CN118015551A - Floating island type monitoring system applied to field ecological wetland - Google Patents

Floating island type monitoring system applied to field ecological wetland Download PDF

Info

Publication number
CN118015551A
CN118015551A CN202410417404.4A CN202410417404A CN118015551A CN 118015551 A CN118015551 A CN 118015551A CN 202410417404 A CN202410417404 A CN 202410417404A CN 118015551 A CN118015551 A CN 118015551A
Authority
CN
China
Prior art keywords
species
monitoring
floating island
module
feature
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.)
Granted
Application number
CN202410417404.4A
Other languages
Chinese (zh)
Other versions
CN118015551B (en
Inventor
陈柲安
李军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong Shirong Information Technology Co ltd
Original Assignee
Shandong Shirong Information Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shandong Shirong Information Technology Co ltd filed Critical Shandong Shirong Information Technology Co ltd
Priority to CN202410417404.4A priority Critical patent/CN118015551B/en
Publication of CN118015551A publication Critical patent/CN118015551A/en
Application granted granted Critical
Publication of CN118015551B publication Critical patent/CN118015551B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G01D21/02—Measuring two or more variables by means not covered by a single other subclass
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
    • G06V10/449—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters
    • G06V10/451—Biologically inspired filters, e.g. difference of Gaussians [DoG] or Gabor filters with interaction between the filter responses, e.g. cortical complex cells
    • G06V10/454—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
    • G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06V10/54—Extraction of image or video features relating to texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06V10/806—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of extracted features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/40—Monitoring or fighting invasive species

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Data Mining & Analysis (AREA)
  • Molecular Biology (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to the technical field of wetland monitoring, in particular to a floating island type monitoring system applied to a field ecological wetland. The system comprises a floating island monitoring unit, wherein the floating island monitoring unit is used for monitoring and recording the ecological environment of the wetland and the ecological behaviors of the green-head submerged ducks; the video monitoring module is used for recording the behavior activities of the green-head submerged ducks and comprises a species identification module, a species classification module and an abnormality early warning module; the storage unit is used for storing monitoring data and aquatic bird species information, the wireless transmission unit is used for carrying out wireless connection on floating island monitoring units distributed at different positions, and a wireless local area network is constructed between monitoring points; the power supply unit supplies power to the floating island monitoring unit and the wireless transmission unit. The multi-mode fusion algorithm is adopted, and the wetland environment parameters are brought into the feature vectors, so that the species can be represented by using richer information sources, the identification accuracy of the foreign invasive species is improved, and the multi-mode fusion is beneficial to reducing misjudgment caused by the influence of the environmental conditions.

Description

Floating island type monitoring system applied to field ecological wetland
Technical Field
The invention relates to the technical field of wetland monitoring, in particular to a floating island type monitoring system applied to a field ecological wetland.
Background
The floating island type monitoring system of the field ecological wetland is a comprehensive facility combining an ecological floating island technology and a modern water quality monitoring technology, and is mainly applied to observation of wild duck population reproduction processes of the field ecological wetland (swamp lake, reed swaying and the like), and short-distance long-time observation of living habits such as nest taking, hatching and the like of green-head submerged ducks.
The existing monitoring system for the green-head submerged ducks depends on a single monitoring means, has certain limitation on species identification and classification, and is particularly influenced by factors such as environmental conditions, illumination changes, individual differences and the like in a complex and changeable wetland environment, so that the accuracy of species identification is insufficient, and the floating island type monitoring system is designed and applied to field ecological wetland.
Disclosure of Invention
The invention aims to provide a floating island type monitoring system for a field ecological wetland, which aims to solve the problems that the existing monitoring system for green-head submerged ducks provided in the background art relies on a single monitoring means, has certain limitation on the identification and classification of species, and is influenced by factors such as environmental conditions, illumination changes, individual differences and the like in a complex and changeable wetland environment, so that the accuracy of species identification is insufficient.
To achieve the above object, the present invention provides a floating island type monitoring system for a field ecological wetland, comprising
The floating island monitoring unit is used for monitoring and recording the ecological environment of the wetland and the ecological behaviors of the green-head submerged ducks, and comprises a multi-path monitoring module and a video monitoring module;
the multi-path monitoring module is used for collecting and recording wetland environment parameters, wherein the wetland environment parameters comprise water temperature, pH value, dissolved oxygen content, turbidity and illumination intensity;
The video monitoring module is used for recording the behavior activities of the green-head submerged ducks, carrying out species identification and analysis, providing foreign species invasion early warning, such as nesting, hatching, brooding and foraging behaviors, and providing visual and detailed data for biologists to study the life habits, the population behaviors and the social structures of the ducks, and comprises a species identification module, a species classification module and an abnormality early warning module;
the storage unit is used for storing monitoring data and aquatic bird species information, and comprises a local storage module and a cloud storage module;
The wireless transmission unit is used for wirelessly connecting floating island monitoring units distributed at different positions, constructing a wireless local area network between monitoring points, and reading monitoring data by a user accessing the cloud storage module in the local area network;
The power supply unit is used for supplying power to the floating island monitoring unit and the wireless transmission unit; specifically, the power supply unit not only supplies power for the equipment on the floating island monitoring unit, but also supplies power for the far-end network transmitting base station and the near-end receiving base station, and at least comprises a solar cell panel, a solar controller and a storage battery; the solar panel is used for converting solar radiation capacity into electric energy to be stored in the storage battery, and the solar controller is a charge-discharge controller and is used for prolonging the service life of the storage battery and preventing the storage battery from being overcharged and deeply charged.
As a further improvement of the technical scheme, the multi-channel monitoring module comprises a water quality sensor and an environment sensor;
The water quality sensor is used for monitoring and recording the temperature, the pH value, the dissolved oxygen content and the turbidity of the wetland water body;
The temperature of the wetland water body has direct influence on the life activities, propagation behaviors and microbial activities of aquatic organisms, and is helpful for researching the influence of the health condition of a water body ecological system and climate change on the wetland ecology;
the pH value reflects the acid-base degree of the water body, is one of important indexes for measuring the quality of water, and the too high or too low pH value can threaten the life of aquatic organisms, particularly green-head submerged ducks;
The content of the dissolved oxygen can reflect the self-cleaning capacity of the water body and the health state of an ecological system, and for waterfowl such as green-head submerged ducks, the sufficient dissolved oxygen is an important condition for survival and reproduction;
the turbidity can indirectly reflect the quantity of suspended matters in the water body, and the high turbidity can influence the survival and the illumination penetration depth of aquatic organisms, so that the growth of wetland vegetation and the balance of the whole ecological system are influenced;
The environmental sensor is used for monitoring the environmental temperature, illumination intensity and biomass density in the wetland environment. The environment temperature and the illumination intensity are used for evaluating the ecological environment comfort level of the wetland and the effective utilization condition of illumination resources;
the illumination intensity can directly influence the quality of the image captured by the video monitoring module, so that the accuracy of species identification is influenced; environmental temperature changes can affect animal behavior patterns and visual characteristics, such as feather status;
The environmental temperature can influence behavior habits of waterfowl, such as activity time and foraging behavior, and behavior habit changes can influence the type and quality of the image sample captured by the species identification module, so that the identification result can be influenced;
The environmental sensor data can be fed back to the species identification module, so that the species identification module can dynamically adjust an identification algorithm according to real-time environmental conditions, such as enhancing image enhancement processing when the illumination condition is poor, or preferentially considering certain characteristic variables under specific environmental conditions, and the identification efficiency and accuracy are improved;
biomass density was used to evaluate wetland plant productivity and richness of waterfowl food resources.
As a further improvement of the technical scheme, the species identification module is used for monitoring aquatic bird species in a monitoring area in real time, the species identification module identifies and distinguishes green-head submerged ducks from other aquatic bird species based on an image processing and machine learning fusion algorithm, the species classification module further identifies foreign invasive species in other aquatic bird species, and the anomaly early warning module timely gives an alarm.
As a further improvement of the technical scheme, the specific steps of the species identification module for identifying and distinguishing the green-head submerged ducks from the foreign invasive species are as follows:
s1.1, preprocessing image data of wet waterfowl captured by a species identification module; noise reduction, enhancement, scaling and cutting operations are carried out on the original image, so that the image quality is ensured to meet the subsequent analysis requirements;
S1.2, extracting contour features, texture features and key point information in image data by adopting an image processing technology, wherein the key point information comprises beak, head shape and body type feature information;
S1.3, training a machine learning model based on a multi-mode fusion algorithm by using waterfowl species information stored in a storage unit, wherein the waterfowl species information comprises contour features, texture features, key point information, a foreign species database and feature labels corresponding to different species, and identifying and distinguishing green-head ducks from other waterfowl species by the trained machine learning model according to real-time image data;
s1.4, identifying foreign invasion natural enemy species from other waterfowl species through a species classification module;
By combining data of different modes, the multi-mode fusion algorithm can utilize richer information sources to characterize species, so that the accuracy and robustness of identifying foreign invasive species are improved, and the multi-mode fusion is helpful for reducing misjudgment caused by the influence of environmental conditions or individual differences;
S1.5, after the species classification module identifies the foreign invasive natural enemy species, the abnormal early warning module timely gives an alarm, sends early warning information to a monitoring center, and records identification time, place and species image data.
As a further improvement of the present technical solution, in S1.2, the specific steps involved in the image processing technology are as follows:
s1.21, acquiring contour information in image data through a Canny algorithm;
s1.22, extracting feather texture features in an image by adopting a gray level co-occurrence matrix algorithm;
S1.23, extracting the beak, head shape and body type characteristic information by using a SIFT algorithm.
As a further improvement of the present technical solution, in S1.3, the machine learning model specifically includes:
Known feature vectors And incorporating the wetland environment parameter E into the feature vector/>In forming a multi-modal feature vector/>Multimodal feature vector/>Wherein the feature vector X comprises a contour feature C, a texture feature D, a beak feature B, a head shape feature H and a body shape feature S, and the multi-modal feature vector/>Mapping to a corresponding species tag Y, wherein the green-head duckling is/>Other waterfowl species are/>；
Aligning the feature vectors extracted from different modes, unifying the feature vectors by means of dimension reduction and dimension increase if the feature dimensions among the modes are inconsistent, then executing feature fusion operation, and selecting a weighted fusion method to fuse the features of the different modes; after the fusion operation, a multi-mode feature vector containing multi-mode information is obtainedThe vector not only contains various local features and global features of original data, but also reflects the relevance among different mode data, so that the vector has stronger expression capability and robustness, and the advantage of incorporating the wetland environment parameters into the feature vector is that the interaction of species characteristics and environment background can be reflected more comprehensively, so that the accuracy and stability of the species identification system are improved;
The probability of classifying the sample into a green head duckling is as follows:
；
The probability of a sample being classified as another waterfowl species is:
；
In the method, in the process of the invention, The probability expressed as green head ducks; /(I)Probability expressed as other waterfowl species; A weight vector representing the green head duckling; /(I) A weight vector representing other waterfowl species; /(I)The representation is/>Vector and feature vector/>For weighted summation of features; /(I)The representation is/>Vector and feature vector/>Is an inner product of (2); /(I)Representing a sigmoid function for mapping the linear prediction value into a (0, 1) interval, thereby representing a probability; /(I)A transpose operation representing a vector; /(I)A bias item representing a green-head latent duck; /(I)Bias terms representing other waterfowl species.
As a further improvement of the technical scheme, in S1.4, the specific steps of identifying the foreign invasive natural enemy species by the species classification module are as follows:
S1.41, constructing a natural enemy data set of the green-head duck according to the foreign species database and the characteristic tag stored in the storage unit, and extracting the characteristic vector of the natural enemy of the green-head duck based on the image processing technology in the step S1.2 And wetland environmental parameters/>In forming a multi-modal feature vector/>Multimodal feature vector/>Wherein the feature vectorIncluding profile features/>Texture features/>Beak characterization/>Head shape feature/>And body shape characteristics/>；
S1.42, taking a data set of natural enemies of the green-head ducks as a training set, training and optimizing weight parameters of the machine learning model in the step S1.3;
the specific process for optimizing the machine learning model is as follows:
Optimizing bias terms by minimizing loss function L And weight vector/>The minimization loss function L is then:
；
In the method, in the process of the invention, Representing the number of samples; /(I)Represents the/>True labels of the individual samples; /(I)Represents model predicted No. >The probability that each sample is natural enemy of the green-head submerged duck;
gradient is calculated for all samples over the entire training set, and then the update parameters are back-propagated
；
；
In the method, in the process of the invention,Representing a learning rate; /(I)Representing weight vector/>A loss function gradient of (2); /(I)Representing bias term/>A loss function gradient of (2); /(I)Representing the optimized weight vector; /(I)Representing the optimized bias term;
and performing iteration continuously until the model converges, so that the model gradually learns to distinguish the foreign invasive species from the local species.
S1.43, identifying an image sample of a potential natural enemy of the foreign invasive species through a trained machine learning model;
s1.44, classifying the results output by the machine learning model in the step S1.43 by adopting a logistic regression classifier, and comparing the results with characteristic labels of natural enemies of the green-head ducks marked in a database to judge whether the species is an external invasion natural enemies;
Because the external invasion natural enemies can have serious influence on the local ecological system, the invasion of the green-head diving ducks can directly threaten the survival and reproduction of the green-head diving ducks, the external invasion natural enemies are identified and early warned, measures can be taken in time to prevent the external invasion natural enemies from being spread, and the living environment of the natural species such as the green-head diving ducks is protected.
As a further improvement of the present technical solution, in S1.43, the probability that the sample is classified as a foreign invasive natural enemy species is:
；
In the method, in the process of the invention, A probability expressed as a foreign invasive natural enemy species; /(I)A weight vector representing a foreign invasive natural enemy species; /(I)A weight vector representing other waterfowl species; /(I)The representation is/>Vector and feature vector/>For weighted summation of features; /(I)A bias term representing a foreign invasive natural enemy species.
As a further improvement of the present technical solution, in S1.44, the specific method for determining whether the species is a foreign invasion natural enemy is as follows:
Setting classification threshold Probability/>, of foreign invasive natural enemy species obtainedAnd classification threshold/>And (3) comparing to obtain:
；
If it is Judging that the natural enemy species is alien;
If it is Judging that the natural enemy species is not exotic;
When if you want At this time, will/>Corresponding feature vector/>And (3) determining whether the species is a foreign invasion natural enemy or not based on feature matching with the feature tag of the green-head latent duck natural enemy marked in the database.
As a further improvement of the technical scheme, the wireless transmission unit comprises a far-end network transmitting base station and a near-end receiving base station, and is used for realizing point-to-point non-line-of-sight high-bandwidth wireless transmission.
Compared with the prior art, the invention has the beneficial effects that:
1. The method is applied to a floating island type monitoring system of the field ecological wetland, adopts a multi-mode fusion algorithm, brings the wetland environment parameters into feature vectors, and can utilize richer information sources to characterize species, so that the accuracy and the robustness of identifying the foreign invasive species are improved, and the multi-mode fusion is beneficial to reducing misjudgment caused by the influence of environmental conditions.
2. The floating island type monitoring system is applied to a floating island type monitoring system of a field ecological wetland, the image processing technology and the machine learning algorithm are adopted to conduct pretreatment and feature extraction on the image, the type of waterfowl in a monitoring sample is rapidly identified, whether the waterfowl shot at present is a green-head submerged duck is judged, the distribution of the waterfowl type and the variation trend of the waterfowl are monitored in real time, the quality and the ecological restoration effect of the habitat can be evaluated, and an effective wetland management and protection strategy is formulated.
Drawings
Fig. 1 is an overall flow diagram of the present invention.
The meaning of each reference sign in the figure is:
1. a floating island monitoring unit; 2. a storage unit; 3. a wireless transmission unit; 4. and a power supply unit.
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.
Examples
Referring to fig. 1, a floating island type monitoring system applied to a field ecological wetland is provided, and comprises a floating island monitoring unit 1, wherein the floating island monitoring unit 1 is used for monitoring and recording the ecological environment of the wetland and the ecological behaviors of a green-head submerged duck, and the floating island monitoring unit 1 comprises a multi-path monitoring module and a video monitoring module;
the multi-path monitoring module is used for collecting and recording wetland environment parameters, wherein the wetland environment parameters comprise water temperature, pH value, dissolved oxygen content, turbidity and illumination intensity;
the multipath monitoring module comprises a water quality sensor and an environment sensor;
The water quality sensor is used for monitoring and recording the temperature, the pH value, the dissolved oxygen content and the turbidity of the wetland water body;
The temperature of the wetland water body has direct influence on the life activities, propagation behaviors and microbial activities of aquatic organisms, and is helpful for researching the influence of the health condition of a water body ecological system and climate change on the wetland ecology;
the pH value reflects the acid-base degree of the water body, is one of important indexes for measuring the quality of water, and the too high or too low pH value can threaten the life of aquatic organisms, particularly green-head submerged ducks;
The content of the dissolved oxygen can reflect the self-cleaning capacity of the water body and the health state of an ecological system, and for waterfowl such as green-head submerged ducks, the sufficient dissolved oxygen is an important condition for survival and reproduction;
the turbidity can indirectly reflect the quantity of suspended matters in the water body, and the high turbidity can influence the survival and the illumination penetration depth of aquatic organisms, so that the growth of wetland vegetation and the balance of the whole ecological system are influenced;
The environmental sensor is used for monitoring the environmental temperature, illumination intensity and biomass density in the wetland environment. The environment temperature and the illumination intensity are used for evaluating the ecological environment comfort level of the wetland and the effective utilization condition of illumination resources;
the illumination intensity can directly influence the quality of the image captured by the video monitoring module, so that the accuracy of species identification is influenced; environmental temperature changes can affect animal behavior patterns and visual characteristics, such as feather status;
The environmental temperature can influence behavior habits of waterfowl, such as activity time and foraging behavior, and behavior habit changes can influence the type and quality of the image sample captured by the species identification module, so that the identification result can be influenced;
The environmental sensor data can be fed back to the species identification module, so that the species identification module can dynamically adjust an identification algorithm according to real-time environmental conditions, such as enhancing image enhancement processing when the illumination condition is poor, or preferentially considering certain characteristic variables under specific environmental conditions, and the identification efficiency and accuracy are improved;
biomass density was used to evaluate wetland plant productivity and richness of waterfowl food resources.
The video monitoring module is used for recording the behavior activities of the green-head submerged ducks, carrying out species identification and analysis, providing foreign species invasion early warning, such as nesting, hatching, brooding and foraging behaviors, and providing visual and detailed data for biologists to study the life habits, the population behaviors and the social structures of the green-head submerged ducks, and comprises a species identification module, a species classification module and an abnormality early warning module;
The species identification module is used for monitoring aquatic bird species in the monitoring area in real time, and based on an image processing and machine learning fusion algorithm, the species identification module identifies and distinguishes the green-head submerged ducks from other aquatic bird species, the species classification module further identifies foreign invasive species in other aquatic bird species, and the anomaly early warning module timely gives an alarm.
The species identification module is used for realizing real-time monitoring of a monitoring area through the monitoring camera.
The specific steps of the species identification module for identifying and distinguishing the green-head submerged ducks from the foreign invasive species are as follows:
s1.1, preprocessing image data of wet waterfowl captured by a species identification module; noise reduction, enhancement, scaling and cutting operations are carried out on the original image, so that the image quality is ensured to meet the subsequent analysis requirements;
S1.2, extracting contour features, texture features and key point information in image data by adopting an image processing technology, wherein the key point information comprises beak, head shape and body type feature information;
in this embodiment, the specific steps involved in the image processing technique are:
S1.21, acquiring contour information in image data through a Canny algorithm; the profile features can reflect unique profile differences for different species;
s1.22, extracting feather texture features in an image by adopting a gray level co-occurrence matrix algorithm;
s1.23, extracting the characteristic information of the shape and the body type of the beak part and the head part by using a SIFT algorithm; by calculating geometric characteristics such as the ratio of the head to the body, the length of the neck, the width and thickness of the chest and the like and combining the color distribution of the head, such as the obvious dark green head of the green head ducks, the accurate distinction of species types is facilitated;
S1.3, training a machine learning model based on a multi-mode fusion algorithm by using waterfowl species information stored in the storage unit 2, wherein the waterfowl species information comprises contour features, texture features, key point information, a foreign species database and feature labels corresponding to different species, and identifying and distinguishing green-head ducks from other waterfowl types by the trained machine learning model according to real-time image data; distinguishing the green-head submerged ducks from other species through the characteristics of body shape, feather color distribution, head color, head shape, beak characteristics and the like;
in this embodiment, the machine learning model is specifically:
Known feature vectors And incorporating the wetland environment parameter E into the feature vector/>In forming a multi-modal feature vector/>Multimodal feature vector/>Wherein the feature vector X comprises a contour feature C, a texture feature D, a beak feature B, a head shape feature H and a body shape feature S, and the multi-modal feature vector/>Mapping to a corresponding species tag Y, wherein the green-head duckling is/>Other waterfowl species are/>；
Aligning the feature vectors extracted from different modes, unifying the feature vectors by means of dimension reduction and dimension increase if the feature dimensions among the modes are inconsistent, then executing feature fusion operation, and selecting a weighted fusion method to fuse the features of the different modes; after the fusion operation, a multi-mode feature vector containing multi-mode information is obtainedThe vector not only contains various local features and global features of original data, but also reflects the relevance among different mode data, so that the vector has stronger expression capability and robustness, and the advantage of incorporating the wetland environment parameters into the feature vector is that the interaction of species characteristics and environment background can be reflected more comprehensively, so that the accuracy and stability of the species identification system are improved;
The living habit and appearance characteristic expression of birds can be influenced by the wetland environment parameters, and the environment factors are brought into the characteristic vectors, so that the actual situation can be better simulated, and the accuracy of the model for identifying the species under different environment conditions is improved; meanwhile, the introduction of the environmental parameters is helpful for understanding the interaction relationship between the species and the environment, so that the model can adjust the identification strategy according to the environmental parameters when facing untrained environmental conditions, the generalization performance of the model is enhanced, and the addition of the environmental parameters is helpful for reducing the false identification caused by environmental influence;
The probability of classifying the sample into a green head duckling is as follows:
；
The probability of a sample being classified as another waterfowl species is:
；
In the method, in the process of the invention, The probability expressed as green head ducks; /(I)Probability expressed as other waterfowl species; A weight vector representing the green head duckling; /(I) A weight vector representing other waterfowl species; /(I)The representation is/>Vector and feature vector/>For weighted summation of features; /(I)The representation is/>Vector and feature vector/>Is an inner product of (2); /(I)Representing a sigmoid function for mapping the linear prediction value into a (0, 1) interval, thereby representing a probability; /(I)A transpose operation representing a vector; /(I)A bias item representing a green-head latent duck; /(I)Bias terms representing other waterfowl species.
Real-time calculation of probability of classifying monitoring samples into green-head ducksThe method can rapidly identify the types of waterfowl in the monitoring sample, judge whether the waterfowl shot at present is a green-head diving duck, monitor the distribution of the waterfowl types and the variation trend thereof in real time, and can help to evaluate the quality and ecological restoration effect of habitat and formulate effective wetland management and protection strategies;
the sigmoid function is specifically:
；
In the method, in the process of the invention, Representing input values/>；/>Representing the result of the sigmoid function; /(I)A base representing natural logarithms, equal to 2.71828;
The meaning of the sigmoid function in the species identification module is to process and interpret the multi-modal feature vector Z and map it to a probability value between 0 and 1, thereby quantifying the probability that the sample is classified into a specific class;
s1.4, identifying foreign invasion natural enemy species from other waterfowl species through a species classification module;
By combining data of different modes, the multi-mode fusion algorithm can utilize richer information sources to characterize species, so that the accuracy and robustness of identifying foreign invasive species are improved, and the multi-mode fusion is helpful for reducing misjudgment caused by the influence of environmental conditions or individual differences;
in this embodiment, the specific steps of the species classification module for identifying the foreign invasive natural enemy species are:
S1.41, constructing a natural enemy data set of the green-head duck according to the foreign species database and the characteristic labels stored in the storage unit 2, and extracting the characteristic vector of the natural enemy of the green-head duck based on the image processing technology in the step S1.2 And wetland environmental parameters/>In forming a multi-modal feature vector/>Multimodal feature vector/>Wherein, feature vector/>Including profile features/>Texture features/>Beak characterization/>Head shape feature/>And body shape characteristics/>；
S1.42, taking a data set of natural enemies of the green-head ducks as a training set, training and optimizing weight parameters of the machine learning model in the step S1.3;
the specific process for optimizing the machine learning model is as follows:
Optimizing bias terms by minimizing loss function L And weight vector/>The minimization loss function L is then:
；
In the method, in the process of the invention, Representing the number of samples; /(I)Represents the/>True labels of the individual samples; /(I)Represents model predicted No. >The probability that each sample is natural enemy of the green-head submerged duck;
gradient is calculated for all samples over the entire training set, and then the update parameters are back-propagated
；
；
In the method, in the process of the invention,Representing a learning rate; /(I)Representing weight vector/>A loss function gradient of (2); /(I)Representing bias term/>A loss function gradient of (2); /(I)Representing the optimized weight vector; /(I)Representing the optimized bias term;
the method comprises the steps of performing iteration continuously until the model converges, so that the model gradually learns to distinguish the foreign invasive species from the local species;
s1.43, identifying an image sample of a potential natural enemy of the foreign invasive species through a trained machine learning model;
wherein the probability of the sample being classified as a foreign invasive natural enemy species is:
；
In the method, in the process of the invention, A probability expressed as a foreign invasive natural enemy species; /(I)A weight vector representing a foreign invasive natural enemy species; /(I)A weight vector representing other waterfowl species; /(I)The representation is/>Vector and feature vector/>For weighted summation of features; /(I)A bias term representing a foreign invasive natural enemy species;
s1.44, classifying the results output by the machine learning model in the step S1.43 by adopting a logistic regression classifier, and comparing the results with characteristic labels of natural enemies of the green-head ducks marked in a database to judge whether the species is an external invasion natural enemies;
Because the external invasion natural enemies can have serious influence on the local ecological system, the invasion of the green-head diving ducks can directly threaten the survival and reproduction of the green-head diving ducks, the external invasion natural enemies are identified and early warned, measures can be taken in time to prevent the external invasion natural enemies from being spread, and the living environment of the natural species such as the green-head diving ducks is protected.
Specifically, the specific method for judging whether the species is a foreign invasion natural enemy comprises the following steps:
Setting classification threshold Probability/>, of foreign invasive natural enemy species obtainedAnd classification threshold/>And (3) comparing to obtain:
；
In the method, in the process of the invention, The value of (2) is set based on expert experience;
If it is Judging that the natural enemy species is alien;
If it is Judging that the natural enemy species is not exotic;
When if you want At this time, will/>Corresponding feature vector/>The characteristic tag of the natural enemy of the green-head submerged duck marked in the database is used for determining whether the species is an external invasion natural enemy or not based on characteristic matching;
the feature matching is specifically to match feature vectors The normalization processing is carried out on each feature in the model, and the processed feature vector/>Similarity calculation is carried out on the characteristics of each green-head latent duck natural enemy sample marked in the database through cosine similarity, and a feature vector/>, based on a maximum similarity matching method, is found outThe green-head latent duck natural enemy sample which is the most similar to the green-head latent duck natural enemy sample in the database is finally judged and confirmed according to the similarity score;
S1.5, after the species classification module identifies the foreign invasive natural enemy species, the abnormal early warning module timely gives an alarm, sends early warning information to a monitoring center, and records identification time, place and species image data.
The floating island type monitoring system applied to the field ecological wetland further comprises a storage unit 2, a wireless transmission unit 3 and a power supply unit 4;
The storage unit 2 is used for storing monitoring data and aquatic bird species information, wherein the storage unit 2 comprises a local storage module and a cloud storage module;
specifically, the local storage module is used for storing and temporarily caching the monitoring data in real time, and has the capability of fast writing and reading, so that all monitoring data collected by the floating island monitoring unit 1 can be stored in real time when no remote network connection exists;
The cloud storage module is used for storing the complete monitoring data set, the waterfowl species information base, the foreign species database, the characteristic tag and the history data of the long-term monitoring record which are uploaded after preliminary processing by the local storage module;
The wireless transmission unit 3 is used for carrying out wireless connection on the floating island monitoring units 1 distributed at different positions, constructing a wireless local area network between the monitoring points, and reading monitoring data by a user accessing the cloud storage module in the local area network; the cloud storage module is mainly responsible for long-term and large-capacity storage requirements and supports remote access and data backup;
the wireless transmission unit 3 comprises a far-end network transmitting base station and a near-end receiving base station, and is used for realizing point-to-point non-line-of-sight high-bandwidth wireless transmission.
The remote floating island monitoring unit 1 is connected with the camera and the remote network transmitting base station through the network cable, signals are transmitted to the near-end receiving base station through the antenna, the near-end receiving base station is connected to the switch through the network cable and is converged with the other camera signal, the local area network is arranged, a client looks over monitoring data in the local area network through the wired connection switch, the influence of the floating platform shaking along with waves on the water surface can be avoided, and the problem that intermittent video pictures occur due to unstable transmission signals of the traditional remote floating island monitoring unit 1 is avoided.
The power supply unit 4 is used for supplying power to the floating island monitoring unit 1 and the wireless transmission unit 3; specifically, the power supply unit 4 not only supplies power to the equipment on the floating island monitoring unit 1, but also supplies power to the far-end network transmitting base station and the near-end receiving base station, and the power supply unit 4 at least comprises a solar panel, a solar controller and a storage battery; the solar panel is used for converting solar radiation capacity into electric energy to be stored in the storage battery, and the solar controller is a charge-discharge controller and is used for prolonging the service life of the storage battery and preventing the storage battery from being overcharged and deeply charged.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the above-described embodiments, and that the above-described embodiments and descriptions are only preferred embodiments of the present invention, and are not intended to limit the invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (10)

1. Be applied to floating island type monitoring system of open-air ecological wetland, its characterized in that: comprising
The floating island monitoring unit (1), the floating island monitoring unit (1) is used for monitoring and recording ecological behaviors of wetland ecological environment and species, and the floating island monitoring unit (1) comprises a multipath monitoring module and a video monitoring module;
The multi-path monitoring module is used for collecting and recording the parameters of the wetland environment;
the video monitoring module is used for recording the behavior activity of species, identifying and analyzing the species and providing foreign species invasion early warning, and comprises a species identification module, a species classification module and an anomaly early warning module;
The storage unit (2) is used for storing monitoring data and aquatic bird species information, wherein the storage unit (2) comprises a local storage module and a cloud storage module;
the wireless transmission unit (3), the wireless transmission unit (3) is used for carrying out wireless connection on floating island monitoring units (1) distributed at different positions, a wireless local area network is constructed between monitoring points, and a user accesses a cloud storage module in the local area network to read monitoring data;
the power supply unit (4), power supply unit (4) are used for supplying power for chinampa monitoring unit (1) and wireless transmission unit (3).
2. The floating island type monitoring system applied to field ecological wetland according to claim 1, wherein: the multi-path monitoring module comprises a water quality sensor and an environment sensor;
The water quality sensor is used for monitoring and recording the temperature, the pH value, the dissolved oxygen content and the turbidity of the wetland water body;
the environmental sensor is used for monitoring the environmental temperature, illumination intensity and biomass density in the wetland environment.
3. The floating island type monitoring system applied to field ecological wetland according to claim 1, wherein: the species identification module is used for monitoring aquatic bird species in a monitoring area in real time, the species identification module identifies and distinguishes species from other aquatic bird species based on an image processing and machine learning fusion algorithm, the species classification module further identifies foreign invasive species in other aquatic bird species, and the anomaly early warning module timely gives an alarm.
4. The floating island type monitoring system applied to the field ecological wetland according to claim 3, wherein: the specific steps of the species identification module for identifying and distinguishing the species from the foreign invasive species are as follows:
S1.1, preprocessing image data of wet waterfowl captured by a species identification module;
S1.2, extracting contour features, texture features and key point information in image data by adopting an image processing technology, wherein the key point information comprises beak, head shape and body type feature information;
s1.3, training a machine learning model based on a multi-mode fusion algorithm by using the waterfowl species information stored in the storage unit (2), and identifying and distinguishing species from other waterfowl species by the trained machine learning model according to real-time image data;
s1.4, identifying foreign invasion natural enemy species from other waterfowl species through a species classification module;
s1.5, after the species classification module identifies the foreign invasive natural enemy species, the abnormal early warning module timely gives an alarm and sends early warning information to the monitoring center.
5. The floating island type monitoring system applied to the field ecological wetland according to claim 4, wherein: in the step S1.2, the specific steps involved in the image processing technology are as follows:
s1.21, acquiring contour information in image data through a Canny algorithm;
s1.22, extracting feather texture features in an image by adopting a gray level co-occurrence matrix algorithm;
S1.23, extracting the beak, head shape and body type characteristic information by using a SIFT algorithm.
6. The floating island type monitoring system applied to the field ecological wetland according to claim 4, wherein: in S1.3, the machine learning model specifically includes:
Known feature vectors And incorporating the wetland environment parameter E into the feature vector/>In forming a multi-modal feature vector/>Multimodal feature vector/>Wherein the feature vector X comprises a contour feature C, a texture feature D, a beak feature B, a head shape feature H and a body shape feature S, and the multi-modal feature vector/>Mapping to a corresponding species tag Y, where the species is/>Other waterfowl species are/>；
The probability of a sample being classified as a species is:
；
The probability of a sample being classified as another waterfowl species is:
；
In the method, in the process of the invention, Probability expressed as species; /(I)Probability expressed as other waterfowl species; /(I)A weight vector representing a species; /(I)A weight vector representing other waterfowl species; /(I)The representation is/>Vector and feature vector/>Is an inner product of (2); /(I)The representation is/>Vector and feature vector/>Is an inner product of (2); /(I)Representing a sigmoid function; /(I)A transpose operation representing a vector; /(I)Bias terms representing species; /(I)Bias terms representing other waterfowl species.
7. The floating island type monitoring system applied to the field ecological wetland according to claim 6, wherein: in the step S1.4, the specific steps of identifying the foreign invasion natural enemy species by the species classification module are as follows:
S1.41, constructing a species natural enemy data set according to the foreign species database and the feature tag stored in the storage unit (2), and extracting the feature vector of the species natural enemy based on the image processing technology in the step S1.2 And wetland environmental parameters/>In forming a multi-modal feature vector/>Multimodal feature vector/>Wherein, feature vector/>Including profile features/>Texture features/>Beak characterization/>Head shape feature/>And body shape characteristics/>；
S1.42, taking a data set of natural enemies of the species as a training set, training and optimizing weight parameters of the machine learning model in the step S1.3;
s1.43, identifying an image sample of a potential natural enemy of the foreign invasive species through a trained machine learning model;
S1.44, classifying the results output by the machine learning model in the step S1.43 by adopting a logistic regression classifier, and comparing the results with characteristic labels of natural enemies of the species marked in a database to judge whether the species is an external invasion natural enemies.
8. The floating island type monitoring system applied to the field ecological wetland according to claim 7, wherein: in S1.43, the probability of the sample being classified as a foreign invasive natural enemy species is:
；
In the method, in the process of the invention, A probability expressed as a foreign invasive natural enemy species; /(I)A weight vector representing a foreign invasive natural enemy species; /(I)A weight vector representing other waterfowl species; /(I)The representation is/>Vector and feature vector/>Is an inner product of (2); /(I)A bias term representing a foreign invasive natural enemy species.
9. The floating island type monitoring system applied to field ecological wetland according to claim 8, wherein: in S1.44, the specific method for determining whether the species is a foreign invasion natural enemy is as follows:
Setting classification threshold Probability/>, of foreign invasive natural enemy species obtainedAnd classification threshold/>And (3) comparing to obtain:
；
If it is Judging that the natural enemy species is alien;
If it is Judging that the natural enemy species is not exotic;
When if you want At this time, will/>Corresponding feature vector/>And determining whether the species is a foreign invasion natural enemy based on the feature matching with the feature tag of the marked species natural enemy in the database.
10. The floating island type monitoring system applied to field ecological wetland according to claim 1, wherein: the wireless transmission unit (3) comprises a far-end network transmitting base station and a near-end receiving base station, and is used for realizing point-to-point non-line-of-sight high-bandwidth wireless transmission.
CN202410417404.4A 2024-04-09 2024-04-09 Floating island type monitoring system applied to field ecological wetland Active CN118015551B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202410417404.4A CN118015551B (en) 2024-04-09 2024-04-09 Floating island type monitoring system applied to field ecological wetland

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202410417404.4A CN118015551B (en) 2024-04-09 2024-04-09 Floating island type monitoring system applied to field ecological wetland

Publications (2)

Publication Number Publication Date
CN118015551A true CN118015551A (en) 2024-05-10
CN118015551B CN118015551B (en) 2024-06-25

Family

ID=90944726

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202410417404.4A Active CN118015551B (en) 2024-04-09 2024-04-09 Floating island type monitoring system applied to field ecological wetland

Country Status (1)

Country Link
CN (1) CN118015551B (en)

Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102065272A (en) * 2009-11-13 2011-05-18 中国科学院遥感应用研究所 Acoustic image integrated wild bird recognition technology-based remote wireless monitoring system
CN102103713A (en) * 2009-12-18 2011-06-22 湖南城市学院 Method and system for monitoring wetland resource and ecological environment
CN105676747A (en) * 2016-03-21 2016-06-15 杭州电子科技大学 Artificial floating-island intelligent monitoring and controlling system and method
KR20190033892A (en) * 2017-09-22 2019-04-01 국립생태원 System and method for monitoring wild animals
CN111291940A (en) * 2020-03-02 2020-06-16 桂林电子科技大学 Student class dropping prediction method based on Attention deep learning model
WO2021022752A1 (en) * 2019-08-07 2021-02-11 深圳先进技术研究院 Multimodal three-dimensional medical image fusion method and system, and electronic device
US20210292805A1 (en) * 2020-03-18 2021-09-23 International Business Machines Corporation Semi-supervised classification of microorganism
US20220027786A1 (en) * 2020-07-24 2022-01-27 Macau University Of Science And Technology Multimodal Self-Paced Learning with a Soft Weighting Scheme for Robust Classification of Multiomics Data
CN114387499A (en) * 2022-01-19 2022-04-22 国家海洋环境监测中心 Island coastal wetland waterfowl identification method, distribution query system and medium
CN114579647A (en) * 2022-03-03 2022-06-03 盐城工业职业技术学院 Fusion model for ecological monitoring data of multi-source heterogeneous wetland
CN115965295A (en) * 2023-03-16 2023-04-14 百鸟数据科技(北京)有限责任公司 Wetland ecosystem monitoring method, computer equipment and storage medium
CN116189884A (en) * 2023-04-24 2023-05-30 成都中医药大学 Multi-mode fusion traditional Chinese medicine physique judging method and system based on facial vision
CN116229518A (en) * 2023-03-17 2023-06-06 百鸟数据科技(北京)有限责任公司 Bird species observation method and system based on machine learning
CN116304600A (en) * 2023-03-06 2023-06-23 四川省林业科学研究院 Foreign invasive species early warning method and system based on big data analysis
CN116778347A (en) * 2023-08-24 2023-09-19 腾讯科技（深圳）有限公司 Data updating method, device, electronic equipment and storage medium
CN220654138U (en) * 2023-07-11 2024-03-26 南京林业大学 Ecological floating island with birds intelligent monitoring and identifying functions

Patent Citations (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102065272A (en) * 2009-11-13 2011-05-18 中国科学院遥感应用研究所 Acoustic image integrated wild bird recognition technology-based remote wireless monitoring system
CN102103713A (en) * 2009-12-18 2011-06-22 湖南城市学院 Method and system for monitoring wetland resource and ecological environment
CN105676747A (en) * 2016-03-21 2016-06-15 杭州电子科技大学 Artificial floating-island intelligent monitoring and controlling system and method
KR20190033892A (en) * 2017-09-22 2019-04-01 국립생태원 System and method for monitoring wild animals
WO2021022752A1 (en) * 2019-08-07 2021-02-11 深圳先进技术研究院 Multimodal three-dimensional medical image fusion method and system, and electronic device
CN111291940A (en) * 2020-03-02 2020-06-16 桂林电子科技大学 Student class dropping prediction method based on Attention deep learning model
US20210292805A1 (en) * 2020-03-18 2021-09-23 International Business Machines Corporation Semi-supervised classification of microorganism
US20220027786A1 (en) * 2020-07-24 2022-01-27 Macau University Of Science And Technology Multimodal Self-Paced Learning with a Soft Weighting Scheme for Robust Classification of Multiomics Data
CN114387499A (en) * 2022-01-19 2022-04-22 国家海洋环境监测中心 Island coastal wetland waterfowl identification method, distribution query system and medium
CN114579647A (en) * 2022-03-03 2022-06-03 盐城工业职业技术学院 Fusion model for ecological monitoring data of multi-source heterogeneous wetland
CN116304600A (en) * 2023-03-06 2023-06-23 四川省林业科学研究院 Foreign invasive species early warning method and system based on big data analysis
CN115965295A (en) * 2023-03-16 2023-04-14 百鸟数据科技(北京)有限责任公司 Wetland ecosystem monitoring method, computer equipment and storage medium
CN116229518A (en) * 2023-03-17 2023-06-06 百鸟数据科技(北京)有限责任公司 Bird species observation method and system based on machine learning
CN116189884A (en) * 2023-04-24 2023-05-30 成都中医药大学 Multi-mode fusion traditional Chinese medicine physique judging method and system based on facial vision
CN220654138U (en) * 2023-07-11 2024-03-26 南京林业大学 Ecological floating island with birds intelligent monitoring and identifying functions
CN116778347A (en) * 2023-08-24 2023-09-19 腾讯科技（深圳）有限公司 Data updating method, device, electronic equipment and storage medium

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
X. ZHU; S. MENA; CISCO SYSTEMS; Z. SARKER; ERICSSON AB;: "Modeling Video Traffic Sources for RMCAT Evaluations draft-ietf-rmcat-video-traffic-model-00", IETF, 15 January 2016 (2016-01-15) *
孙玉芳;姜丽华;李刚;陈宝瑞;张宏斌;: "外来植物入侵遥感监测预警研究进展", 中国农业资源与区划, no. 08, 25 August 2016 (2016-08-25) *

Also Published As

Publication number Publication date
CN118015551B (en) 2024-06-25

Similar Documents

Publication Publication Date Title
CN109145032B (en) Intelligent monitoring method and system for bee breeding
CN109591982A (en) A kind of no worker monitor underwater culture robot
CN107622236B (en) Crop disease diagnosis and early warning method based on swarm and gradient lifting decision tree algorithm
CN109472883A (en) Patrol pool method and apparatus
CN115067243B (en) Fishery monitoring and analyzing method, system and storage medium based on Internet of things technology
Stehfest et al. Network analysis of acoustic tracking data reveals the structure and stability of fish aggregations in the ocean
CN115879771B (en) Intelligent control method and system for dissolved oxygen in aquaculture pond
CN116034904B (en) Pig health monitoring system and method based on track type inspection robot
TW202002772A (en) Intelligent deep learning agricultural and fishery cultivation system to make processing operations, such as modification of cultivation environmental status corresponding response for growth cultivation, according to the organism growth and environment statuses
CN114898405B (en) Portable broiler chicken anomaly monitoring system based on edge calculation
Zhang et al. From beasts to bytes: Revolutionizing zoological research with artificial intelligence
CN113887413A (en) Sika behavior recognition and health early warning method and system
CN118015551B (en) Floating island type monitoring system applied to field ecological wetland
CN116935439A (en) Automatic monitoring and early warning method and automatic monitoring and early warning system for delivery of pregnant sheep
CN109359624B (en) Neural network-based prediction method and prediction device for feeding behavior of laying hens
CN111950598A (en) Method for individually classifying swimming crab groups based on K-Means algorithm and application
KR102678374B1 (en) Digital twin-based smart farm method, management device and computer program
CN112380486B (en) Dynamic monitoring and evaluating method and system for sea water ornamental fish culture state
Saville et al. A Mariculture Fish Mortality Prediction Using Machine Learning Based Analysis of Water Quality Monitoring
Kumar et al. Estimating fish weight growth in aquaponic farming through machine learning techniques
Gladju et al. Potential applications of data mining in aquaculture
CN118355874B (en) Oyster reef proliferation repair regional control method
CN118036903B (en) Odontobutis obscura behavior data-based seedling culture management system and method
CN117611380B (en) Fish disease early warning method and system
Nuanmeesri et al. Transfer Learning Artificial Neural Network-based Ensemble Voting of Water Quality Classification for Different Types of Farming

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