CN118130394B - Water-air cooperation type water quality monitoring system and method based on spectral imaging - Google Patents
Water-air cooperation type water quality monitoring system and method based on spectral imaging Download PDFInfo
- Publication number
- CN118130394B CN118130394B CN202410226162.0A CN202410226162A CN118130394B CN 118130394 B CN118130394 B CN 118130394B CN 202410226162 A CN202410226162 A CN 202410226162A CN 118130394 B CN118130394 B CN 118130394B
- Authority
- CN
- China
- Prior art keywords
- data
- water
- water quality
- remote sensing
- mobile monitoring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/58—Extraction of image or video features relating to hyperspectral data
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
- G06V10/803—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of input or preprocessed data
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/17—Terrestrial scenes taken from planes or by drones
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Remote Sensing (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Spectrometry And Color Measurement (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to the technical field of water quality monitoring systems, in particular to a water-air cooperation type water quality monitoring system and method based on spectral imaging. The system comprises: a data acquisition module; a data transmission and processing module; the processing and storage module of the spectrum data; the processing and analyzing module of the remote sensing image; the evaluation and prediction module of the water quality parameters; an adaptive calibration strategy module. According to the invention, the detection precision of the remote sensing spectrometry under the illumination conditions of different areas can be effectively improved by carrying out multiple calibration on the spectrum data; by adopting the water-air collaborative navigation mode, the safety and the accuracy of each mobile monitoring terminal in the task execution process can be ensured, and the problems of tracking and positioning of the air and water targets in a complex environment can be effectively overcome; through the adaptive training of the multi-light scene, the influence of the illumination condition change on the water quality detection precision of the water quality monitoring system can be reduced, so that the water quality monitoring effect with higher precision and stronger adaptability is realized.
Description
Technical Field
The invention relates to the technical field of water quality monitoring systems, in particular to a water-air cooperation type water quality monitoring system and method based on spectral imaging.
Background
The remote sensing spectrometry is a technology for realizing non-contact and remote detection of the composition of earth surface materials and environmental conditions by analyzing the response characteristics of the earth objects to electromagnetic radiation with different wavelengths, and has the capability of synchronously observing and acquiring information in a large area and simultaneously detecting various water quality parameters, and also has higher monitoring efficiency and accuracy, so that the remote sensing spectrometry is widely applied to the field of water quality monitoring and is commonly used for early finding and tracking pollution changes of water bodies.
However, the detection accuracy of the remote sensing spectroscopy under the illumination conditions of different areas may be affected by various factors, which are specifically expressed as follows: on one hand, due to the fact that illumination is uneven, such as factors of cloud cover, fog diffusion, sun altitude angle, incident angle and the like, which change with time and season, the radiation brightness of the surface of the ground object changes, so that the spectrum response of the surface of the ground object on a remote sensing image is different from a standard state, the shadow effect in complex topography also easily covers real spectrum information, classification recognition errors are increased, spectrum differences among ground objects in the same area are increased, and classification and quantitative analysis are affected; on the other hand, under the complex illumination condition, the influence of the atmosphere on the optical signals is enhanced, so that the atmosphere correction process is more complex and the accuracy is reduced; in addition, the albedo of the ground object varies due to different illumination conditions, which may also cause a change in the shape or position of the spectrum curve, and thus reduce the accuracy of spectrum matching and identification. Therefore, how to optimize the remote sensing data processing technology under various illumination environments to improve the detection precision of the remote sensing spectrometry is a current urgent problem to be solved.
Disclosure of Invention
In view of the shortcomings of the prior art, the invention provides a water-air cooperation type water quality monitoring system and method based on spectral imaging.
A water-air cooperative water quality monitoring system based on spectral imaging, comprising:
the data acquisition module comprises an air mobile monitoring terminal, a water mobile monitoring terminal and a remote sensing monitoring terminal and is used for acquiring water quality monitoring data of a target water area;
the data transmission and processing module is used for processing and analyzing the data acquired by the monitoring terminal and distributing part of data processing tasks to nearby server terminals for cooperative processing through a high-speed wireless communication network;
The spectrum data processing and storing module is used for carrying out noise reduction and spurious removal on the spectrum data, and carrying out safe encryption and periodical backup on the data;
the processing and analyzing module is used for carrying out image segmentation, feature extraction and classification operation on the remote sensing images, and combining multi-scale fusion strategies to fuse the remote sensing images with different resolutions so as to obtain fine landform information;
The water quality parameter evaluation and prediction module is used for constructing an evaluation model of various water quality parameters based on the remote sensing image and the water quality data, and predicting the water quality parameters by using a deep learning method so as to early warn water quality abnormal conditions in advance;
And the self-adaptive calibration strategy module is used for executing a self-adaptive calibration strategy and carrying out cooperative calibration by utilizing the spectrum data of the air mobile monitoring terminal and the plurality of water mobile monitoring terminals.
A water-air cooperation type water quality monitoring method based on spectrum imaging comprises the following steps:
S1, deploying an on-water mobile monitoring terminal carrying a spectrometer in a target water area, acquiring and calibrating spectrum data under different illumination conditions according to a path, simultaneously carrying out same-step high-altitude remote sensing and water surface monitoring in the same time period, collecting spectrum data of the same coordinate position, calibrating by combining satellite remote sensing data, processing and analyzing the acquired data, and establishing an accurate relation model between ground actual measurement and remote sensing inversion results;
S2, eliminating deviation caused by atmospheric scattering, absorption and aerosol in spectrum data acquired by a high-altitude remote sensing platform through dark pixel correction, relative radiation correction and absolute radiation correction technologies respectively, and simulating various illumination conditions by utilizing a multi-source data assimilation algorithm so as to correct earth surface reflectivity change and measurement errors caused by uneven illumination;
S3, aiming at the water area environmental characteristics, dividing the water area to be detected into a plurality of illumination scene categories, and independently training a water quality parameter inversion model for each scene to enhance the pertinence and the accuracy of the model, and simultaneously, carrying out feedback correction and optimization on the remote sensing inversion model by utilizing field monitoring data based on a monitoring system so as to keep the performance of the remote sensing inversion model stable under various illumination conditions;
S4, comprehensively utilizing multi-angle and multi-period spectrum data from the remote sensing monitoring terminal, the air mobile monitoring terminal and the water mobile monitoring terminal, constructing a water quality parameter inversion model by combining a deep learning technology, analyzing the trend of the water quality parameter in the long-term continuous remote sensing data along with the change of time and illumination conditions, and further optimizing model parameters to adapt to the water quality monitoring requirements under various illumination environments.
Preferably, the specific steps of the step S1 are as follows:
s11, selecting a proper point distribution scheme according to a water area to be detected, and installing and configuring a water mobile monitoring terminal with a spectrum sensor;
S12, setting a timing or triggering type acquisition mode, and keeping the acquisition time of the spectrum sensor overlapped with the acquisition time period of the high-altitude remote sensing data;
S13, measuring the water body transmission and reflection spectrums under various wavelengths by utilizing a spectrum sensor of the water mobile monitoring terminal;
S14, collecting high-altitude remote sensing data and water surface spectrometer data in the same time period;
s15, the two groups of data are corresponding to each other according to geographic position and time, and the difference of water quality parameters is compared and analyzed;
S16, establishing a statistical relation model between the ground actual measurement and the remote sensing inversion, and correcting and cross-verifying the data based on the statistical relation model.
Preferably, the specific steps of the step S2 are as follows:
s21, collecting radiation brightness data of the same period and the same region;
S22, calculating aerosol content, humidity and gas absorption coefficient affecting spectrum signals in the atmosphere by using an atmosphere transmission model;
S23, correcting and removing noise by using dark pixels, and performing relative radiation correction by using bright pixels to obtain apparent spectral reflectivity;
s24, comparing and calibrating with reference to the actually measured solar constant or the synchronous satellite data on the ground to finish absolute radiation correction;
s25, calculating the sun incidence angle and illumination intensity of the target water area at different time and space positions;
S26, simulating actual reflection characteristics of the water body under different illumination conditions by using a multi-source data assimilation algorithm, and correcting the remote sensing image correspondingly.
Preferably, the specific steps of the step S3 are as follows:
s31, dividing the water area to be measured into a plurality of illumination scene categories according to the environmental characteristics and illumination conditions of the water area;
s32, respectively carrying out feature extraction and normalization processing on the data of each scene category;
S33, training different water quality parameter inversion models aiming at different illumination scenes, and evaluating the performance of the models by using cross verification;
S34, periodically acquiring the latest water quality actual measurement data from the field monitoring equipment, and taking the latest water quality actual measurement data as a basis for iteratively updating the model;
s35, continuously optimizing the model structure and parameters according to the new data, and improving the model prediction accuracy through online learning or incremental learning;
s36, combining learning experience and field actual conditions, and adjusting a model structure and a calibration method to form a working flow of dynamic adjustment and continuous optimization.
Preferably, the specific steps of the step S4 are as follows:
S41, integrating spectrum data of different remote sensing platforms in the same region under different time periods and different visual angles;
s42, performing radiometric calibration, geometric correction and cloud cover removal operation on the data;
S43, constructing a fusion model by using a machine learning algorithm, and extracting water quality parameters by combining multi-angle and multi-time phase characteristics;
s44, constructing a time sequence database, and recording long-term remote sensing and field monitoring data;
s45, analyzing the trend of the water quality parameters along with the change of time and illumination and seasonal rules;
S46, the feedback data of the influence of the illumination trend and the law on the water quality are combined, inversion model parameters are optimized, and the adaptability and the prediction capability of the model under complex illumination conditions are improved.
Preferably, the method further comprises the step of optimizing the path of the mobile monitoring terminal, and specifically comprises the following steps:
based on the air mobile monitoring terminal carrying a laser ranging and communication system, transmitting laser pulses to the target water mobile monitoring terminal to guide the target water mobile monitoring terminal to synchronously move along with the air mobile monitoring terminal, meanwhile, analyzing reflection and scattering characteristics of laser when the laser propagates in air, and judging whether an air barrier exists or not in real time by utilizing a high-sensitivity sensor and an intelligent algorithm;
And integrating the data acquired by the aerial mobile monitoring terminal with the underwater environment data collected by the water mobile monitoring terminal to construct a three-dimensional space situation map so as to accurately confirm whether the aerial blockage exists or not, and dynamically optimizing the mobile monitoring paths of the aerial mobile monitoring terminal and the water mobile monitoring terminal according to the three-dimensional space situation map.
Preferably, the method further comprises the step of carrying out adaptive calibration on each mobile monitoring terminal, wherein the adaptive calibration specifically comprises the following steps:
dividing the water body into a plurality of areas, wherein each area has relatively fixed illumination conditions and water quality parameters;
Grouping the mobile monitoring terminals in each area, and periodically exchanging calibration data by the member mobile monitoring terminals of each group to reflect the change of the spectrum conditions in the group;
When a group of member mobile monitoring terminals reach a new area, firstly, carrying out preliminary processing on spectrum data of the group of member mobile monitoring terminals, extracting main color and reflectivity characteristics, then comparing the characteristics with the existing calibration data, and calculating the deviation of a spectrum response curve;
According to the deviation value, a new calibration weight is distributed to the member mobile monitoring terminals of the current group, so that the spectral response curve of each member mobile monitoring terminal gradually converges to a global optimal state;
after the complete self-adaptive calibration is performed once, the calibration weight information of the mobile monitoring terminals of the members of the current group is shared with the members of other groups so as to achieve the global optimal solution.
Preferably, the aerial mobile monitoring terminal is an unmanned aerial vehicle, the water mobile monitoring terminal is an intelligent buoy or an unmanned ship, and the remote sensing monitoring terminal is a remote sensing satellite.
Preferably, the water quality parameters include chlorophyll a concentration, turbidity, chemical oxygen demand, dissolved organics, suspended particulate matter, and nitrogen and phosphorus compounds.
The beneficial effects of the invention are as follows: according to the invention, the detection precision of the remote sensing spectrometry under the illumination conditions of different areas can be effectively improved by carrying out multiple calibration on the spectrum data; by adopting the water-air collaborative navigation mode, the safety and the accuracy of each mobile monitoring terminal in the task execution process can be ensured, and the problems of tracking and positioning of the air and water targets in a complex environment can be effectively overcome; through the adaptive training of the multi-light scene, the influence of the illumination condition change on the water quality detection precision of the water quality monitoring system can be reduced, so that the water quality monitoring effect with higher precision and stronger adaptability is realized.
Drawings
FIG. 1 is a schematic structural diagram of a water-air cooperative water quality monitoring system based on spectral imaging according to an embodiment of the present invention;
Fig. 2 is a schematic flow chart of a water-air cooperation type water quality monitoring method based on spectral imaging according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a water-air cooperative water quality monitoring system based on spectral imaging includes:
and a data acquisition module: the system comprises an air mobile monitoring terminal, a water mobile monitoring terminal and a remote sensing monitoring terminal, wherein the remote sensing monitoring terminal is used for acquiring water quality monitoring data of a target water area; the aerial mobile monitoring terminal is an unmanned aerial vehicle, the water mobile monitoring terminal is an intelligent buoy or an unmanned ship, and the remote sensing monitoring terminal is a remote sensing satellite.
And the data transmission and processing module is used for: the system is used for processing the data acquired by the monitoring terminal and distributing part of the data processing tasks to nearby server terminals for cooperative processing through a high-speed wireless communication network, and can use a high-speed wireless communication technology (such as 5G, 6G and the like) and a data compression algorithm to reduce the data transmission time and bandwidth requirements, and an edge computing technology is used for distributing part of the data processing tasks to the nearby server terminals, so that the pressure of a central server can be relieved;
And the spectrum data processing and storing module is used for: the method is used for carrying out noise reduction and spurious elimination on the spectrum data, carrying out safe encryption and periodical backup on the data, such as carrying out noise reduction and spurious elimination on the spectrum data by adopting wavelet transformation and a convolutional neural network, setting a periodical backup mechanism, and carrying out safe encryption on the data;
And a remote sensing image processing and analyzing module: the method is used for carrying out image segmentation, feature extraction and classification operations on the remote sensing images, combining a multi-scale fusion strategy to fuse the remote sensing images with different resolutions so as to obtain fine landform information, and adopting a deep learning technology (such as a convolutional neural network, a cyclic neural network and the like) for processing the remote sensing images so as to realize more accurate image segmentation, feature extraction and classification, and using the multi-scale fusion strategy to fuse the remote sensing images with different resolutions so as to obtain finer landform information;
and a water quality parameter evaluation and prediction module: the method is used for constructing evaluation models (such as a support vector machine, an artificial neural network and the like) of various water quality parameters based on remote sensing images and water quality data, and predicting the water quality parameters by using a machine learning method so as to early warn water quality abnormal conditions in advance;
An adaptive calibration strategy module: the method is used for executing a self-adaptive calibration strategy, carrying out cooperative calibration by utilizing the spectrum data of the air mobile monitoring terminal and the plurality of water mobile monitoring terminals, designing a self-adaptive calibration strategy for improving the detection precision of a remote sensing spectrometry under different area illumination conditions, carrying out cooperative calibration by utilizing the spectrum data of the air mobile monitoring terminal and the plurality of water mobile monitoring terminals, for example, dividing the plurality of water mobile monitoring terminals into different groups, periodically exchanging calibration data, and carrying out dynamic adjustment on the data according to the spectrum differences inside and outside the groups.
Referring to fig. 2, a water-air cooperation type water quality monitoring method based on spectral imaging comprises the following steps:
s1, acquiring and calibrating spectrum data under different illumination conditions according to a path by deploying an intelligent buoy or an unmanned ship carrying a spectrometer in a target water area so as to assist in satellite remote sensing data calibration; and simultaneously, simultaneously carrying out step-by-step aerial remote sensing and water surface monitoring in the same time period, collecting spectrum information of the same place or adjacent area, and establishing an accurate relation model between ground actual measurement and remote sensing inversion results by comparing and analyzing the data. The method comprises the following specific steps:
s11, selecting a proper point distribution strategy according to a water area to be detected, and installing and configuring an intelligent buoy or an unmanned ship with a spectrum sensor;
s12, setting a timing or triggering type acquisition mode, and ensuring that a spectrum sensor and a time period for acquiring high-altitude remote sensing data are overlapped as much as possible;
S13, measuring the water body transmission and reflection spectrums under various wavelengths by utilizing a spectrum sensor of the water mobile monitoring terminal; it is necessary to ensure that the spectroscopic sensor of the buoy/unmanned ship is calibrated strictly in the laboratory.
S14, collecting high-altitude remote sensing data and water surface spectrometer data at the same time or in a similar time period;
S15, the two groups of data are corresponding to each other according to geographic position and time, and the difference of water quality parameters (such as chlorophyll a concentration, turbidity, chemical oxygen demand, dissolved organic matters, suspended particulate matters, nitrogen and phosphorus compounds and the like) is compared and analyzed;
S16, establishing a statistical relation model between ground actual measurement and remote sensing inversion for subsequent data correction and cross verification.
S2, firstly, eliminating deviation caused by atmospheric scattering, absorption and aerosol in spectrum data acquired by a high-altitude remote sensing platform through the atmospheric correction technologies such as dark pixel correction, relative radiation correction and absolute radiation correction, and then simulating various illumination conditions (comprising different sun incidence angles and sunlight intensities) by utilizing a multi-source data assimilation algorithm such as a 6S model so as to accurately correct earth surface reflectivity change and measurement errors caused by uneven illumination; the method comprises the following specific steps:
s21, collecting radiation brightness data of the same period and the same region, wherein the radiation brightness data comprise bright pixels (water surface or land reflection) and dark pixels (approximate non-reflection region);
s22, calculating the influence of the atmosphere on the spectrum signal, such as aerosol content, humidity, gas absorption coefficient and the like by using an atmosphere transmission model;
S23, correcting and removing noise by using dark pixels, and performing relative radiation correction by using bright pixels to obtain apparent spectral reflectivity;
S24, comparing and calibrating with reference to the actually measured solar constant or the synchronous satellite data on the ground so as to finish absolute radiation correction;
s25, calculating the sun incidence angle and illumination intensity of the target water area at different time and space positions;
S26, simulating actual reflection characteristics of the water body under different illumination conditions by using a multi-source data assimilation algorithm, and correcting the remote sensing image correspondingly.
S3, subdividing the water area to be measured into various illumination scene categories according to the water area environmental characteristics, and independently training a water quality parameter inversion model for each scene, so that the pertinence and the accuracy of the model are enhanced. Meanwhile, the system continuously utilizes the field monitoring data to carry out feedback correction and optimization on the remote sensing inversion model, so that the performance of the remote sensing inversion model under various illumination conditions is ensured to be stable and reliable, the influence of illumination condition changes on the water quality detection precision of a remote sensing spectrometry is reduced, and the water quality monitoring effect with higher precision and stronger adaptability is realized; the method comprises the following specific steps:
s31, dividing the water area to be measured into a plurality of illumination scene categories according to the environmental characteristics (such as open lake surface, estuary, along sea-tangle and the like) of the water area and illumination conditions;
s32, respectively carrying out feature extraction and normalization processing on the data of each scene category;
S33, training different water quality parameter inversion models aiming at different illumination scenes, and evaluating the performance of the models by using cross verification;
S34, periodically acquiring the latest water quality actual measurement data from the field monitoring equipment, and taking the latest water quality actual measurement data as a basis for iteratively updating the model;
s35, continuously optimizing the model structure and parameters according to the new data, and improving the model prediction accuracy through online learning or incremental learning;
S36, combining expert knowledge and field actual conditions, and adjusting a model structure and a calibration method to form a dynamic adjustment and continuous optimization workflow.
S4, comprehensively utilizing multi-angle and multi-period spectrum data from a multi-source remote sensing monitoring terminal, an air mobile monitoring terminal and a water mobile monitoring terminal, constructing a more accurate water quality parameter inversion model by combining machine learning or deep learning technology, and reducing the influence of single-view illumination difference on water quality monitoring accuracy; meanwhile, through further analysis of trends of water quality parameters in the long-term continuous remote sensing data along with changes of time and illumination conditions, model parameters are further optimized, so that the model can realize a high-efficiency water quality monitoring function in various illumination environments; the method comprises the following specific steps:
S41, integrating spectrum data of different remote sensing platforms (such as remote sensing satellites and unmanned aerial vehicles) in the same area under different time periods and different visual angles;
s42, preprocessing the optical data, including radiometric calibration, geometric correction, cloud cover removal and the like;
s43, constructing a fusion model by using a machine learning algorithm (such as a random forest, a support vector machine, a deep neural network and the like), and extracting water quality parameters by combining multi-angle and multi-time phase characteristics;
s44, constructing a time sequence database, and recording long-term remote sensing and field monitoring data;
s45, analyzing the trend of the water quality parameters along with the change of time and illumination and seasonal rules;
S46, comprehensively considering water quality parameter change data caused by influence factors of illumination change, optimizing inversion model parameters, and improving adaptability and prediction capability of the model under complex illumination conditions.
As a preferred embodiment of the present invention, the method further includes path optimization for the mobile monitoring terminal, and specifically includes the following steps:
the unmanned aerial vehicle is used for carrying a laser ranging and communication system, accurate laser pulses are emitted to a target intelligent buoy to guide the target intelligent buoy to move along with the unmanned aerial vehicle, in the process, the characteristics of reflection, scattering and the like of laser when the laser propagates in the air are analyzed, and a high-sensitivity sensor and an intelligent algorithm are used for judging whether an air barrier exists in real time; the two-way communication link established between the unmanned aerial vehicle and the intelligent buoy rapidly transmits the information of the air obstacle to the in-water mobile monitoring terminal, so that the risk is avoided in real time;
Simultaneously, unmanned aerial vehicle carries out the synchronous comparison of water space data with intelligent buoy: the data acquired in the air and the underwater environment data collected by the underwater mobile monitoring terminal are integrated together to jointly construct a three-dimensional space situation map, so that whether the air blockage exists or not is accurately confirmed, and the moving scheme of the unmanned aerial vehicle and the intelligent buoy is dynamically optimized according to the situation.
As another preferred embodiment of the invention, the method further comprises the step of adaptively calibrating each mobile monitoring terminal, and specifically comprises the following steps:
dividing the water body into a plurality of areas, wherein each area has relatively fixed illumination conditions and water quality parameters;
Dividing the mobile monitoring terminals in each area into a plurality of groups, wherein the number of the members of each group can be determined according to actual conditions, and the mobile monitoring terminals of the members of each group exchange calibration data periodically so as to reflect the change of the spectrum conditions in the group;
When a group of member mobile monitoring terminals reach a new area, firstly, carrying out preliminary processing on spectrum data of the group of member mobile monitoring terminals, extracting main color and reflectivity characteristics, then comparing the characteristics with the existing calibration data, and calculating the deviation of a spectrum response curve;
according to the deviation value, a new calibration weight is distributed to the member mobile monitoring terminals of the current group, and the new calibration weight is calculated by combining the calibration weights of all the member mobile monitoring terminals under the condition that a certain confidence coefficient requirement is met, so that the spectral response curve of each member mobile monitoring terminal gradually converges to a global optimal state;
after the complete self-adaptive calibration is performed once, the calibration weight information of the mobile monitoring terminals of the members of the current group is shared with the members of other groups so as to achieve the global optimal solution.
The above embodiments are merely illustrative of the principles of the present invention and its effectiveness, and are not intended to limit the invention. Modifications and variations may be made to the above-described embodiments by those skilled in the art without departing from the spirit and scope of the invention. Accordingly, it is intended that all equivalent modifications and variations of the invention be covered by the claims, which are within the ordinary skill of the art, be within the spirit and scope of the present disclosure.
Claims (6)
1. The water-air cooperation type water quality monitoring method based on the spectral imaging is characterized by adopting a water-air cooperation type water quality monitoring system based on the spectral imaging, and the water-air cooperation type water quality monitoring system based on the spectral imaging comprises the following steps:
the data acquisition module comprises an air mobile monitoring terminal, a water mobile monitoring terminal and a remote sensing monitoring terminal and is used for acquiring water quality monitoring data of a target water area;
the data transmission and processing module is used for processing and analyzing the data acquired by the monitoring terminal and distributing part of data processing tasks to nearby server terminals for cooperative processing through a high-speed wireless communication network;
The spectrum data processing and storing module is used for carrying out noise reduction and spurious removal on the spectrum data, and carrying out safe encryption and periodical backup on the data;
the processing and analyzing module is used for carrying out image segmentation, feature extraction and classification operation on the remote sensing images, and combining multi-scale fusion strategies to fuse the remote sensing images with different resolutions so as to obtain fine landform information;
The water quality parameter evaluation and prediction module is used for constructing an evaluation model of various water quality parameters based on the remote sensing image and the water quality monitoring data, and predicting the water quality parameters by using a deep learning method so as to early warn the abnormal condition of the water quality in advance;
The self-adaptive calibration strategy module is used for executing a self-adaptive calibration strategy and carrying out cooperative calibration by utilizing spectrum data of the air mobile monitoring terminal and the plurality of water mobile monitoring terminals;
The water-air cooperation type water quality monitoring method based on spectral imaging specifically comprises the following steps:
S1, deploying an overwater mobile monitoring terminal carrying a spectrometer in a target water area, acquiring and calibrating spectral data under different illumination conditions according to a path, simultaneously carrying out same-step high-altitude remote sensing and water surface monitoring in the same time period, collecting spectral data of the same coordinate position, calibrating by combining the high-altitude remote sensing data, processing and analyzing the acquired data, and establishing an accurate relation model between ground actual measurement and remote sensing inversion results;
S2, eliminating deviation caused by atmospheric scattering, absorption and aerosol in spectrum data acquired by a high-altitude remote sensing platform through dark pixel correction, relative radiation correction and absolute radiation correction technologies respectively, and simulating various illumination conditions by utilizing a multi-source data assimilation algorithm so as to correct earth surface reflectivity change and measurement errors caused by uneven illumination;
S3, aiming at the water area environmental characteristics, dividing the water area to be detected into a plurality of illumination scene categories, and independently training a water quality parameter inversion model for each scene to enhance the pertinence and the accuracy of the model, and simultaneously, carrying out feedback correction and optimization on the remote sensing inversion model by utilizing field monitoring data based on a monitoring system to keep the stability of the performance of the remote sensing inversion model under various illumination conditions, wherein the specific steps of the step S3 are as follows:
s31, dividing the water area to be measured into a plurality of illumination scene categories according to the environmental characteristics and illumination conditions of the water area;
s32, respectively carrying out feature extraction and normalization processing on the data of each scene category;
S33, training different water quality parameter inversion models aiming at different illumination scenes, and evaluating the performance of the models by using cross verification;
S34, periodically acquiring the latest water quality actual measurement data from the field monitoring equipment, and taking the latest water quality actual measurement data as a basis for iteratively updating the model;
s35, continuously optimizing the model structure and parameters according to the new data, and improving the model prediction accuracy through online learning or incremental learning;
s36, combining learning experience and field actual conditions, and adjusting a model structure and a calibration method to form a working flow of dynamic adjustment and continuous optimization;
S4, comprehensively utilizing multi-angle and multi-period spectrum data from a remote sensing monitoring terminal, an air mobile monitoring terminal and a water mobile monitoring terminal, constructing a water quality parameter inversion model by combining a deep learning technology, analyzing the trend of the water quality parameter in the long-term continuous remote sensing data along with the change of time and illumination conditions, and further optimizing model parameters to adapt to the water quality monitoring requirements under various illumination environments, wherein the specific steps of the step S4 are as follows:
s41, integrating spectrum data of different remote sensing platforms in the same area under different time periods and different visual angles;
s42, performing radiometric calibration, geometric correction and cloud cover removal operation on the data;
S43, constructing a fusion model by using a machine learning algorithm, and extracting water quality parameters by combining multi-angle and multi-time phase characteristics;
s44, constructing a time sequence database, and recording long-term remote sensing and field monitoring data;
s45, analyzing the trend of the water quality parameters along with the change of time and illumination and seasonal rules;
S46, the feedback data of the influence of the illumination change trend and the seasonal rule on the water quality are combined, inversion model parameters are optimized, and the adaptability and the prediction capability of the model under complex illumination conditions are improved;
The method also comprises the step of optimizing the path of the mobile monitoring terminal, and specifically comprises the following steps:
Based on the air mobile monitoring terminal carrying a laser ranging and communication system, transmitting laser pulses to the target water mobile monitoring terminal to guide the target water mobile monitoring terminal to synchronously move along with the air mobile monitoring terminal, meanwhile, analyzing reflection and scattering characteristics of laser when the laser propagates in air, and judging whether an air barrier exists or not in real time by utilizing a high-sensitivity sensor and an intelligent algorithm;
And integrating the data acquired by the aerial mobile monitoring terminal with the underwater environment data collected by the water mobile monitoring terminal to construct a three-dimensional space situation map so as to accurately confirm whether the aerial blockage exists or not, and dynamically optimizing the mobile monitoring paths of the aerial mobile monitoring terminal and the water mobile monitoring terminal according to the three-dimensional space situation map.
2. The water-air cooperation type water quality monitoring method based on spectral imaging according to claim 1, wherein the specific steps of the step S1 are as follows:
s11, selecting a proper point distribution scheme according to a water area to be detected, and installing and configuring a water mobile monitoring terminal with a spectrometer;
S12, setting a timing or triggering type acquisition mode, and keeping the acquisition time of the spectrometer overlapped with the acquisition time period of the high-altitude remote sensing data;
S13, measuring the water body transmission and reflection spectrums under various wavelengths by utilizing a spectrometer of the water mobile monitoring terminal;
S14, collecting high-altitude remote sensing data and water surface spectrometer data in the same time period;
s15, the two groups of data are corresponding to each other according to geographic position and time, and the difference of water quality parameters is compared and analyzed;
S16, establishing a statistical relation model between the ground actual measurement and the remote sensing inversion, and correcting and cross-verifying the data based on the statistical relation model.
3. The water-air cooperation type water quality monitoring method based on spectral imaging according to claim 1, wherein the specific steps of the step S2 are as follows:
s21, collecting radiation brightness data of the same period and the same region;
S22, calculating aerosol content, humidity and gas absorption coefficient affecting spectrum signals in the atmosphere by using an atmosphere transmission model;
S23, correcting and removing noise by using dark pixels, and performing relative radiation correction by using bright pixels to obtain apparent spectral reflectivity;
s24, comparing and calibrating with reference to the actually measured solar constant or the synchronous satellite data on the ground to finish absolute radiation correction;
S25, calculating the sun incidence angle and illumination intensity of the target water area at different time and space positions;
S26, simulating actual reflection characteristics of the water body under different illumination conditions by using a multi-source data assimilation algorithm, and correcting the remote sensing image correspondingly.
4. The water-air collaborative water quality monitoring method based on spectral imaging according to claim 1, further comprising performing adaptive calibration on each mobile monitoring terminal, wherein the adaptive calibration specifically comprises the following steps:
dividing the water body into a plurality of areas, wherein each area has relatively fixed illumination conditions and water quality parameters;
Grouping the mobile monitoring terminals in each area, and periodically exchanging calibration data by the member mobile monitoring terminals of each group to reflect the change of the spectrum conditions in the group;
When a group of member mobile monitoring terminals reach a new area, firstly, carrying out preliminary processing on spectrum data of the group of member mobile monitoring terminals, extracting main color and reflectivity characteristics, then comparing the characteristics with the existing calibration data, and calculating the deviation of a spectrum response curve;
According to the deviation value, a new calibration weight is distributed to the member mobile monitoring terminals of the current group, so that the spectral response curve of each member mobile monitoring terminal gradually converges to a global optimal state;
after the complete self-adaptive calibration is performed once, the calibration weight information of the mobile monitoring terminals of the members of the current group is shared with the members of other groups so as to achieve the global optimal solution.
5. The water-air collaborative water quality monitoring method based on spectral imaging according to claim 1, wherein the air mobile monitoring terminal is an unmanned aerial vehicle, the water mobile monitoring terminal is an intelligent buoy or unmanned ship, and the remote sensing monitoring terminal is a remote sensing satellite.
6. A method for collaborative water quality monitoring based on spectral imaging according to claim 1 wherein the water quality parameters include chlorophyll a concentration, turbidity, chemical oxygen demand, dissolved organics, suspended particulate matter, and nitrogen-phosphorous compounds.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410226162.0A CN118130394B (en) | 2024-02-29 | 2024-02-29 | Water-air cooperation type water quality monitoring system and method based on spectral imaging |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410226162.0A CN118130394B (en) | 2024-02-29 | 2024-02-29 | Water-air cooperation type water quality monitoring system and method based on spectral imaging |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118130394A CN118130394A (en) | 2024-06-04 |
| CN118130394B true CN118130394B (en) | 2024-10-15 |
Family
ID=91237177
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410226162.0A Active CN118130394B (en) | 2024-02-29 | 2024-02-29 | Water-air cooperation type water quality monitoring system and method based on spectral imaging |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118130394B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119935917B (en) * | 2025-03-11 | 2025-07-18 | 西安欣源测控技术有限公司 | Old-goaf water quality detection system based on multi-parameter detection technology |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110865040A (en) * | 2019-11-29 | 2020-03-06 | 深圳航天智慧城市系统技术研究院有限公司 | Sky-ground integrated hyperspectral water quality monitoring and analyzing method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2547359C (en) * | 2003-11-26 | 2012-11-27 | Florida Environmental Research Institute, Inc. | Spectral imaging system |
| CN209727916U (en) * | 2019-01-16 | 2019-12-03 | 河北华厚天成环保技术有限公司 | Water quality on-line analysis system |
| CN110887800B (en) * | 2019-11-27 | 2021-01-15 | 中国科学院西安光学精密机械研究所 | A Data Calibration Method for Spectroscopic Water Quality Online Monitoring System |
| CN116577475A (en) * | 2023-04-13 | 2023-08-11 | 北京工业大学 | A fully autonomous water quality detection unmanned ship system |
-
2024
- 2024-02-29 CN CN202410226162.0A patent/CN118130394B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110865040A (en) * | 2019-11-29 | 2020-03-06 | 深圳航天智慧城市系统技术研究院有限公司 | Sky-ground integrated hyperspectral water quality monitoring and analyzing method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118130394A (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liang et al. | Close-range remote sensing of forests: The state of the art, challenges, and opportunities for systems and data acquisitions | |
| US12072242B2 (en) | Hyperspectral sensing system and processing methods for hyperspectral data | |
| Thompson et al. | Real-time remote detection and measurement for airborne imaging spectroscopy: a case study with methane | |
| US11650190B2 (en) | Hyperspectral sensing system and methods | |
| US20240361200A1 (en) | Apparatuses and methods for gas flux measurements | |
| KR101863123B1 (en) | System for mapping river water-bloom map using automatic driving unmanned air vehicle and unmanned floating body of moving type | |
| Damm et al. | FLD-based retrieval of sun-induced chlorophyll fluorescence from medium spectral resolution airborne spectroscopy data | |
| Moses et al. | Atmospheric correction for inland waters | |
| CN112051226A (en) | Method for estimating total suspended matter concentration of offshore area based on unmanned aerial vehicle-mounted hyperspectral image | |
| CN112348882A (en) | Low-altitude target tracking information fusion method and system based on multi-source detector | |
| CN118130394B (en) | Water-air cooperation type water quality monitoring system and method based on spectral imaging | |
| CN113155740A (en) | Calibration reference field BRDF characteristic analysis method and system | |
| Biney et al. | Bor uvka | |
| Liu et al. | Multi-temporal monitoring of urban river water quality using UAV-borne multi-spectral remote sensing | |
| CN119762848A (en) | Water pollution source identification method and system based on unmanned aerial vehicle multispectral image | |
| Tulet et al. | MAP-IO: an atmospheric and marine observatory program on board Marion Dufresne over the Southern Ocean | |
| EP4412442B1 (en) | Method and system for providing a site specific fertilizer recommendation | |
| CN117952450A (en) | Ecological environment quality factor analysis method and device, storage medium and processor | |
| KR102727554B1 (en) | Method for artificial intelligence learning of algae using ground-mounted hyperspectral sensor unit and systems for predicting algae | |
| Berezowski et al. | UAV-Based Hyperspectral Ultraviolet-Visible Interpolated Reflectance Images for Remote Sensing of Leaf Area Index | |
| Zhang et al. | Comparison of scanning aerosol lidar and in situ measurements of aerosol physical properties and boundary layer heights | |
| CN119915390B (en) | Atmospheric correction method and device for radiation signals of coastal waters under zenith angles of sun | |
| CN116380811B (en) | A cloud detection method based on oxygen absorption bands | |
| Mandal et al. | Application of Prisma Hyperspectral Data for PM 2.5 Estimation: A Case Study on New Delhi, India | |
| CN120318681A (en) | A method and device for determining the optimal vegetation index of Spartina alterniflora |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |