CN117456395B - Sea and land two-domain garbage recycling planning method based on machine vision - Google Patents

Sea and land two-domain garbage recycling planning method based on machine vision Download PDF

Info

Publication number
CN117456395B
CN117456395B CN202311797667.4A CN202311797667A CN117456395B CN 117456395 B CN117456395 B CN 117456395B CN 202311797667 A CN202311797667 A CN 202311797667A CN 117456395 B CN117456395 B CN 117456395B
Authority
CN
China
Prior art keywords
garbage
garbage disposal
disposal device
algorithm
sea
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202311797667.4A
Other languages
Chinese (zh)
Other versions
CN117456395A (en
Inventor
谢泽文
刘长红
柯镇宇
陈奎庚
张成云
邹涛
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.)
Guangzhou University
Original Assignee
Guangzhou 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 Guangzhou University filed Critical Guangzhou University
Priority to CN202311797667.4A priority Critical patent/CN117456395B/en
Publication of CN117456395A publication Critical patent/CN117456395A/en
Application granted granted Critical
Publication of CN117456395B publication Critical patent/CN117456395B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/17—Terrestrial scenes taken from planes or by drones
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • 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
    • G06N3/12—Computing arrangements based on biological models using genetic models
    • G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N7/00—Computing arrangements based on specific mathematical models
    • G06N7/02—Computing arrangements based on specific mathematical models using fuzzy logic
    • 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
    • G06Q10/30—Administration of product recycling or disposal
    • 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/26—Government or public services
    • 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
    • G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
    • 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
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G—PHYSICS
    • G06—COMPUTING; 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
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/07—Target detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Business, Economics & Management (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Biophysics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Tourism & Hospitality (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Molecular Biology (AREA)
  • Economics (AREA)
  • Biomedical Technology (AREA)
  • General Business, Economics & Management (AREA)
  • Algebra (AREA)
  • Strategic Management (AREA)
  • Marketing (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Fuzzy Systems (AREA)
  • Sustainable Development (AREA)
  • Remote Sensing (AREA)
  • Entrepreneurship & Innovation (AREA)

Abstract

The invention discloses a sea and land two-domain garbage recycling planning method based on machine vision, which comprises the following steps: detecting garbage data by using a target detection algorithm, detecting and identifying biological data by using a thermal imaging and digital image segmentation algorithm, and performing map modeling by using a slam modeling technology to segment sea and land so as to obtain a two-domain map; acquiring the number and coordinates of the garbage disposal devices and performing garbage recycling path pre-planning; each garbage disposal device performs garbage recycling tasks according to the garbage recycling path pre-planning; calculating wave interference degree when executing garbage recycling tasks, and constructing a fuzzy relation matrix according to the wave interference degree and the current moving speed; outputting a corresponding control moving speed according to the detected biological condition; calculating an operational risk degree when the garbage disposal device is close to garbage; constructing a fuzzy relation matrix according to the operation risk degree and the current operation speed, and outputting and controlling the operation speed; and recording garbage recycling task data, optimizing parameter setting by using an optimization algorithm, and improving the recycling efficiency of sea and land garbage.

Description

Sea and land two-domain garbage recycling planning method based on machine vision
Technical Field
The invention belongs to the technical field of garbage collection and machine vision, and particularly relates to a sea-land two-domain garbage collection planning method based on machine vision.
Background
With the development of ocean resources by human beings, the environment at the offshore is severely polluted, and the garbage existing on the beach and the ocean can be moved under the influence of sea waves, so that the garbage at the offshore has the characteristic of alternately appearing in the two areas of the sea and the land. The garbage generated on the beach by human beings can be rolled to the offshore position, the ocean garbage at the offshore position can be shot to the coast, so that the garbage can dynamically coexist in two areas of the sea and the land, and the garbage cleaning in the two areas of the offshore position is extremely difficult. In the prior art, the garbage recovery strategy mainly comprises only a single sea area or land (beach), so that the scheme for the garbage recovery decision of the sea-land two-area is less, and the method is difficult to adapt to the characteristic of dynamic movement of offshore garbage in the sea-land two-area. In addition, the biological safety and the influence of offshore exposure are still few, the influence of garbage danger on the living beings is not considered, and finally, the prior art is difficult to make a good decision plan for some fuzzy concepts such as the distance of sea waves, the speed of the current device and the like.
Disclosure of Invention
The invention mainly aims to overcome the defects and shortcomings of the prior art, and provides a sea-land two-domain garbage recycling planning method based on machine vision, which considers the influence of sea wave interference, garbage danger and biological factors, realizes the recycling planning of sea-land two-domain garbage by calculating the sea wave interference degree and operation risk degree, accurately controls a garbage recycling device, improves recycling efficiency and reduces the influence on living things.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a sea and land two-domain garbage recycling planning method based on machine vision comprises the following steps:
the unmanned aerial vehicle is controlled to detect garbage data in an area to be planned by using a target detection algorithm, and biological data is detected and identified by using a thermal imaging and digital image segmentation algorithm; performing map modeling on the area to be planned by using a slam modeling technology according to the garbage data and the biological data, and dividing ocean and land by using OpenCV to obtain a two-domain map; the garbage data comprises garbage types, garbage amounts and garbage positions; the biological data includes biological species, biological quantity, and biological location;
acquiring the number and coordinates of garbage disposal devices in a region to be planned, and performing garbage recycling path pre-planning; the garbage disposal device is provided with a target detection algorithm;
each garbage disposal device performs garbage recycling tasks according to the garbage recycling path pre-planning;
when the garbage recycling task is executed, the distance L between sea waves and garbage is used for 1 Distance L from garbage disposal device to garbage 2 Current moving speed V of garbage disposal apparatus F Calculating the sea wave interference degree O, and according to the sea wave interference degree O and the current moving speed V of the garbage disposal device F Constructing a fuzzy relation matrix, and outputting a corresponding control moving speed according to the detected biological condition; the calculation formula of the sea wave interference degree O is as follows:
O=(W 2 L 2 -W 1 L 1 )×W 3 V F ，
wherein L is 1 For the distance between the sea wave and the garbage, if no sea wave is detected, L 1 Is 0; w (W) 1 、W 2 、W 3 Are all weight coefficients;
when the garbage disposal device is close to garbage, recognizing the garbage hazard level D according to the garbage hazard level, counting the biological quantity B in the visual range of the garbage disposal device, and combining the current operation speed V of the garbage disposal device W Calculating the operation risk degree R; according to the operation risk degree R and the current operation speed V W Constructing a fuzzy relation matrix and outputting and controlling the operation speedThe method comprises the steps of carrying out a first treatment on the surface of the The calculation formula of the operation risk degree R is as follows:
R = (W 4 D+W 5 B) ×W 6 V W ，
wherein B is the number of organisms surrounding the waste, b=0 when no organisms surrounding the waste are detected; w (W) 4 、W 5 、W 6 Are all weight coefficients;
and recording data of each garbage collection task of each garbage treatment device, and optimizing parameter setting by using an optimization algorithm.
As an optimal technical scheme, the target detection algorithm is obtained by collecting a garbage data set training neural network of a region to be planned through an unmanned aerial vehicle carrying a light filter;
the optical filter is used during large exposure of the sun;
the garbage collection path is pre-planned, and specifically comprises the following steps:
using the number of garbage treatment devices as the clustering number of garbage in the two-domain map, and performing cluster analysis on the garbage data by using a clustering algorithm to obtain the central coordinates of each cluster;
dividing the responsible areas of the garbage disposal devices on the two-domain map according to the clustering result;
in the responsible area of each garbage disposal device, taking the position of the garbage disposal device as a starting point, taking organisms as barriers, and carrying out garbage recycling path pre-planning by using a path planning algorithm by taking garbage in the responsible area as a terminal point;
the goal of the path planning algorithm is to avoid living beings and complete the recovery of all the garbage in the responsible area with the shortest path.
As a preferable technical scheme, the method is characterized in that the method is based on the sea wave interference degree O and the current moving speed V of the garbage disposal device F The fuzzy relation matrix is constructed, specifically:
the argument of the sea wave interference degree fuzzy matrix is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the sea wave interference degree O, which is expressed as follows:
，
OO represents no wave interference, OS represents weak wave interference, OM represents general wave interference, and OL represents strong wave interference;
the domain of the current moving speed fuzzy matrix of the garbage disposal device is defined as {0,3,6,9}, and the membership function is as follows:
，
constructing a fuzzy matrix of the current moving speed of the garbage disposal device, wherein the fuzzy matrix is expressed as follows:
，
wherein FO represents that the garbage disposal device is currently stationary, FS represents that the garbage disposal device is currently moving slowly, FM represents that the garbage disposal device is moving at a current normal speed, and FL represents that the garbage disposal device is moving rapidly;
based on the fuzzy matrix of the current moving speed and the wave interference, constructing a fuzzy relation matrix M of the current moving speed and the wave interference 1 ：
。
As an optimal technical scheme, in the moving process of the garbage disposal device, firstly, according to the fuzzy relation matrix M 1 Control movement speed of output garbage disposal apparatus=V F ×M 1 ；
Then, according to the biological condition detected by the garbage disposal device, the maximum membership method is used for controlling the moving speedUpdating current moving speed V of garbage disposal apparatus F Comprising:
when the garbage disposal apparatus detects living things in the visual range, the current moving speed V of the garbage disposal apparatus F =Ln (e+b), where e is a natural index;
when the garbage disposal apparatus does not detect the living being in the visual range, the current moving speed V of the garbage disposal apparatus F =；
The control of the moving speedThe membership function of (2) is:
。
as an optimal technical scheme, the garbage risk is identified and classified according to a target detection algorithm, wherein the garbage risk comprises no risk, weak risk, general risk and high risk; the garbage risk level is that the garbage risk level is quantized to obtain D= {0,1,2,3}, namely, the risk level is not 0, the risk level is weak 1, the risk level is generally 2, and the risk level is high 3;
the operation risk degree R and the current operation speed V of the garbage disposal device W Carrying out blurring processing to construct a corresponding blurring matrix, specifically:
the domain of the fuzzy matrix of the operation risk degree is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the operation risk degree R, which is expressed as follows:
，
wherein RO represents no risk of operation, RS represents weak risk of operation, RM represents general risk of operation, and RL represents high risk of operation;
the domain of the current operation speed fuzzy matrix of the garbage disposal device is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the current operation speed of the garbage disposal device, wherein the fuzzy matrix is expressed as:
，
wherein WL represents the quick operation of the garbage disposal device, WM represents the normal operation of the garbage disposal device, WS represents the slow operation of the garbage disposal device, and WO represents the stop operation of the garbage disposal device.
As a preferable technical solution, the output control operation speed specifically includes:
constructing a fuzzy relation matrix M of the current operation speed and operation risk degree of the garbage disposal device 2 ：
，
According to the fuzzy relation matrix M 2 Control operation speed of output garbage disposal apparatus= V W ×M 2 ；
According to the control operation speed by using the maximum membership methodUpdating current operating speed V of garbage disposal apparatus W =The method comprises the steps of carrying out a first treatment on the surface of the The control operation speed +.>The membership function of (2) is:
。
as an optimal technical scheme, a defogging algorithm and a night vision algorithm are also deployed on the unmanned aerial vehicle and the garbage disposal device;
solar panels are arranged on the unmanned aerial vehicle and the garbage disposal device;
the optimization algorithm is one or two of a genetic algorithm, an ant colony algorithm, a simulated annealing algorithm and a machine learning algorithm.
As an optimal technical scheme, when the two optimization algorithms are adopted, the optimization is performed by using the corresponding optimization algorithm according to the data volume of the garbage collection task, specifically:
according to the recorded garbage collection task data, performing quality evaluation on the garbage collection task executed by each garbage treatment device, wherein an evaluation formula is as follows:
F(R,O,E) = (α∑R+β∑O-γE)×δT，
wherein R is the operation risk degree, O is the sea wave interference degree, E is the energy consumption of the garbage disposal device, and T is the total time for completing the garbage recycling task; alpha, beta, gamma and delta are weight coefficients; the operation risk degree R and the sea wave interference degree O are obtained discretely;
setting each parameter in the fuzzy relation matrix as an influence factor; the parameters comprise current moving speed, sea wave interference degree, current operation speed and operation risk degree;
when the recorded garbage collection task data is less than 1000 pieces, optimizing parameter setting by using a genetic algorithm, an ant colony algorithm or a simulated annealing algorithm;
when the recorded garbage collection task data is not less than 1000 pieces, optimizing parameter setting by using a machine learning algorithm.
As a preferable technical scheme, the optimizing parameter setting using the genetic algorithm specifically includes:
initializing a genetic algorithm, taking the most completed garbage collection tasks as a main target, taking the lowest operation risk degree, the lowest sea wave interference degree and the least energy consumption of a garbage treatment device as secondary targets, and constructing an objective function; the objective function is expressed as:
Q = K 1 ×n/N+K 2 ×F(R,O,E)，
wherein N is the number of times that a certain garbage disposal device completes the garbage collection task, and N represents the number of times that all garbage disposal devices complete the garbage collection task; k (K) 1 、K 2 Is a weight coefficient;
using genetic algorithm to carry out coding, fitness function calculation, selection operation, crossover operation and mutation operation to output an optimal population;
and decoding the optimal population to obtain optimal parameters.
As a preferable technical solution, the optimizing parameter setting using the machine learning algorithm specifically includes:
the neural network is selected as a model of a machine learning algorithm, a garbage collection task is completed as a main target, the operation risk degree is minimum, the sea wave interference degree is minimum, the garbage treatment device energy consumption is minimum as a secondary target, and the quality evaluation is used as a model output label;
dividing recorded garbage collection task data according to a proportion to obtain a training set, a testing set and a verification set;
training, testing and verifying the model of the machine learning algorithm to finally obtain the optimal parameters.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. because the offshore rubbish is easily affected by sea waves and has the specificity of coexistence of two areas of sea and land, the invention provides a planning method for rubbish recovery of two areas of service; the factors such as sea wave interference, garbage dangers (such as glass ceramic breakage and battery forced explosion), biological influence and the like are fully considered, the unmanned aerial vehicle is used for pre-planning, fine planning such as sea wave interference degree and operation risk degree is combined and calculated, more excellent garbage operation and path planning are realized, and garbage recovery efficiency is improved.
2. According to the invention, by constructing the fuzzy relation matrix and adopting a fuzzy control method to make fuzzy decisions in the garbage device, the fuzzy control with adjustable parameters is realized, the dynamic control of the garbage treatment device is improved, the influence of biological factors is considered, and the harm to living beings in garbage recycling operation is reduced.
3. According to the invention, the garbage collection operation is further optimized according to a large amount of data generated when the garbage treatment device executes the task, and the related parameters in the garbage collection operation are optimized through an optimization algorithm, so that the accuracy and the efficiency of garbage collection are further improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is an overall flowchart of a sea-land two-domain garbage collection planning method based on machine vision in an embodiment of the invention.
Fig. 2 is a flow chart of the garbage collection path pre-planning in the embodiment of the invention.
FIG. 3 is a graph of the result of the pre-planning in an embodiment of the present invention.
Fig. 4 is a flowchart of outputting a corresponding control movement speed in an embodiment of the present invention.
Fig. 5 is a flowchart of outputting a corresponding garbage operation speed in an embodiment of the present invention.
FIG. 6 is a flow chart of optimizing parameter settings using an optimization algorithm in an embodiment of the invention.
Detailed Description
In order to enable those skilled in the art to better understand the present application, the following description will make clear and complete descriptions of the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.
As shown in fig. 1, the sea-land two-domain garbage collection planning method based on machine vision in this embodiment includes the following steps:
s1, controlling an unmanned aerial vehicle to detect garbage data in a region to be planned by using a target detection algorithm, and detecting and identifying biological data by using a thermal imaging and digital image segmentation algorithm; performing map modeling on the area to be planned by using a slam modeling technology according to the garbage data and the biological data, and dividing ocean and land by using OpenCV to obtain a two-domain map; the garbage data comprise garbage types, garbage quantity, garbage positions and the like; biological data includes biological species, biological quantity, biological location, and the like;
s2, acquiring the number and coordinates of the garbage disposal devices in the area to be planned, and performing garbage recycling path pre-planning; the garbage disposal device is also provided with a target detection algorithm;
s3, each garbage disposal device performs a garbage recycling task according to the garbage recycling path pre-planning;
s4, when the garbage recycling task is executed, according to the distance L between sea waves and garbage 1 Distance L from garbage disposal device to garbage 2 Current moving speed V of garbage disposal apparatus F Calculating the sea wave interference degree O, and according to the sea wave interference degree O and the current moving speed V of the garbage disposal device F Constructing a fuzzy relation matrix and outputting corresponding control according to the detected biological conditionA moving speed; the calculation formula of the sea wave interference degree O is as follows:
O=(W 2 L 2 -W 1 L 1 )×W 3 V F ；
wherein L is 1 For the distance between the sea wave and the garbage, if no sea wave is detected, L 1 0, otherwise L 1 Is an actual value; w (W) 1 、W 2 、W 3 Are all weight coefficients;
s5, when the garbage disposal device is close to garbage, recognizing the garbage danger level D according to the garbage danger degree, counting the biomass B in the visual range of the garbage disposal device, and combining the current operation speed V of the garbage disposal device W Calculating the operation risk degree R; according to the operation risk degree R and the current operation speed V W Constructing a fuzzy relation matrix and outputting and controlling the operation speed; the calculation formula of the operation risk degree R is as follows:
R = (W 4 D+W 5 B) ×W 6 V W ；
wherein B is the number of organisms around the garbage, b=0 when no organisms around the garbage are detected, otherwise B is the actual value; w (W) 4 、W 5 、W 6 Are all weight coefficients;
s6, recording data of each garbage collection task of each garbage treatment device, and optimizing parameter setting by using an optimization algorithm.
Furthermore, in order to reduce the influence of reflection of light in the ocean, the target detection algorithm is obtained by collecting a rubbish data set of a region to be planned through an unmanned aerial vehicle carrying a light filter and training a neural network after marking is completed. The filter is mounted for improving the garbage recognition accuracy of the target detection algorithm by physically reducing the influence of exposure, and the filter is specifically bandpass filter for filtering out long waves of 600-900 nm; meanwhile, the optical filter is used in the large exposure of the sun. In this embodiment, the target detection algorithm adopts a yolov4 target detection algorithm (which may also be a yolov5 target detection algorithm, a yolov7 target detection algorithm, etc.), accurately identifies and screens out floating garbage on the sea and garbage on the beach, and calculates and counts the number of the floating garbage and the garbage on the beach; and meanwhile, the garbage target is subjected to distance measurement and marked on the two-domain map.
Further, as shown in fig. 2, the pre-planning of the garbage collection path is specifically as follows:
using the number of garbage treatment devices as the clustering number of garbage in the two-domain map, and performing cluster analysis on the garbage data by using a clustering algorithm to obtain the central coordinates of each cluster;
dividing the responsible areas of the garbage disposal devices on the two-domain map according to the clustering result; if the number of the garbage disposal devices is 3, the number of the garbage clusters is 3, and then one garbage disposal device is responsible for one area;
in the responsible area of each garbage disposal device, taking the position of the garbage disposal device as a starting point, taking organisms as barriers, and carrying out garbage recycling path pre-planning by using a path planning algorithm by taking garbage in the responsible area as a terminal point; the objective of the path planning algorithm is to avoid living things and complete recovery of all garbage in the responsible area with the shortest path.
In this embodiment, the clustering algorithm uses a K-Means clustering algorithm, the path planning algorithm uses an a-x algorithm or a dijkstra algorithm, and the result of the final garbage collection path pre-planning is shown in fig. 3, where 4 garbage disposal devices are put into the system, the clustering result is 4, and each garbage disposal device is responsible for an area while avoiding living beings.
Further, in the process that the garbage treatment device moves to the garbage position, as the garbage in the ocean can move under the influence of sea waves and organisms, and then the garbage position is changed, the moving speed of the garbage treatment device is subjected to fuzzy control by calculating the interference degree of the sea waves and constructing a fuzzy relation matrix. As shown in FIG. 4, first, the sea wave disturbance degree O and the current moving speed V of the garbage disposal apparatus are constructed F Specifically:
the argument of the sea wave interference degree fuzzy matrix is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the sea wave interference degree O, which is expressed as follows:
，
OO represents no wave interference, OS represents weak wave interference, OM represents general wave interference, and OL represents strong wave interference;
the domain of the current moving speed fuzzy matrix of the garbage disposal device is defined as {0,3,6,9}, and the membership function is as follows:
，
constructing a fuzzy matrix of the current moving speed of the garbage disposal device, wherein the fuzzy matrix is expressed as follows:
，
where FO indicates that the refuse treatment apparatus is currently stationary, FS indicates that the refuse treatment apparatus is currently moving slowly, FM indicates that the refuse treatment apparatus is moving at a current normal speed, and FL indicates that the refuse treatment apparatus is moving rapidly.
Then constructing a fuzzy relation matrix M of the current moving speed and the wave interference based on the fuzzy matrix of the current moving speed and the wave interference 1 ：
，
Finally, in the moving process of the garbage device, according to the fuzzy relation matrix M 1 Control movement speed of output garbage disposal apparatus=V F ×M 1 And according to the biological condition detected by the garbage disposal device, the maximum membership method is used for controlling the moving speed>Updating refuse treatment apparatusCurrent moving speed V F Comprising:
when the garbage disposal device detects living things in the visual range, the speed needs to be reduced to avoid hurting living things, and the current moving speed V of the garbage disposal device F =Ln (e+B), where e is a natural exponent, ensuring that the denominator is not 0;
when the garbage disposal apparatus does not detect the living being in the visual range, the current moving speed V of the garbage disposal apparatus F =；
Controlling the speed of movementThe membership function of (2) is:
。
further, when the garbage disposal device works close to the garbage, if the garbage is a fragile product (such as glass, ceramic, electric bulb, battery, etc.), the garbage may damage surrounding organisms (such as fragments scratch caused by breakage of glass, ceramic, etc., secondary damage caused by explosion of the battery, etc.) and affect the working efficiency; therefore, the operation risk degree is calculated by identifying the garbage risk level and combining the surrounding biological data, and the fuzzy matrix is constructed by combining the current operation speed, so that the operation speed is adjusted, and the influence on the surrounding biological data is avoided. The operation speed in the present application may be a conveyor belt speed, a garbage sorting speed, a garbage compression speed, etc. of the garbage treatment device, and is determined according to the actual function of the garbage treatment device; since the speed of operation is determined by the device itself, and the present application considers the external effects of the risk of refuse and biological factors, no further discussion will be made. As shown in fig. 5, first, a garbage hazard level D is identified according to a garbage hazard level; the garbage risk is identified and classified according to the target detection algorithm and the material risk, including no risk (such as plastic bottles and pop cans), weak risk (such as batteries), general risk (such as ceramics and tiles) and high risk (such as glass bottles and electric bulbs); then, the garbage risk level is that the garbage risk level is further quantized to obtain D= {0,1,2,3}, namely, the risk level is not 0, the risk level is weak 1, the risk level is generally 2, and the risk level is high 3;
secondly, the operation risk degree R and the current operation speed V of the garbage disposal device W Carrying out blurring processing to construct a corresponding blurring matrix, specifically:
the domain of the fuzzy matrix of the operation risk degree is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the operation risk degree R, which is expressed as follows:
，
wherein RO represents no risk of operation, RS represents weak risk of operation, RM represents general risk of operation, and RL represents high risk of operation;
the domain of the current operation speed fuzzy matrix of the garbage disposal device is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the current operation speed of the garbage disposal device, wherein the fuzzy matrix is expressed as:
，
wherein WL represents the quick operation of the garbage disposal device, WM represents the normal operation of the garbage disposal device, WS represents the slow operation of the garbage disposal device, and WO represents the stop operation of the garbage disposal device.
Then construct garbageFuzzy relation matrix M of current operation speed and operation risk degree of processing device 2 ：
，
Then according to the fuzzy relation matrix M 2 Control operation speed of output garbage disposal apparatus= V W ×M 2 ；
According to the control operation speed by using the maximum membership methodUpdating current operating speed V of garbage disposal apparatus W =；
Controlling the speed of operationThe membership function of (2) is:
。
furthermore, in order to ensure the precision of the unmanned aerial vehicle and the garbage disposal device for recognizing garbage and living things, a defogging algorithm and a night vision algorithm are deployed on the unmanned aerial vehicle and the garbage disposal device, so that the reliability of path planning is ensured. And solar panels are arranged on the unmanned aerial vehicle and the garbage disposal device, so that energy consumption is reduced. Meanwhile, the condition of executing garbage collection tasks is different every time, so that parameter setting needs to be optimized, and the optimization algorithm in the method can be one or two of a genetic algorithm, an ant colony algorithm, a simulated annealing algorithm and a machine learning algorithm.
Further, as shown in fig. 6, when the optimization algorithms are two, the optimization algorithm needs to be optimized according to the data volume of the garbage collection task, specifically:
according to the recorded garbage collection task data, performing quality evaluation on the garbage collection task executed by each garbage treatment device, wherein an evaluation formula is as follows:
F(R,O,E) = (α∑R+β∑O-γE)×δT，
wherein R is the operation risk degree, O is the sea wave interference degree, E is the energy consumption of the garbage disposal device, and T is the total time for completing the garbage recycling task; alpha, beta, gamma and delta are weight coefficients; note that the operation risk level R and the sea wave interference level O are obtained discretely rather than always;
setting each parameter in the fuzzy relation matrix as an influence factor; the parameters comprise current moving speed, sea wave interference degree, current operation speed and operation risk degree;
and then optimizing parameter setting by using a corresponding optimization algorithm according to the data size of the garbage collection task, wherein the method comprises the following steps:
when the recorded garbage collection task data is less than 1000 pieces, optimizing parameter setting by using a genetic algorithm, an ant colony algorithm or a simulated annealing algorithm;
when the recorded garbage collection task data is not less than 1000 pieces, optimizing parameter setting by using a machine learning algorithm.
Further, the optimization of the parameter settings using the genetic algorithm is specifically:
initializing a genetic algorithm, taking the most completed garbage collection tasks as a main target, taking the lowest operation risk degree, the lowest sea wave interference degree and the least energy consumption of a garbage treatment device as secondary targets, and constructing an objective function, wherein the objective function is expressed as follows:
Q = K 1 ×n/N+K 2 ×F(R,O,E)，
wherein N is the number of times that a certain garbage disposal device completes the garbage collection task, and N represents the number of times that all garbage disposal devices complete the garbage collection task; k (K) 1 、K 2 Is a weight coefficient;
using genetic algorithm to carry out coding, fitness function calculation, selection operation, crossover operation and mutation operation to output an optimal population;
and decoding the optimal population to obtain optimal parameters.
Further, the optimization of the parameter settings using the machine learning algorithm is specifically:
the neural network is selected as a model of a machine learning algorithm, a garbage collection task is completed as a main target, the operation risk degree is minimum, the sea wave interference degree is minimum, the garbage treatment device energy consumption is minimum as a secondary target, and the quality evaluation is used as a model output label;
dividing recorded garbage collection task data according to a proportion to obtain a training set, a testing set and a verification set;
training, testing and verifying the model of the machine learning algorithm to finally obtain the optimal parameters.
It should be noted that, for the sake of simplicity of description, the foregoing method embodiments are all expressed as a series of combinations of actions, but it should be understood by those skilled in the art that the present invention is not limited by the order of actions described, as some steps may be performed in other order or simultaneously in accordance with the present invention.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above examples, and any other changes, modifications, substitutions, combinations, and simplifications that do not depart from the spirit and principle of the present invention should be made in the equivalent manner, and the embodiments are included in the protection scope of the present invention.

Claims (10)

1. The sea-land two-domain garbage recycling planning method based on machine vision is characterized by comprising the following steps of:
the unmanned aerial vehicle is controlled to detect garbage data in an area to be planned by using a target detection algorithm, and biological data is detected and identified by using a thermal imaging and digital image segmentation algorithm; performing map modeling on the area to be planned by using a slam modeling technology according to the garbage data and the biological data, and dividing ocean and land by using OpenCV to obtain a two-domain map; the garbage data comprises garbage types, garbage amounts and garbage positions; the biological data includes biological species, biological quantity, and biological location;
acquiring the number and coordinates of garbage disposal devices in a region to be planned, and performing garbage recycling path pre-planning; the garbage disposal device is provided with a target detection algorithm;
each garbage disposal device performs garbage recycling tasks according to the garbage recycling path pre-planning;
when the garbage recycling task is executed, the distance L between sea waves and garbage is used for 1 Distance L from garbage disposal device to garbage 2 Current moving speed V of garbage disposal apparatus F Calculating the sea wave interference degree O, and according to the sea wave interference degree O and the current moving speed V of the garbage disposal device F Constructing a fuzzy relation matrix, and outputting a corresponding control moving speed according to the detected biological condition; the calculation formula of the sea wave interference degree O is as follows:
O=(W 2 L 2 -W 1 L 1 )×W 3 V F ，
wherein L is 1 For the distance between the sea wave and the garbage, if no sea wave is detected, L 1 Is 0; w (W) 1 、W 2 、W 3 Are all weight coefficients;
when the garbage disposal device is close to garbage, recognizing the garbage hazard level D according to the garbage hazard level, counting the biological quantity B in the visual range of the garbage disposal device, and combining the current operation speed V of the garbage disposal device W Calculating the operation risk degree R; according to the operation risk degree R and the current operation speed V W Constructing a fuzzy relation matrix and outputting and controlling the operation speed; the calculation formula of the operation risk degree R is as follows:
R = (W 4 D+W 5 B) ×W 6 V W ，
wherein B is the number of organisms within the visual range of the garbage disposal apparatus, and b=0 when no organisms within the visual range are detected; w (W) 4 、W 5 、W 6 Are all weight coefficients;
and recording data of each garbage collection task of each garbage treatment device, and optimizing parameter setting by using an optimization algorithm.
2. The machine vision-based sea-land two-domain garbage collection planning method according to claim 1, wherein the target detection algorithm is obtained by acquiring a garbage data set training neural network of a region to be planned through an unmanned aerial vehicle carrying a light filter;
the optical filter is used during large exposure of the sun;
the garbage collection path is pre-planned, and specifically comprises the following steps:
using the number of garbage treatment devices as the clustering number of garbage in the two-domain map, and performing cluster analysis on the garbage data by using a clustering algorithm to obtain the central coordinates of each cluster;
dividing the responsible areas of the garbage disposal devices on the two-domain map according to the clustering result;
in the responsible area of each garbage disposal device, taking the position of the garbage disposal device as a starting point, taking organisms as barriers, and carrying out garbage recycling path pre-planning by using a path planning algorithm by taking garbage in the responsible area as a terminal point;
the goal of the path planning algorithm is to avoid living beings and complete the recovery of all the garbage in the responsible area with the shortest path.
3. The machine vision-based sea-land two-domain garbage collection planning method according to claim 1, wherein the sea-wave interference degree O and the current moving speed V of the garbage disposal device are used for F The fuzzy relation matrix is constructed, specifically:
the argument of the sea wave interference degree fuzzy matrix is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the sea wave interference degree O, which is expressed as follows:
，
OO represents no wave interference, OS represents weak wave interference, OM represents general wave interference, and OL represents strong wave interference;
the domain of the current moving speed fuzzy matrix of the garbage disposal device is defined as {0,3,6,9}, and the membership function is as follows:
，
constructing a fuzzy matrix of the current moving speed of the garbage disposal device, wherein the fuzzy matrix is expressed as follows:
，
wherein FO represents that the garbage disposal device is currently stationary, FS represents that the garbage disposal device is currently moving slowly, FM represents that the garbage disposal device is moving at a current normal speed, and FL represents that the garbage disposal device is moving rapidly;
based on the fuzzy matrix of the current moving speed and the wave interference, constructing a fuzzy relation matrix M of the current moving speed and the wave interference 1 ：
。
4. The machine vision-based sea-land two-domain garbage collection planning method according to claim 3, wherein the garbage disposal device is moved according to the fuzzy relation matrix M 1 Control movement speed of output garbage disposal apparatus=V F ×M 1 ；
Then, according to the garbage disposal deviceThe detected biological condition is controlled according to the moving speed by using a maximum membership methodUpdating current moving speed V of garbage disposal apparatus F Comprising:
when the garbage disposal apparatus detects living things in the visual range, the current moving speed V of the garbage disposal apparatus F =Ln (e+b), where e is a natural index;
when the garbage disposal apparatus does not detect the living being in the visual range, the current moving speed V of the garbage disposal apparatus F =；
The control of the moving speedThe membership function of (2) is:
。
5. the machine vision-based sea-land two-domain garbage collection planning method according to claim 1, wherein the garbage risk is identified and classified according to a target detection algorithm, and the method comprises no risk, weak risk, general risk and high risk; the garbage risk level is that the garbage risk level is quantized to obtain D= {0,1,2,3}, namely, the risk level is not 0, the risk level is weak 1, the risk level is generally 2, and the risk level is high 3;
the operation risk degree R and the current operation speed V of the garbage disposal device W Carrying out blurring processing to construct a corresponding blurring matrix, specifically:
the domain of the fuzzy matrix of the operation risk degree is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the operation risk degree R, which is expressed as follows:
，
wherein RO represents no risk of operation, RS represents weak risk of operation, RM represents general risk of operation, and RL represents high risk of operation;
the domain of the current operation speed fuzzy matrix of the garbage disposal device is defined as {0,1,2,3}, and the membership function is as follows:
，
constructing a fuzzy matrix of the current operation speed of the garbage disposal device, wherein the fuzzy matrix is expressed as:
，
wherein WL represents the quick operation of the garbage disposal device, WM represents the normal operation of the garbage disposal device, WS represents the slow operation of the garbage disposal device, and WO represents the stop operation of the garbage disposal device.
6. The machine vision-based sea-land two-domain garbage collection planning method according to claim 5, wherein the output control operation speed is specifically:
constructing a fuzzy relation matrix M of the current operation speed and operation risk degree of the garbage disposal device 2 ：
，
According to the fuzzy relation matrix M 2 Control operation of output garbage disposal apparatusSpeed of speed= V W ×M 2 ；
According to the control operation speed by using the maximum membership methodUpdating current operating speed V of garbage disposal apparatus W =/>The method comprises the steps of carrying out a first treatment on the surface of the The control operation speed +.>The membership function of (2) is:
。
7. the machine vision-based sea-land two-domain garbage collection planning method according to claim 1, wherein a defogging algorithm and a night vision algorithm are further deployed on the unmanned aerial vehicle and the garbage disposal device;
solar panels are arranged on the unmanned aerial vehicle and the garbage disposal device;
the optimization algorithm is one or two of a genetic algorithm, an ant colony algorithm, a simulated annealing algorithm and a neural network algorithm.
8. The machine vision-based sea-land two-domain garbage collection planning method according to claim 7, wherein when the two optimization algorithms are adopted, the optimization is performed by using the corresponding optimization algorithm according to the data volume of the garbage collection task, specifically:
according to the recorded garbage collection task data, performing quality evaluation on the garbage collection task executed by each garbage treatment device, wherein an evaluation formula is as follows:
F(R,O,E) = (α∑R+β∑O-γE)×δT，
wherein R is the operation risk degree, O is the sea wave interference degree, E is the energy consumption of the garbage disposal device, and T is the total time for completing the garbage recycling task; alpha, beta, gamma and delta are weight coefficients; the operation risk degree R and the sea wave interference degree O are obtained discretely;
setting each parameter in the fuzzy relation matrix as an influence factor; the parameters comprise current moving speed, sea wave interference degree, current operation speed and operation risk degree;
when the recorded garbage collection task data is less than 1000 pieces, optimizing parameter setting by using a genetic algorithm, an ant colony algorithm or a simulated annealing algorithm;
and when the recorded garbage collection task data is not less than 1000 pieces, optimizing parameter setting by using a neural network algorithm.
9. The machine vision-based sea-land two-domain garbage collection planning method according to claim 8, wherein the optimizing parameter setting using genetic algorithm is specifically:
initializing a genetic algorithm, taking the most completed garbage collection tasks as a main target, taking the lowest operation risk degree, the lowest sea wave interference degree and the least energy consumption of a garbage treatment device as secondary targets, and constructing an objective function; the objective function is expressed as:
Q = K 1 ×n/N+K 2 ×F(R,O,E)，
wherein N is the number of times that a certain garbage disposal device completes the garbage collection task, and N represents the number of times that all garbage disposal devices complete the garbage collection task; k (K) 1 、K 2 Is a weight coefficient;
using genetic algorithm to carry out coding, fitness function calculation, selection operation, crossover operation and mutation operation to output an optimal population;
and decoding the optimal population to obtain optimal parameters.
10. The machine vision-based sea-land two-domain garbage collection planning method according to claim 8, wherein the optimizing parameter setting using the neural network algorithm is specifically:
the neural network is selected as a model of a neural network algorithm, the task of garbage collection is completed as a main target, the operation risk degree is minimum, the sea wave interference degree is minimum, the garbage treatment device energy consumption is minimum as a secondary target, and the quality evaluation is used as a model output label;
dividing recorded garbage collection task data according to a proportion to obtain a training set, a testing set and a verification set;
training, testing and verifying the model of the neural network algorithm to finally obtain the optimal parameters.
CN202311797667.4A 2023-12-26 2023-12-26 Sea and land two-domain garbage recycling planning method based on machine vision Active CN117456395B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311797667.4A CN117456395B (en) 2023-12-26 2023-12-26 Sea and land two-domain garbage recycling planning method based on machine vision

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311797667.4A CN117456395B (en) 2023-12-26 2023-12-26 Sea and land two-domain garbage recycling planning method based on machine vision

Publications (2)

Publication Number Publication Date
CN117456395A CN117456395A (en) 2024-01-26
CN117456395B true CN117456395B (en) 2024-03-29

Family

ID=89595189

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311797667.4A Active CN117456395B (en) 2023-12-26 2023-12-26 Sea and land two-domain garbage recycling planning method based on machine vision

Country Status (1)

Country Link
CN (1) CN117456395B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117893204B (en) * 2024-03-18 2024-05-10 泸州市兴泸环保发展有限公司 Data-driven garbage collection prediction method and device and electronic equipment

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107248014A (en) * 2017-06-27 2017-10-13 安徽师范大学 Intelligent garbage based on quantum cuckoo searching algorithm reclaims paths planning method
CN108104089A (en) * 2018-01-18 2018-06-01 蔡璟 A kind of removing type garbage recovering apparatus waterborne and its method of work
CN113985876A (en) * 2021-10-27 2022-01-28 广州大学 Marine garbage recycling path planning method and system based on foraging of bionic fishes
CN115454081A (en) * 2022-09-20 2022-12-09 鹏城实验室 Ocean garbage recycling method, unmanned boat, unmanned aerial vehicle and system
CN115933715A (en) * 2023-01-10 2023-04-07 莆田学院 Marine garbage cleaning robot and path planning method, device and medium thereof
CN116152115A (en) * 2023-04-04 2023-05-23 湖南融城环保科技有限公司 Garbage image denoising processing method based on computer vision

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20210214055A1 (en) * 2019-11-27 2021-07-15 Boris Maydanik Ocean cleanup autonomous system (ocas)

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107248014A (en) * 2017-06-27 2017-10-13 安徽师范大学 Intelligent garbage based on quantum cuckoo searching algorithm reclaims paths planning method
CN108104089A (en) * 2018-01-18 2018-06-01 蔡璟 A kind of removing type garbage recovering apparatus waterborne and its method of work
CN113985876A (en) * 2021-10-27 2022-01-28 广州大学 Marine garbage recycling path planning method and system based on foraging of bionic fishes
CN115454081A (en) * 2022-09-20 2022-12-09 鹏城实验室 Ocean garbage recycling method, unmanned boat, unmanned aerial vehicle and system
CN115933715A (en) * 2023-01-10 2023-04-07 莆田学院 Marine garbage cleaning robot and path planning method, device and medium thereof
CN116152115A (en) * 2023-04-04 2023-05-23 湖南融城环保科技有限公司 Garbage image denoising processing method based on computer vision

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
基于树莓派的小型水域智能垃圾清理机器人系统设计与实现;汤伟 等;科学技术与工程;20191208(第34期);第242-250页 *
基于流体涡旋向心吸力的改进型水面垃圾回收装置;郑烁桐 等;《科技创新与应用》;20211231(第15期);第11-14页 *

Also Published As

Publication number Publication date
CN117456395A (en) 2024-01-26

Similar Documents

Publication Publication Date Title
CN117456395B (en) Sea and land two-domain garbage recycling planning method based on machine vision
CN109784190A (en) A kind of automatic Pilot scene common-denominator target Detection and Extraction method based on deep learning
CN109325960B (en) Infrared cloud chart cyclone analysis method and analysis system
CN109766884A (en) A kind of airfield runway foreign matter detecting method based on Faster-RCNN
CN106874954A (en) The method and relevant apparatus of a kind of acquisition of information
CN109346103B (en) Audio detection method for road tunnel traffic incident
CN109543615B (en) Double-learning-model target tracking method based on multi-level features
CN111723657B (en) River foreign matter detection method and device based on YOLOv3 and self-optimization
CN116310785B (en) Unmanned aerial vehicle image pavement disease detection method based on YOLO v4
CN109876569B (en) Online detection device, system and method for purified gas index of bag-type dust remover
CN110119747A (en) A kind of coal rock detection method based on radioscopic image
CN108074014A (en) A kind of method for the phase fitful wind that detects a typhoon
CN111186656A (en) Target garbage classification method and intelligent garbage can
CN110689051A (en) Intelligent identification method for corrosion mode in gas pipeline based on transfer learning
CN112591333A (en) Automatic garbage classification device and method based on artificial intelligence
CN115240069A (en) Real-time obstacle detection method in full-fog scene
CN115862073A (en) Transformer substation harmful bird species target detection and identification method based on machine vision
CN110516625A (en) A kind of method, system, terminal and the storage medium of rubbish identification classification
CN105303162A (en) Target proposed algorithm-based insulator recognition algorithm for aerial images
CN113989594A (en) Tomato picking device based on shelf life prediction and maturity judgment and use method thereof
CN114882410B (en) Tunnel dome lamp fault detection method and system based on improved positioning loss function
Yang et al. Research on subway pedestrian detection algorithms based on SSD model
CN111767874A (en) Pavement disease detection method based on deep learning
CN114049134A (en) Pollution source online monitoring data counterfeiting identification method
CN117274212A (en) Bridge underwater structure crack detection method

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