CN113487143A - Fish shoal feeding decision method and device, electronic equipment and storage medium - Google Patents

Fish shoal feeding decision method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN113487143A
CN113487143A CN202110662613.1A CN202110662613A CN113487143A CN 113487143 A CN113487143 A CN 113487143A CN 202110662613 A CN202110662613 A CN 202110662613A CN 113487143 A CN113487143 A CN 113487143A
Authority
CN
China
Prior art keywords
fish
feeding
fed
feed
type
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.)
Pending
Application number
CN202110662613.1A
Other languages
Chinese (zh)
Inventor
安冬
张宇宁
位耀光
李道亮
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.)
China Agricultural University
Original Assignee
China Agricultural University
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 China Agricultural University filed Critical China Agricultural University
Priority to CN202110662613.1A priority Critical patent/CN113487143A/en
Publication of CN113487143A publication Critical patent/CN113487143A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
    • G06F18/2132—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on discrimination criteria, e.g. discriminant analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/2431—Multiple classes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/205—Parsing
    • G06F40/216—Parsing using statistical methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/284—Lexical analysis, e.g. tokenisation or collocates
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/20—Natural language analysis
    • G06F40/279—Recognition of textual entities
    • G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
    • 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
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/02—Agriculture; Fishing; Mining

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Computational Linguistics (AREA)
  • Human Resources & Organizations (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Strategic Management (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Educational Administration (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Molecular Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computing Systems (AREA)
  • Biophysics (AREA)
  • General Business, Economics & Management (AREA)
  • Biomedical Technology (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Tourism & Hospitality (AREA)
  • Evolutionary Biology (AREA)
  • Probability & Statistics with Applications (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Agronomy & Crop Science (AREA)
  • Animal Husbandry (AREA)
  • Marine Sciences & Fisheries (AREA)

Abstract

The invention provides a fish school feeding decision method, a device, electronic equipment and a storage medium, wherein the method comprises the following steps: obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules; the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; and inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of the fed feed and the feeding amount. The method, the device, the electronic equipment and the storage medium provided by the invention realize accurate feeding of the fish school, and avoid the condition that the water quality and the water body are influenced and the health of the fish school is harmful because the feeding is not accurate by manual decision.

Description

Fish shoal feeding decision method and device, electronic equipment and storage medium
Technical Field
The invention relates to the technical field of intelligent feeding, in particular to a fish school feeding decision method, a fish school feeding decision device, electronic equipment and a storage medium.
Background
With the increasing status of aquaculture industry in the fishery economic industry of China, the problems of overhigh culture density, excessive bait feeding and the like are caused, so that the bait is dissolved out and the water body is polluted, thereby influencing the growth and health of cultured animals and improving the culture cost and the culture risk.
Because the degree of satiation of aquatic organisms is not well controlled, the growth of the aquatic organisms is limited due to insufficient feeding, the residual baits are dissolved in the aquaculture water body due to excessive feeding, and the baits are used as main sources of nitrogen and phosphorus pollutants and become solid or suspended particle pollutants to cause water body pollution.
However, since the aquatic animals must be fed through the medium, the deterioration of the water quality of the water body can seriously affect the health of the aquatic animals. Meanwhile, the discharge of the aquaculture wastewater can also aggravate the pollution to the surrounding environment (lakes, rivers, sea areas and the like), so that the reduction of the influence of aquaculture production on the environment can be realized through an accurate feeding strategy.
Therefore, how to avoid the situations of water quality pollution and aquatic animal health hazard caused by inaccurate feeding amount control in the traditional manual feeding decision is still a problem to be solved by the technical personnel in the field.
Disclosure of Invention
The invention provides a fish school feeding decision method, a device, electronic equipment and a storage medium, which are used for solving the problems of water quality pollution and aquatic animal health harms caused by inaccurate feeding amount control in the traditional manual feeding decision. Because fish nutrition feeding knowledge base is established with the default rule on a large amount of fish culture technical data of collecting, including the corresponding relation link in the knowledge base, specific type's fish should throw the feeding volume and the fodder model of specific type fodder when specific breed density and specific growth stage promptly, consequently, go based on the breed density and the biomass index of the fish crowd of waiting to throw the feed inquiry can obtain accurate feeding volume and fodder model in the knowledge base, is favorable to accurate throwing the feed, avoids influencing the quality of water and avoids harmful fish crowd healthy.
The invention provides a fish school feeding decision method, which comprises the following steps:
obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules;
the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type;
collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image;
and inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of the fed feed and the feeding amount.
According to the fish school feeding decision method provided by the invention, the fish nutrition feeding knowledge base is constructed according to the preset rules, and the method specifically comprises the following steps:
initially cleaning txt file metadata information of the fish nutrition feeding data to obtain data nodes;
carrying out secondary cleaning on the data nodes based on the LDA model to obtain the probability of any vocabulary under any theme in any document;
selecting the vocabulary with the highest occurrence probability under any topic as the elements in the relation chain;
wherein the subject is the type of fish, feed type and feed model selected manually.
According to the fish school feeding decision method provided by the invention, the data nodes are cleaned for the second time based on the LDA model to obtain the probability of any vocabulary under any theme in any document, and the method specifically comprises the following steps:
determining a corpus W, a vocabulary V and a theme vector T of the data nodes;
wherein W ═ { W ═ W1，w2，...，wn}，V＝{v1，v2，...，vm}，T＝{t1，t2，...，tkN is the total number of documents, m is the total number of words, and k is the total number of topics;
for any topic in each document, any vocabulary v is determinedjIn any document wiFrequency d of occurrence inijConstructing a frequency second order matrix Dn×m；
The frequency second order matrix Dn×mInputting LDA model, outputting document-theme second-order matrix Mn×kAnd topic-vocabulary second order matrix Qk×m；
Based on the document-subject second-order matrix Mn×kAnd the topic-vocabulary second order matrix Qk×mDetermining a document-vocabulary second order matrix S for each topicn×m。
According to the fish school feeding decision method provided by the invention, the obtaining of fish nutrition feeding data specifically comprises the following steps:
determining fish nutrition feeding data based on artificially collected feed factory feeding manuals and fish culture technical information obtained through network crawling;
the network-crawled fish culture technology information comprises webpage information of a crawled culture technology education website and fish culture thesis information in a literature website.
According to the fish school feeding decision method provided by the invention, the determining of the biomass information of the fish school to be fed based on the underwater image specifically comprises the following steps:
determining image coordinate system coordinates of fish in the fish school to be fed, which are collected in the underwater image;
converting the coordinates of the image coordinate system into coordinates of a world coordinate system based on a camera coordinate system conversion rule;
determining a length and a width of the fish based on the world coordinate system coordinates, determining a surface area and a fish volume mass of the fish based on the length and the width;
wherein the biomass information comprises length, width, surface area and fish volume mass of the fish.
According to the fish school feeding decision method provided by the invention, the determining of the breeding density of the fish school to be fed based on the underwater image specifically comprises the following steps:
inputting the underwater image into a density map extraction model, and outputting a density map of the fish school to be fed;
the density map extraction model is obtained by training based on a sample underwater image and a corresponding density map label, and the neural network construction in the density map extraction model training process comprises the steps of generating a low-resolution density map LR-CNN network and generating a high-resolution density map HR-CNN network;
determining the culture density of the fish school to be fed based on the density map.
According to the fish school feeding decision method provided by the invention, the fish nutrition feeding knowledge base is inquired based on the biomass information and the culture density to determine the type of fed feed and the feeding amount, and the method specifically comprises the following steps:
introducing target biomass information of a fish swarm to be fed, target breeding density, artificially determined target fish type and target feed type into the fish nutrition feeding knowledge base for searching, and outputting a target feeding feed model and a target feeding amount corresponding to a search result target relation chain;
wherein the target relation chain comprises a corresponding relation chain of the type of the target fish, the type of the target feed, the target breeding density, the target biomass index, the target feeding amount and the target feed model.
The invention also provides a fish school feeding decision device, comprising:
the building unit is used for obtaining fish nutrition feeding data and building a fish nutrition feeding knowledge base according to preset rules;
the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type;
the collecting unit is used for collecting an underwater image of a fish school to be fed and determining biomass information and culture density of the fish school to be fed based on the underwater image;
and the decision unit is used for inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type and the feeding amount of the fed feed.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the steps of the fish school feeding decision method.
The invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of a method of fish herd feeding decision as described in any one of the above.
According to the fish school feeding decision method, the fish school feeding decision device, the electronic equipment and the storage medium, the fish nutrition feeding knowledge base is constructed according to the preset rules by acquiring the fish nutrition feeding data; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density, and determining the type of fed feed and the feeding amount; the preset rule is analysis processing of fish nutrition feeding data based on a linear discriminant analysis LDA model, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type. The fish nutrition feeding knowledge base is constructed according to preset rules on the collected large amount of fish culture technical data, the knowledge base comprises corresponding relation links, namely the feeding amount and the feed type of specific type of feed should be fed when specific culture density and specific growth stage of specific type of fish, so that accurate feeding amount and feed type can be obtained by inquiring the knowledge base based on the determined culture density and biomass index of the fish swarm to be fed, accurate feeding is facilitated, water quality and water body are prevented from being influenced, and the health of the fish swarm is prevented from being harmful. Therefore, the method, the device, the electronic equipment and the storage medium provided by the embodiment of the invention realize accurate feeding of the fish school and avoid the condition that the water quality and the water body are influenced and the health of the fish school is harmful because the feeding is not accurately decided by manpower.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a fish school feeding decision method provided by the present invention;
FIG. 2 is a flow chart of nutrient feeding data acquisition provided by the present invention;
FIG. 3 is a flow chart of the calculation of fish growth indicators according to the present invention;
FIG. 4 is a flow chart of a process for obtaining biomass information of fish according to the present invention;
FIG. 5 is a schematic view of a fish feeding decision process according to the present invention;
fig. 6 is a schematic structural diagram of a fish school feeding decision device provided by the present invention;
fig. 7 is a schematic physical structure diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The problems of water quality pollution of water body and harm to the health of aquatic animals caused by inaccurate feeding amount control generally exist in the traditional manual feeding decision. A fish shoal feeding decision method of the present invention is described below with reference to fig. 1 to 5. Fig. 1 is a schematic flow chart of a fish school feeding decision method provided by the present invention, as shown in fig. 1, the method includes:
step 110, obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules;
the preset rule is analysis processing of fish nutrition feeding data based on a linear discriminant analysis LDA model, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type.
Specifically, fish nutrition feeding data is obtained firstly, the data is obtained in various manners, such as querying an employee operation manual from a farm, querying a feed use manual from a feed factory, or crawling relevant information of fish cultivation from the internet, a target webpage is set in a crawling process, such as an aquaculture website, a cultivation website, an aquaculture website and a literature website, wherein the literature website can crawl fish cultivation technical articles by setting relevant keywords, and a large amount of fish nutrition feeding data can be obtained by any one or any combination of the above methods. Then, based on the obtained massive fish nutrition feeding data, a fish nutrition feeding knowledge base is constructed according to preset rules, wherein the knowledge base is further defined to comprise a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed model, for example, one relation chain in the knowledge base can be crucian-plant type feed-100 per mu of pond-10 cm in length and 300 g in weight-1 kg per mu of pond-diatom essence, and the relation chain indicates that, in the pond for culturing crucian, 1kg of diatom essence feed needs to be fed per mu of pond if the current growth vigor of the crucian is 10 cm in length and 300 g in weight and the culture density is 100 per mu of pond. The knowledge base can be constructed to include data of the relationship chain, and the relationship chain can be obtained only by processing collected large amount of fish nutrition feeding data according to preset rules, which are explained herein.
And 120, acquiring an underwater image of the fish school to be fed, and determining the biomass information and the culture density of the fish school to be fed based on the underwater image.
Specifically, the method comprises the steps of collecting an underwater image of a fish school to be fed, carrying out image processing on the underwater image to identify fish in the underwater image, determining average biomass information of all the fish in the underwater image, wherein the biomass information generally comprises the length, width, area and weight of each fish or sampled fish in the image, processing the underwater image by the culture density through a specific rule to obtain a density map, and counting mark points in the density map to obtain the number of the fish school in the space range of the underwater image. The specific rule is described here, and there are many ways to determine the specific rule of the density map, for example, training based on machine learning obtains a density map extraction model, or using a key feature extraction algorithm in the image to identify fish included in the image, and the like, and this is not limited here.
And step 130, inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density, and determining the type and the feeding amount of the fed feed.
Specifically, after the biomass information and the cultivation density of the fish swarm to be fed are determined in step 120, the biomass information and the cultivation density can be used to go to the fish nutrition feeding knowledge base constructed in step 110 for query, so as to obtain the most appropriate feed feeding amount and the specific model of the feed to be fed under the conditions of the current growth stage and the cultivation density of the fish swarm to be fed. It should be noted that the type of the fish to be fed is provided by the breeder, and the type of the feed is also provided by the breeder, for example, the feed for crucian can be plant type or small fish and shrimp type, the breeder is required to determine the type of the feed, and the breeder can consider from an economic aspect or the time-consuming cost of feeding human, and the like, and is not limited herein.
According to the method provided by the invention, a fish nutrition feeding knowledge base is constructed according to preset rules by acquiring fish nutrition feeding data; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density, and determining the type of fed feed and the feeding amount; the preset rule is analysis processing of fish nutrition feeding data based on a linear discriminant analysis LDA model, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type. The fish nutrition feeding knowledge base is constructed according to preset rules on the collected large amount of fish culture technical data, the knowledge base comprises corresponding relation links, namely the feeding amount and the feed type of specific type of feed should be fed when specific culture density and specific growth stage of specific type of fish, so that accurate feeding amount and feed type can be obtained by inquiring the knowledge base based on the determined culture density and biomass index of the fish swarm to be fed, accurate feeding is facilitated, water quality and water body are prevented from being influenced, and the health of the fish swarm is prevented from being harmful. Therefore, the method provided by the embodiment of the invention realizes accurate feeding of the fish school and avoids the condition that the feeding is inaccurate by manual decision to influence the water quality and the water body and the health of the fish school.
Based on the above embodiment, in the method, the constructing a fish nutrition feeding knowledge base according to the preset rule specifically includes:
initially cleaning txt file metadata information of the fish nutrition feeding data to obtain data nodes;
carrying out secondary cleaning on the data nodes based on the LDA model to obtain the probability of any vocabulary under any theme in any document;
selecting the vocabulary with the highest occurrence probability under any topic as the elements in the relation chain;
wherein the subject is the type of fish, feed type and feed model selected manually.
Specifically, the embodiment further defines a specific way of constructing the fish nutrition feeding knowledge base according to preset rules. The specific construction method comprises two data cleaning treatments:
primary cleaning: and importing the metadata information txt file of the fish nutrition feeding data into a PilotEdit tool, automatically converting the collected metadata information into data nodes through the tool (the conversion means that all the punctuations are converted into commas so as to carry out subsequent secondary processing), and then manually checking to delete possible pictures or messy codes.
Secondary cleaning: and (3) carrying out secondary cleaning on the document data by utilizing the LDA theme model, wherein the specific process is as follows:
the probability of a word appearing under a certain theme of a certain document can be obtained by assuming that the probability of the word appearing under the same theme, the probability of the word appearing under the same theme of the same document and the product of the two probabilities are known, namely the probability of the appearance of any word under any theme in any document can be obtained, and then the word with the maximum probability of the appearance under any theme is selected as an element in the relation chain.
It should be noted here that the type, feed type and feed type of fish subject to be manually selected, for example, subjects that may include crucian, grass carp, catfish, megalobrama, perch, vegetal feed, cereal feed, animal feed, diatoms, shrimp, corn, etc., are manually set according to the functional requirements of the farming using the herd feeding decision device.
Based on the above embodiment, in the method, the performing secondary cleaning on the data node based on the LDA model to obtain the probability of any vocabulary appearing under any topic in any document specifically includes:
determining a corpus W, a vocabulary V and a theme vector T of the data nodes;
wherein W ═ { W ═ W1，w2，...，wn}，V＝{v1，v2，...，vm}，T＝{t1，t2，...，tkN is the total number of documents, m is the total number of words, and k is the total number of topics;
for any topic in each document, any vocabulary v is determinedjIn any document wiFrequency d of occurrence inijConstructing a frequency second order matrix Dn×m；
The frequency second order matrix Dn×mInputting LDA model, outputting document-theme second-order matrix Mn×kAnd topic-vocabulary second order matrix Qk×m；
Based on the document-subject second-order matrix Mn×kAnd the topic-vocabulary second order matrix Qk×mDetermining a document-vocabulary second order matrix S for each topicn×m。
Specifically, assuming that the probability of a word appearing under the same topic, the probability of a topic appearing under the same document, and the product of the two probabilities are known, the probability of a word appearing under a topic of a document can be obtained:
1): now there are n txt documents forming a corpus W ═ W1，w2，...，wnIn total, m words form a vocabulary table V ═ V1，v2，...，vmA total of k topic words forming a topic vector T ═ T1，t2，...，tk}；
2): forming a second order matrix D from W and Vn×mFor each document, any theme is extracted from the theme distribution, and a word d is randomly extracted from the word distribution corresponding to the extracted themeijD isn×mAs an input term for the LDA model, where the value d in the second order matrixijRepresents a word vjAt wiThe frequency of occurrence of;
3): repeating the process 2) until each word in the whole document is traversed;
4): after passing through the LDA model, a second-order matrix M of the document-subject is outputn×kSecond order matrix Q of sum topic-vocabularyk×mThe former is used for describing the distribution of the current document on k subjects, and the row vector is marked as thetai(i 1, 2.. times.n), the column vector represents the specific subject word, denoted as Zj(j ═ 1, 2,. k); the latter is used for describing the distribution of the theme on m vocabularies;
5): combining the matrix M generated in 4)n×kAnd Qk×mMultiplying to obtain a matrix S representing the probability of a word under a topic appearing in a documentn×mDeleting the value of 0 according to the distribution condition of the matrix;
6): and finally, manually screening once again to form nutrition feeding data which is stored in a CSV format according to the type and the data category of the fish, and constructing a fish nutrition feeding knowledge base according to a relation flow of the type of the fish, the feed type, the fish culture density, the fish biomass index, the feeding amount and the feed model by combining a neo4j database construction technology.
Based on the above embodiment, in the method, the obtaining of fish nutrition feeding data specifically includes:
determining fish nutrition feeding data based on artificially collected feed factory feeding manuals and fish culture technical information obtained through network crawling;
the network-crawled fish culture technology information comprises webpage information of a crawled culture technology education website and fish culture thesis information in a literature website.
Specifically, the nutrition feeding data can be obtained through two ways: factory collection and web crawling. Wherein, factory collection comprises an operator manual and a feeder feeding manual of a farm, which both need manual collection; the information sources crawled in the web crawl include two main categories: the first is a webpage or website related to fish culture technology and information, such as an aquaculture net, a culture net, an aquaculture net and the like. Fig. 2 is a flow chart for acquiring nutrition feeding data provided by the present invention, as shown in fig. 2, the main flow is: firstly, a web crawler is used for downloading a whole station or a specified page by using webzip; secondly, acquiring webpage content by using the jsup for online webpage content; and finally, screening out nutrition feeding data according to the fish keywords, and storing the nutrition feeding data as a txt file. And secondly, crawling the literature web documents. The main process is as follows: firstly, crawling the literature web documents by using a python crawler according to a multi-keyword mode (the information of crawled articles comprises article names, authors, article keywords, article summaries, article reference documents and article detail pages url), and storing the crawled information; and then, downloading the full text of the document according to the url link of the crawled article detail page, and storing the full text of the document as a txt file.
The nutrition feeding data comprises: the type, percentage of nutrients, feeding amount, and environmental requirements (herein referred to as pond temperature and dissolved oxygen concentration) of the feed required by the fish in different growth stages.
Based on the above embodiment, in the method, the determining biomass information of the fish school to be fed based on the underwater image specifically includes:
determining image coordinate system coordinates of fish in the fish school to be fed, which are collected in the underwater image;
converting the coordinates of the image coordinate system into coordinates of a world coordinate system based on a camera coordinate system conversion rule;
determining a length and a width of the fish based on the world coordinate system coordinates, determining a surface area and a fish volume mass of the fish based on the length and the width;
wherein the biomass information comprises length, width, surface area and fish volume mass of the fish.
Specifically, the biomass estimation is carried out on a real-time fish image, and comprises two parts: counting fishes and estimating the weight of the fishes to obtain biomass information of the fishes; fig. 3 is a flowchart of the calculation of the fish growth index provided by the present invention, and as shown in fig. 3, the specific process of the calculation is as follows:
a. calculating the fish body mass according to the surface area of the length and the width of each fish;
b. integrating the fish quality and length of each fish, the fish culture density of a feeding decision system and the growth stage information to obtain the growth index of each fish;
c. and taking the average value of the growth indexes of each fish as the fish growth index in the feeding decision system.
Further, obtaining biomass information of the fish comprises: length, width and surface area of the fish. The method adopts a parallax distance measurement method in binocular vision length estimation and combines a single-factor prediction model, selects underwater fish images, and comprises the following specific processes: two cameras with completely same internal parameters are selected and placed in parallel, so that the optical axes of the cameras are parallel to each other, another pair of coordinate axes are collinear, the two imaging planes are coplanar, and the optical centers of the two cameras have a fixed distance d, so that the world coordinates of image points are solved only by the internal parameters of the cameras.
Under the special camera configuration described above, assume C1Coordinate system is O1x1y1z1，C2Coordinate system is O2x2y2z2Focal length f, camera separation d, and the coordinates of any spatial point P at C1Is as follows (x)1，y1，z1) At C2In the coordinate system is (x)2，y2，z2) To the left of the image point in the left camera is (u)1，v1) The image point coordinate in the right camera is (u)2，v2)。
According to the camera coordinate system relation formula and the shooting proportion relation of the world coordinate system relation formula, the left camera and the right camera have the following relation:
Figure BDA0003115958590000131
X＝x1＝x2+d，Y＝y1＝y2，Z＝z1＝z2
Figure BDA0003115958590000132
from the above formula
x1-x2＝d
Figure BDA0003115958590000133
Figure BDA0003115958590000134
Further, it is possible to obtain:
Figure BDA0003115958590000135
three-dimensional coordinates of spatial points can be calculated from the above equation:
Figure BDA0003115958590000136
Figure BDA0003115958590000137
Figure BDA0003115958590000138
and calculating the length and width of the fish through the coordinate change of the space point, thereby obtaining the surface area and the fish volume mass and obtaining the biomass information of the fish.
Based on the above embodiment, in the method, the determining the culture density of the fish school to be fed based on the underwater image specifically includes:
inputting the underwater image into a density map extraction model, and outputting a density map of the fish school to be fed;
the density map extraction model is obtained by training based on a sample underwater image and a corresponding density map label, and the neural network construction in the density map extraction model training process comprises the steps of generating a low-resolution density map LR-CNN network and generating a high-resolution density map HR-CNN network;
determining the culture density of the fish school to be fed based on the density map.
Specifically, fish farming density information is acquired. The method for estimating the density map in deep learning is adopted, and the specific flow is as follows: aiming at an underwater fish real-time image, firstly, generating a density map with low resolution; further refinement then generates a high resolution density map. The network structure is composed of two CNN branches, one branch is used for generating a low-resolution density map, and the other branch is used for generating a high-resolution density map on the basis of the generated low-resolution density map and the extracted feature map.
The specific process is as follows: the input is a triplet D { (X)1，Y1，Z1)，...，(Xn，Yn，Zn) In which XiIs an input image, YiIs a density map of the same resolution as the original image, ZiIs a low resolution density map.
For the branch LR-CNN (image becomes one quarter of original image) that generates the low resolution density map, the input is XiThen, it can be expressed by the following formula:
Figure BDA0003115958590000141
wherein the content of the first and second substances,
Figure BDA0003115958590000142
is the generated low-resolution density map, flIs a mapping function that converts the picture to a low resolution density map;
for generating high resolution branch HR-CNN (a bilinear interpolation method is used to restore the 1/4 size image to the original image size), the input is Xi，
Figure BDA0003115958590000143
It is represented as follows:
Figure BDA0003115958590000144
wherein f ishIs a mapping function that converts the picture into a high resolution density map;
which minimizes the loss function L (theta)l，θh) Learning of thetal，θh：
Figure BDA0003115958590000151
Where ρ islAnd ρhAre all hyper-parameters, based on formulas
Figure BDA0003115958590000152
And formula
Figure BDA0003115958590000153
For the above formula L (θ)l，θh) The resulting loss function is simplified as follows:
Figure BDA0003115958590000154
and finally, using a high-resolution density map output by HR-CNN as a final output result, wherein the number of density points in the density map is the fish culture density.
Fig. 4 is a flow chart of a process for obtaining biomass information of fish according to the present invention, as shown in fig. 4, the left branch in fig. 4 is used to obtain fish culture density, the right branch in fig. 4 is used to obtain fish surface area and fish volume, and finally the fish culture density and fish surface area weight are used as fish biomass information to query the fish nutrition feeding knowledge base to obtain the type of feed to be fed and the feeding amount of the fish to be fed at the current time.
Based on the above embodiment, in the method, the querying the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type and the feeding amount of the fed feed specifically includes:
introducing target biomass information of a fish swarm to be fed, target breeding density, artificially determined target fish type and target feed type into the fish nutrition feeding knowledge base for searching, and outputting a target feeding feed model and a target feeding amount corresponding to a search result target relation chain;
wherein the target relation chain comprises a corresponding relation chain of the type of the target fish, the type of the target feed, the target breeding density, the target biomass index, the target feeding amount and the target feed model.
Specifically, according to the fish growth indexes and the fish satiety degree, determining various required culture index values by combining a nutrition feeding knowledge base, and finishing feeding decision;
and leading the fish growth index and the culture density as well as the type and the target feed type of the predetermined target fish into a nutrition feeding knowledge base for searching, and further obtaining two culture feeding index numerical values of the feed type and the feeding amount corresponding to the current growth stage.
Fig. 5 is a schematic view of a fish school feeding decision process provided by the present invention, and as shown in fig. 5, the process specifically includes:
1. acquiring various nutrition feeding data and real-time fish images in an intelligent fish feeding decision system;
2. analyzing and processing the nutrition feeding data, and integrating into a fish nutrition feeding knowledge base;
3. analyzing and processing the fish image to obtain a fish growth index and a fish satiety degree;
4. and determining various required culture index values according to the fish growth indexes and the fish satiety degree by combining a nutrition feeding knowledge base, and finishing feeding decision.
The fish school feeding decision device provided by the present invention is described below, and the fish school feeding decision device described below and the fish school feeding decision method described above may be referred to in correspondence with each other.
Fig. 6 is a schematic structural diagram of a fish school feeding decision device provided by the present invention, as shown in fig. 6, the fish school feeding decision device includes a generation and construction unit 610, an acquisition unit 620 and a decision unit 630, wherein,
the constructing unit 610 is used for acquiring fish nutrition feeding data and constructing a fish nutrition feeding knowledge base according to preset rules;
the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type;
the acquisition unit 620 is configured to acquire an underwater image of a fish school to be fed, and determine biomass information and culture density of the fish school to be fed based on the underwater image;
the decision unit 630 is configured to query the fish nutrition feeding knowledge base based on the biomass information and the culture density, and determine a type of fed feed and a feeding amount.
According to the fish school feeding decision-making device provided by the invention, a fish nutrition feeding knowledge base is constructed according to preset rules by acquiring fish nutrition feeding data; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density, and determining the type of fed feed and the feeding amount; the preset rule is analysis processing of fish nutrition feeding data based on a linear discriminant analysis LDA model, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type. The fish nutrition feeding knowledge base is constructed according to preset rules on the collected large amount of fish culture technical data, the knowledge base comprises corresponding relation links, namely the feeding amount and the feed type of specific type of feed should be fed when specific culture density and specific growth stage of specific type of fish, so that accurate feeding amount and feed type can be obtained by inquiring the knowledge base based on the determined culture density and biomass index of the fish swarm to be fed, accurate feeding is facilitated, water quality and water body are prevented from being influenced, and the health of the fish swarm is prevented from being harmful. Therefore, the device provided by the embodiment of the invention realizes accurate feeding of fish schools and avoids the condition that the feeding is inaccurate through manual decision to influence the water quality and the water body and the health of the fish schools.
On the basis of the above embodiment, in the fish school feeding decision device, the constructing a fish nutrition feeding knowledge base according to the preset rule specifically includes:
initially cleaning txt file metadata information of the fish nutrition feeding data to obtain data nodes;
carrying out secondary cleaning on the data nodes based on the LDA model to obtain the probability of any vocabulary under any theme in any document;
selecting the vocabulary with the highest occurrence probability under any topic as the elements in the relation chain;
wherein the subject is the type of fish, feed type and feed model selected manually.
On the basis of the above embodiment, in the fish school feeding decision apparatus, the performing secondary cleaning on the data nodes based on the LDA model to obtain the probability of any vocabulary under any topic in any document specifically includes:
determining a corpus W, a vocabulary V and a theme vector T of the data nodes;
wherein W ═ { W ═ W1，w2，…，wn}，V＝{v1，v2，...，vm}，T＝{t1，t2，...，tkN is the total number of documents, m is the total number of words, and k is the total number of topics;
for any topic in each document, any vocabulary v is determinedjIn any document wiFrequency d of occurrence inijConstructing a frequency second order matrix Dn×m；
The frequency second order matrix Dn×mInputting LDA model, outputting document-theme second-order matrix Mn×kAnd topic-vocabulary second order matrix Qk×m；
Based on the document-subject second-order matrix Mn×kAnd the topic-vocabulary second order matrix Qk×mDetermining a document-vocabulary second order matrix S for each topicn×m。
On the basis of the above embodiment, in the fish school feeding decision device, the acquiring fish nutrition feeding data specifically includes:
determining fish nutrition feeding data based on artificially collected feed factory feeding manuals and fish culture technical information obtained through network crawling;
the network-crawled fish culture technology information comprises webpage information of a crawled culture technology education website and fish culture thesis information in a literature website.
On the basis of the above embodiment, in the fish school feeding decision apparatus, the determining biomass information of the fish school to be fed based on the underwater image specifically includes:
determining image coordinate system coordinates of fish in the fish school to be fed, which are collected in the underwater image;
converting the coordinates of the image coordinate system into coordinates of a world coordinate system based on a camera coordinate system conversion rule;
determining a length and a width of the fish based on the world coordinate system coordinates, determining a surface area and a fish volume mass of the fish based on the length and the width;
wherein the biomass information comprises length, width, surface area and fish volume mass of the fish.
On the basis of the above embodiment, in the fish school feeding decision device, the determining the culture density of the fish school to be fed based on the underwater image specifically includes:
inputting the underwater image into a density map extraction model, and outputting a density map of the fish school to be fed;
the density map extraction model is obtained by training based on a sample underwater image and a corresponding density map label, and the neural network construction in the density map extraction model training process comprises the steps of generating a low-resolution density map LR-CNN network and generating a high-resolution density map HR-CNN network;
determining the culture density of the fish school to be fed based on the density map.
On the basis of the above embodiment, in the fish school feeding decision device, the decision unit is specifically configured to:
introducing target biomass information of a fish swarm to be fed, target breeding density, artificially determined target fish type and target feed type into the fish nutrition feeding knowledge base for searching, and outputting a target feeding feed model and a target feeding amount corresponding to a search result target relation chain;
wherein the target relation chain comprises a corresponding relation chain of the type of the target fish, the type of the target feed, the target breeding density, the target biomass index, the target feeding amount and the target feed model.
Fig. 7 is a schematic physical structure diagram of an electronic device provided in the present invention, and as shown in fig. 7, the electronic device may include: a processor (processor)710, a communication Interface (Communications Interface)720, a memory (memory)730, and a communication bus 740, wherein the processor 710, the communication Interface 720, and the memory 730 communicate with each other via the communication bus 740. Processor 710 may invoke logic instructions in memory 730 to perform a shoal feeding decision method comprising: obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules; the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; and inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of the fed feed and the feeding amount.
In addition, the logic instructions in the memory 730 can be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the fish herd feeding decision method provided by the above methods, the method comprising: obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules; the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; and inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of the fed feed and the feeding amount.
In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, the computer program being implemented by a processor to perform the method for fish herd feeding decision provided by the above methods, the method comprising: obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules; the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type; collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image; and inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of the fed feed and the feeding amount.
The above-described server embodiments are only illustrative, and the units described as separate components may or may not be physically separate, and components displayed as units may or may not be physical units, may be located in one place, or may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. A fish shoal feeding decision-making method is characterized by comprising the following steps:
obtaining fish nutrition feeding data, and constructing a fish nutrition feeding knowledge base according to preset rules;
the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type;
collecting an underwater image of a fish school to be fed, and determining biomass information and culture density of the fish school to be fed based on the underwater image;
and inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of the fed feed and the feeding amount.
2. A fish school feeding decision method according to claim 1, wherein the constructing a fish nutrition feeding knowledge base according to a preset rule specifically comprises:
initially cleaning txt file metadata information of the fish nutrition feeding data to obtain data nodes;
carrying out secondary cleaning on the data nodes based on the LDA model to obtain the probability of any vocabulary under any theme in any document;
selecting the vocabulary with the highest occurrence probability under any topic as the elements in the relation chain;
wherein the subject is the type of fish, feed type and feed model selected manually.
3. A fish school feeding decision method according to claim 2, wherein the secondary cleaning of the data nodes based on the LDA model to obtain the probability of any vocabulary under any topic in any document specifically comprises:
determining a corpus W, a vocabulary V and a theme vector T of the data nodes;
wherein W ═ { W ═ W1,w2,…,wn}，V＝{v1,v2,…,vm}，T＝{t1,t2,…,tkN is the total number of documents, m is the total number of words, and k is the total number of topics;
for any topic in each document, any vocabulary v is determinedjIn any document wiFrequency d of occurrence inijConstructing a frequency second order matrix Dn×m；
The frequency second order matrix Dn×mInputting LDA model, outputting document-theme second-order matrix Mn×kAnd topic-vocabulary second order matrix Qk×m；
Based on the document-subject second-order matrix Mn×kAnd the topic-vocabulary second order matrix Qk×mDetermining a document-vocabulary second order matrix S for each topicn×m。
4. A fish farm feeding decision method according to any one of claims 1-3, wherein the obtaining fish nutritional feeding data specifically comprises:
determining fish nutrition feeding data based on artificially collected feed factory feeding manuals and fish culture technical information obtained through network crawling;
the network-crawled fish culture technology information comprises webpage information of a crawled culture technology education website and fish culture thesis information in a literature website.
5. A fish farm feeding decision method according to any one of claims 1-3, wherein the determining biomass information of the fish farm to be fed based on the underwater image specifically comprises:
determining image coordinate system coordinates of fish in the fish school to be fed, which are collected in the underwater image;
converting the coordinates of the image coordinate system into coordinates of a world coordinate system based on a camera coordinate system conversion rule;
determining a length and a width of the fish based on the world coordinate system coordinates, determining a surface area and a fish volume mass of the fish based on the length and the width;
wherein the biomass information comprises length, width, surface area and fish volume mass of the fish.
6. A fish farm feeding decision method according to any one of claims 1-3, wherein the determining of the farming density of the fish farm to be fed based on the underwater image comprises:
inputting the underwater image into a density map extraction model, and outputting a density map of the fish school to be fed;
the density map extraction model is obtained by training based on a sample underwater image and a corresponding density map label, and the neural network construction in the density map extraction model training process comprises the steps of generating a low-resolution density map LR-CNN network and generating a high-resolution density map HR-CNN network;
determining the culture density of the fish school to be fed based on the density map.
7. A fish swarm feeding decision method according to any one of claims 1-3, wherein the querying the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type of fed feed and the feeding amount specifically comprises:
introducing target biomass information of a fish swarm to be fed, target breeding density, artificially determined target fish type and target feed type into the fish nutrition feeding knowledge base for searching, and outputting a target feeding feed model and a target feeding amount corresponding to a search result target relation chain;
wherein the target relation chain comprises a corresponding relation chain of the type of the target fish, the type of the target feed, the target breeding density, the target biomass index, the target feeding amount and the target feed model.
8. A shoal feeding decision device, comprising:
the building unit is used for obtaining fish nutrition feeding data and building a fish nutrition feeding knowledge base according to preset rules;
the preset rule is analysis processing of fish nutrition feeding data of an LDA model based on linear discriminant analysis, and the fish nutrition feeding knowledge base comprises a relation chain of fish type-feed type-fish culture density-fish biomass index-feeding amount-feed type;
the collecting unit is used for collecting an underwater image of a fish school to be fed and determining biomass information and culture density of the fish school to be fed based on the underwater image;
and the decision unit is used for inquiring the fish nutrition feeding knowledge base based on the biomass information and the culture density to determine the type and the feeding amount of the fed feed.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the fish herd feeding decision method as claimed in any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of fish herd feeding decision as claimed in any one of claims 1 to 7.
CN202110662613.1A 2021-06-15 2021-06-15 Fish shoal feeding decision method and device, electronic equipment and storage medium Pending CN113487143A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110662613.1A CN113487143A (en) 2021-06-15 2021-06-15 Fish shoal feeding decision method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110662613.1A CN113487143A (en) 2021-06-15 2021-06-15 Fish shoal feeding decision method and device, electronic equipment and storage medium

Publications (1)

Publication Number Publication Date
CN113487143A true CN113487143A (en) 2021-10-08

Family

ID=77934947

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110662613.1A Pending CN113487143A (en) 2021-06-15 2021-06-15 Fish shoal feeding decision method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113487143A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114240686A (en) * 2022-02-24 2022-03-25 深圳市旗扬特种装备技术工程有限公司 Wisdom fishery monitoring system
CN116630080A (en) * 2023-07-25 2023-08-22 广东省农业科学院动物科学研究所 Method and system for determining capacity of aquatic product intensive culture feed based on image recognition
CN117223666A (en) * 2023-09-15 2023-12-15 中国科学院水生生物研究所 Precise feeding intelligent system development platform for aquaculture environment and use method
CN117556980A (en) * 2024-01-11 2024-02-13 中国水产科学研究院南海水产研究所 Unmanned feeding ship path planning analysis method, system and storage medium

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105512347A (en) * 2016-01-27 2016-04-20 北京航空航天大学 Information processing method based on geographic topic model
CN106777043A (en) * 2016-12-09 2017-05-31 宁波大学 A kind of academic resources acquisition methods based on LDA
CN107526819A (en) * 2017-08-29 2017-12-29 江苏飞搏软件股份有限公司 A kind of big data the analysis of public opinion method towards short text topic model
CN109272546A (en) * 2018-09-14 2019-01-25 中国农业大学 A kind of fry length measurement method and system
CN109992673A (en) * 2019-04-10 2019-07-09 广东工业大学 A kind of knowledge mapping generation method, device, equipment and readable storage medium storing program for executing
CN110352888A (en) * 2019-06-28 2019-10-22 唐山哈船科技有限公司 A kind of automatic charging machine people and its feeding method
CN111125355A (en) * 2018-10-31 2020-05-08 北京国双科技有限公司 Information processing method and related equipment
CN111657189A (en) * 2020-06-15 2020-09-15 张玉红 Fishery aquaculture daily monitoring management system based on big data
CN111797204A (en) * 2020-07-01 2020-10-20 北京三快在线科技有限公司 Text matching method and device, computer equipment and storage medium

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105512347A (en) * 2016-01-27 2016-04-20 北京航空航天大学 Information processing method based on geographic topic model
CN106777043A (en) * 2016-12-09 2017-05-31 宁波大学 A kind of academic resources acquisition methods based on LDA
CN107526819A (en) * 2017-08-29 2017-12-29 江苏飞搏软件股份有限公司 A kind of big data the analysis of public opinion method towards short text topic model
CN109272546A (en) * 2018-09-14 2019-01-25 中国农业大学 A kind of fry length measurement method and system
CN111125355A (en) * 2018-10-31 2020-05-08 北京国双科技有限公司 Information processing method and related equipment
CN109992673A (en) * 2019-04-10 2019-07-09 广东工业大学 A kind of knowledge mapping generation method, device, equipment and readable storage medium storing program for executing
CN110352888A (en) * 2019-06-28 2019-10-22 唐山哈船科技有限公司 A kind of automatic charging machine people and its feeding method
CN111657189A (en) * 2020-06-15 2020-09-15 张玉红 Fishery aquaculture daily monitoring management system based on big data
CN111797204A (en) * 2020-07-01 2020-10-20 北京三快在线科技有限公司 Text matching method and device, computer equipment and storage medium

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114240686A (en) * 2022-02-24 2022-03-25 深圳市旗扬特种装备技术工程有限公司 Wisdom fishery monitoring system
CN116630080A (en) * 2023-07-25 2023-08-22 广东省农业科学院动物科学研究所 Method and system for determining capacity of aquatic product intensive culture feed based on image recognition
CN116630080B (en) * 2023-07-25 2024-01-26 广东省农业科学院动物科学研究所 Method and system for determining capacity of aquatic product intensive culture feed based on image recognition
CN117223666A (en) * 2023-09-15 2023-12-15 中国科学院水生生物研究所 Precise feeding intelligent system development platform for aquaculture environment and use method
CN117556980A (en) * 2024-01-11 2024-02-13 中国水产科学研究院南海水产研究所 Unmanned feeding ship path planning analysis method, system and storage medium
CN117556980B (en) * 2024-01-11 2024-03-22 中国水产科学研究院南海水产研究所 Unmanned feeding ship path planning analysis method, system and storage medium

Similar Documents

Publication Publication Date Title
CN113487143A (en) Fish shoal feeding decision method and device, electronic equipment and storage medium
CN107667903B (en) Livestock breeding living body weight monitoring method based on Internet of things
Chen et al. Feed intake prediction model for group fish using the MEA-BP neural network in intensive aquaculture
Sant’Ana et al. Weighing live sheep using computer vision techniques and regression machine learning
Li et al. Characterizing gully cross section and modelling gully volume in hilly loess region of western Shanxi Province.
Jelić Mrčelić et al. An Overview of Atlantic Bluefin Tuna Farming Sustainability in the Mediterranean with Special Regards to the Republic of Croatia
Guan et al. Cutting tool wear recognition based on MF-DFA feature and LS-SVM algorithm.
Lane The Mass Diagnosis of Ankylostome Infestation.(Part I.).
Liu et al. Wood defect recognition based on optimized convolution neural network algorithm.
Teng et al. Non-contact sow body condition scoring method based on Kinect sensor.
Bazrafkan et al. Capability of LISS III data for forest canopy density mapping in Zagros forests (Case study: Marivan forests).
Guo et al. Improvement on image detection algorithm of droplets deposition characteristics.
Lugert et al. Evaluating the most suitable nonlinear growth model for turbot (Scophthalmus maximus) in aquaculture 2 (weight application): Multi‐criteria model selection and growth prediction
Przytulska et al. Comparative assessment of polyvinylpyrrolidone type of membranes based on porosity analysis.
Jing et al. Three-dimensional distribution of fish using an imaging sonar.
Caubet et al. NEIGHBOUR-IN: Image processing software for spatial analysis of animal grouping
Roy et al. Comparative efficiency of different artificial intelligence based models for predicting density dependent saltwater intrusion processes in coastal aquifers and saltwater intrusion management utilizing the best performing model.
Subba et al. Relative condition factor, length-weight relationship and sex ratio of copper mahseer, Neolissochilus hexagonolepis (McClelland, 1839) from Tamor River, Nepal.
Ramírez-Gutiérrez et al. Eggshell deformation detection applying computer vision.
Chen et al. Identification of CT image defects in wood based on convolution neural network.
Oyedele et al. Impact of training and development on the quality of fish produced: Evidence from fish farmers in Odogbolu, Ogun State, Nigeria
Dehnoei et al. A new strategy to reduce factors number in water quality measurements using Taguchi method-case study: the Karaj River.
Langlands Reports of the Committee upon the Physiology of Vision. IV. Experiments on Binocular Vision.
Liu et al. Research on image recognition of five bark texture images based on deep learning.
Abdanan et al. Computational estimation of L* a* b* units from RGB using machine vision.

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