolution trend analysis module uses the spatial distribution inversion model to analyze and display the evolution trend of the oil film and the distribution of pollutants.

[0080] Step S8: The results are displayed on the user interface for the operator and environmental management personnel to view and analyze.

[0081] In this embodiment, the collected oil film hyperspectral monitoring data and the spatial distribution information of the oil film position are demonstrated through the spatial distribution inversion model as follows:

[0082] First, the hyperspectral data and the location information of the oil film spatial distribution are loaded.

[0083] import numpy as np

[0084] import matplotlib.pyplot as plt

[0085] #Assume that the oil film hyperspectral data is stored in a file, and each line is a band of data

[0086] def load_hyperspectral_data(filename):

[0087] return np.loadtxt(filename)

[0088] #Assume that the oil film spatial distribution data is stored in the file, and each point has location information and pollutant concentration

[0089] def load_spill_distribution(filename):

[0090] return np.loadtxt(filename)

[0091] #Example file name

[0092] hyperspectral_file='hyperspectral_data.txt'

[0093] spill_distribution_file='spill_distribution.txt'

[0094] # Load data

[0095] hyperspectral_data=load_hyperspectral_data(hyperspectral_file)

[0096] spill_distribution_data=load_spill_distribution(spill_distribution_file)

[0097] During the demonstration, it is assumed that, using a simplified linear model for spatial distribution inversion, the pollutant concentration at each spatial point can be estimated by linearly combining the hyperspectral data.

[0098] def spatial_inversion_model(hyperspectral_data,spill_distribution_data):

[0099] num_points=spill_distribution_data.shape[0]

[0100] num_bands=hyperspectral_data.shape[1]

[0101] # Initialize the estimated pollutant concentration distribution

[0102] estimated_concentration=np.zeros(num_points)

[0103] #Simple linear model: Assume that the weights corresponding to each band are known weights = np.random.rand(num_bands) #Assume that the weights of each band are randomly initialized

[0104] for i in range(num_points):

[0105] estimated_concentration[i]=np.dot(hyperspectral_data[i],weights)

[0106] return estimated_concentration

[0107] # Perform spatial distribution inversion

[0108] estimated_concentration=spatial_inversion_model(hyperspectral_data,spill_distribution_data)

[0109] Finally, the inversion results are visualized to show the spatial distribution of oil film contaminants.

[0110] #Visualize the distribution of oil film pollutant concentration

[0111] plt.figure(figsize=(10,6))

[0112] plt.scatter(spill_distribution_data[:,0],spill_distribution_data[:,1],c=estimated_concentration,cmap='viridis')

[0113] plt.colorbar(label="Estimated Concentration")

[0114] plt.xlabel('X Coordinate')

[0115] plt.ylabel('Y Coordinate')

[0116] plt.title('Estimated Spill Concentration Distribution')

[0117] plt.grid(True)

[0118] plt.show().

[0119] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0120] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A hyperspectral monitoring system platform, characterized in that: include Spectral imaging module, used to obtain high-resolution spectral data of the target area within the spectral range of 400-1000nm; A data acquisition module, configured to acquire data acquired by the spectral imaging module at a rate of ≥90 fps; Image processing module, used to process and analyze the acquired spectral data in real time, including high-precision slit design and inhomogeneity correction technology; An identification module configured to identify any petroleum-based contaminants in real time through an artificial intelligence algorithm with an identification accuracy rate of not less than 85%; The evolution trend analysis module is used to quantitatively describe the evolution trend of oil film thickness and oil film type in the monitoring area based on the spatial distribution inversion model.

2. A hyperspectral monitoring system platform according to claim 1, characterized in that: The spectral imaging module includes a high-resolution spectral imaging sensor, an optical lens system, an optical fiber bundle and a spectrometer. The optical lens system focuses light onto the spectral imaging sensor and transmits spectral data to the spectrometer for processing through the optical fiber bundle.

3. A hyperspectral monitoring system platform according to claim 1, characterized in that: The data acquisition module includes a high-speed data acquisition card, a processor unit and a memory. The data output by the spectral imaging module is acquired in real time by the high-speed data acquisition card, then transmitted to the processor unit for real-time processing, and stored in the memory.

4. A hyperspectral monitoring system platform according to claim 1, characterized in that: The image processing module includes an image processor, a digital signal processor and an image stitching algorithm. The data output by the data acquisition module is processed by the image processor for image enhancement, slit design and inhomogeneity correction.

5. The hyperspectral monitoring system platform according to claim 1, characterized in that: The recognition module includes an artificial intelligence processor, a classification algorithm and a real-time recognition library. The data processed by the image processing module is input into the recognition module, and the real-time recognition algorithm is executed by the artificial intelligence processor.

6. A hyperspectral monitoring system platform according to claim 1, characterized in that: The evolution trend analysis module includes an evolution trend analyzer, a spatial distribution inversion model and a data visualization module. The data collected by the data collection module is input into the evolution trend analysis module for further analysis.

7. The hyperspectral monitoring system platform according to claim 1, characterized in that: The spectral resolution of the spectral imaging module is ≤2nm.

8. The hyperspectral monitoring system platform according to claim 1, characterized in that: The image processing module also includes a wide-field-of-view large-format image stitching technology that can cover a wide area and provide continuous high-resolution images.

9. The hyperspectral monitoring system platform according to claim 1, characterized in that: The identification module can identify the oil film on the first and second types of clean water surface in real time, and the identification accuracy rate is not less than 90%.

10. The hyperspectral monitoring system platform according to claim 1, characterized in that: The data acquisition module is installed on a shore-based platform with a height not exceeding 10 meters, and its spatial resolution is 0.08m.

Images