CN114565635B - Unmanned ship system capable of intelligently identifying river channel garbage and performing classified collection - Google Patents
Unmanned ship system capable of intelligently identifying river channel garbage and performing classified collection Download PDFInfo
- Publication number
- CN114565635B CN114565635B CN202210218162.7A CN202210218162A CN114565635B CN 114565635 B CN114565635 B CN 114565635B CN 202210218162 A CN202210218162 A CN 202210218162A CN 114565635 B CN114565635 B CN 114565635B
- Authority
- CN
- China
- Prior art keywords
- garbage
- area
- cleaning
- point
- unmanned ship
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Human Resources & Organizations (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Economics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Game Theory and Decision Science (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Development Economics (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an unmanned ship system for intelligently identifying and classifying and collecting river channel garbage, which belongs to the technical field of river channel garbage cleaning and comprises an area division module, a garbage identification module, a route planning module, a recycling and classifying module and a server; a river channel garbage cleaning area is divided through an area dividing module, and a garbage identifying module identifies garbage in the garbage cleaning area; the route planning module plans a cleaning route of the river channel garbage, acquires the central coordinate of a garbage image coordinate area in real time, marks the central coordinate as a central point, establishes a central point dynamic distribution map according to the central point acquired in real time, merges representative areas of the central point in the dynamic distribution map to obtain a dynamic garbage merging map, marks the coordinate of the unmanned ship, calculates the shortest path of the unmanned ship for cleaning all the garbage based on an ant colony algorithm, and marks the corresponding shortest path as a whole-course planning path; the intelligent cleaning of river channel rubbish is realized, so that the cleaning of the river channel rubbish is faster and safer.
Description
Technical Field
The invention belongs to the technical field of river channel garbage cleaning, and particularly relates to an unmanned ship system for intelligently identifying and classifying and collecting river channel garbage.
Background
Garbage such as vegetable leaves, plastic bags, beverage bottles, tree branches and leaves, packing boxes, clothes and the like can float on the water surface of a river in urban and rural areas, and the garbage on the water surface is artificial and is also generated by factors such as wind and rain in nature. The garbage can have great influence on the environment and the water quality of the river channel. In order to beautify the river channel environment and clean the water body, the garbage in the river channel needs to be cleaned; however, when garbage is cleaned in a river channel, a manual salvage mode is mostly adopted, so that the cost is high, the efficiency is low, workers are always in danger of falling into water, and therefore an unmanned ship system for intelligently identifying and classifying and collecting the garbage in the river channel is needed at present, and the problem of cleaning the garbage in the river channel is solved.
Disclosure of Invention
In order to solve the problems existing in the scheme, the invention provides an unmanned ship system for intelligently identifying and classifying and collecting river channel garbage.
The purpose of the invention can be realized by the following technical scheme:
an unmanned ship system for intelligently identifying and classifying and collecting river channel garbage comprises an area division module, a garbage identification module, a route planning module, a recycling and classifying module and a server;
dividing a river channel garbage cleaning area through an area dividing module, and identifying garbage in the garbage cleaning area through a garbage identification module; the route planning module plans a cleaning route of the river channel garbage, acquires the central coordinate of a garbage image coordinate area in real time, marks the central coordinate as a central point, establishes a central point dynamic distribution map according to the central point acquired in real time, merges representative areas of the central point in the dynamic distribution map to obtain a dynamic garbage merging map, marks the coordinate of the unmanned ship, calculates the shortest paths of the unmanned ship for cleaning all garbage based on an ant colony algorithm, and marks the corresponding shortest paths as a whole-course planning path; acquiring a region starting point of the unmanned ship entering each garbage combination region and a region ending point of the unmanned ship leaving the corresponding garbage combination region according to the whole-course planning path; acquiring the cleaning width of the unmanned ship, and setting a cleaning route extending over the garbage merging area according to the cleaning width, the area starting point and the area terminal point;
then classifying the garbage recycled by the unmanned ship through a recycling and classifying module;
the method for performing representative region merging on the central point in the dynamic garbage merging map comprises the following steps:
step SA1: setting a maximum combination radius and a maximum distance between two central points, and marking the maximum distance between the two central points as a distribution distance; optionally selecting one central point from all the central points as a p point;
step SA2: calculating the distances from the point p to all the central points in the dynamic garbage combination graph in real time, and marking as the calculated distances; all center points for which the calculated distance is less than the distribution distance are labeled as p 1 Forming a class, determining the class center of the class, and calculating the class radius L of the class according to the class center 1 ；
Step SA3: calculating p in real time 1 Distances from the point to all the remaining center points, and the center points with the distances smaller than the distribution distance are marked as p 2 Point, p 1 Point sum p 2 Forming a new class by the points, determining the class center of the class, and calculating the class radius L of the class according to the class center 2 And so on until the label p i Point, p i The point represents the center point of the ith calculation; obtaining a class radius L i Wherein i =1,2,3，……，n；
class radius L i And when the combined radius is not less than the maximum combined radius, combining the central points, and performing garbage coordinate area replacement on the central points in the combined area.
The method for completing the center point combination in the step SA3 comprises the following steps:
when the class radius L i When the radius is equal to the maximum merging radius, the merging of the central points is completed;
when class radius L i When the maximum merging radius is larger than the maximum merging radius, removing the p farthest from the center of the category in the category 2 Point, recalculating the category center and the category radius of the category, and iterating until the p farthest from the category center is removed 2 After point counting, when the calculated category radius is not larger than the maximum combination radius, the combination of the central points is completed;
the method for setting the cleaning route extending over the garbage combination area according to the cleaning width, the area starting point and the area ending point comprises the following steps:
the area and the boundary contour of each garbage coordinate area in the dynamic garbage combination map are identified, the reciprocating times and the corresponding reciprocating times routes for cleaning the garbage in the garbage coordinate areas by the unmanned ship are calculated, the starting points and the end points corresponding to the reciprocating times are obtained and marked as the reciprocating starting points and the reciprocating end points, the reciprocating starting points and the reciprocating end points are marked at the corresponding positions in the dynamic garbage combination map, the area starting points, the area end points, the reciprocating starting points, the reciprocating end points and the reciprocating times routes in the dynamic garbage combination map are used as the necessary routes, the shortest route of the unmanned ship passing the necessary routes is calculated based on an ant colony algorithm and marked as the cleaning route.
Further, the working method of the region division module comprises the following steps:
acquiring a river channel map needing garbage cleaning, identifying longitude and latitude coordinates of a river channel boundary, acquiring the model of an unmanned ship, setting buffer areas in the river channel map according to the model of the unmanned ship, and marking areas among the buffer areas in the river channel map as action areas;
the method comprises the steps of setting an area unit, obtaining unmanned ship coordinates, updating the obtained unmanned ship coordinates in a river map, obtaining the river map in a range corresponding to the unmanned ship coordinates, framing a garbage cleaning area in the river map by an operator through the area unit, identifying an area boundary framed by the operator, integrating the identified framing area boundary and an action area boundary, and marking the integrated area as a garbage cleaning area.
Further, the working method of the garbage identification module comprises the following steps:
acquiring images of the garbage cleaning areas, establishing a grid graph in the images of the garbage cleaning areas, and numbering the images of the garbage cleaning areas corresponding to each grid; segmenting the garbage cleaning area image according to grids in the grid map to obtain a plurality of segmented images;
acquiring a water surface image when the river channel is free of garbage, carrying out size processing on the acquired water surface image, and marking the processed image as a standard image; calculating the similarity between all the segmented images and the standard image, judging whether the calculated similarity meets the similarity requirement, and marking the corresponding segmented images as normal images when the calculated similarity meets the similarity requirement; and when the calculated similarity value is judged not to meet the similarity requirement, marking the corresponding segmentation image as a garbage image, and acquiring a coordinate area corresponding to the garbage image.
Further, the method for acquiring the image of the garbage cleaning area comprises the following steps:
the method comprises the steps of identifying boundary coordinates of a garbage cleaning area, inputting the acquired boundary coordinates of the garbage cleaning area into an external image acquisition device, carrying out image acquisition on the garbage cleaning area through the external image acquisition device, identifying coordinate boundaries in an acquired image, and segmenting the acquired image according to the identified coordinate boundaries to obtain an image of the garbage cleaning area.
Further, the size of each grid in the grid map is the same.
Further, the working method of the recycling classification module comprises the following steps:
acquiring detection images of the recovered garbage, training a deep learning neural network, processing a first detection image through the deep learning neural network to obtain a detection result, generating a first control signal when the detection result is matched, and sending the corresponding recovered garbage into a recovery area; and when the detection result is not matched, generating a second control signal, and sending the corresponding recovered garbage into the non-recovery area.
Compared with the prior art, the invention has the beneficial effects that: the intelligent cleaning of the river channel garbage is realized, the human resources are greatly liberated, and the cleaning of the river channel garbage is quicker and safer; by arranging the region division module, an operator can conveniently divide garbage cleaning regions and data support is provided for subsequent garbage identification; by setting the segmentation image and calculating the similarity between the segmentation image and the standard image, the position of the rubbish can be conveniently and rapidly identified, and the subsequent planning of a river rubbish cleaning route is facilitated.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a schematic block diagram of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the following embodiments, and it should be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
As shown in fig. 1, an unmanned ship system for intelligently identifying and classifying and collecting river channel garbage comprises an area division module, a garbage identification module, a route planning module, a recovery classification module and a server;
the region division module is used for dividing a river channel garbage cleaning region, and the specific method comprises the following steps:
acquiring a river map needing garbage cleaning, and identifying longitude and latitude coordinates of a river boundary, wherein the river boundary is a river bank or a bridge not meeting traffic conditions; the method comprises the steps of obtaining the model of the unmanned ship, setting a buffer area in a river map according to the model of the unmanned ship, namely setting the buffer area at a distance of a plurality of meters from a river bank according to the model of the unmanned ship, and avoiding the unmanned ship from being stranded; marking the area between the buffer areas in the river map as an action area;
the method comprises the steps that a region unit is arranged, and the region unit is used for dividing garbage cleaning regions by operators, obtaining unmanned ship coordinates, updating the obtained unmanned ship coordinates in a river map, and obtaining the river map of a range corresponding to the unmanned ship coordinates; an operator frames a garbage cleaning area in a river map through an area unit, identifies the framed area boundary of the operator, integrates the identified framed area boundary with an action area boundary, and marks the integrated area as the garbage cleaning area.
In one embodiment, a method for integrating identified box selection area boundaries with action zone boundaries comprises:
when the border of the frame selection area comprises the action area border, taking an area formed by the action area border and the other two frame selection area borders as an integrated area; when the frame selection area boundary does not include an action area boundary, overlapping the frame selection area boundary in the same direction as the action area boundary with the action area boundary to form an integrated area; it is equivalent to that the frame selection area boundary is only used for determining the length of the integration area, and the action area boundary is used as the width of the integration area.
In one embodiment, another method of integrating identified box region boundaries with action zone boundaries includes:
when the width of the river channel is especially wide and exceeds a preset value, the frame selection area is directly used as an integration area.
The garbage recognition module is used for recognizing garbage in a garbage cleaning area, and the specific method comprises the following steps:
recognizing boundary coordinates of a garbage cleaning area, inputting the acquired boundary coordinates of the garbage cleaning area into an external image acquisition device, wherein the external image acquisition device is an existing image acquisition device with coordinates, acquiring images of the garbage cleaning area through the external image acquisition device, recognizing coordinate boundaries in the acquired images, segmenting the acquired images according to the recognized coordinate boundaries to obtain images of the garbage cleaning area, establishing a grid graph in the images of the garbage cleaning area, and enabling the size of each grid in the grid graph to be the same; the grid size is set by an expert group according to the image of the garbage cleaning area and the size of the unmanned ship recovery area; numbering the images of the garbage cleaning area corresponding to each grid; segmenting the garbage cleaning area image according to grids in the grid map to obtain a plurality of segmented images;
acquiring a water surface image when the river channel is free of rubbish, and performing size processing on the acquired water surface image, wherein the size processing is to cut the water surface image into an image with a standard size; the standard size is the grid size; marking the processed image as a standard image; calculating the similarity between all the segmentation images and the standard image, wherein the similarity between the segmentation images and the standard image can be calculated by using the conventional image similarity calculation method; judging whether the calculated similarity value meets the similarity requirement or not, wherein the similarity requirement is discussed and set by an expert group; when the calculated similarity value is judged to meet the similarity requirement, marking the corresponding segmented image as a normal image; when the calculated similarity value is judged not to meet the similarity requirement, marking the corresponding segmentation image as a garbage image, and acquiring a coordinate area corresponding to the garbage image; the garbage image is that the image has garbage.
The route planning module is used for planning a river channel garbage cleaning route, and the specific method comprises the following steps:
acquiring the central coordinate of a garbage image coordinate area in real time, and marking the central coordinate as a central point, wherein the river garbage can move due to the influence of various aspects such as water flow, air flow and the fluctuation of a garbage cleaning process, so that the central point position needs to be acquired and updated in real time; establishing a central point dynamic distribution diagram according to a central point acquired in real time, merging representative areas of the central point in the dynamic distribution diagram to obtain a dynamic garbage merging diagram, marking coordinates of the unmanned ship, calculating shortest paths for cleaning all garbage of the unmanned ship based on an ant colony algorithm, and marking the corresponding shortest paths as a whole-course planning path; acquiring a region starting point of the unmanned ship entering each garbage combination region and a region ending point of the unmanned ship leaving the corresponding garbage combination region according to the whole-course planning path; and acquiring the cleaning width of the unmanned ship, and setting a cleaning route extending all over the garbage merging area according to the cleaning width, the area starting point and the area ending point.
The method for performing representative region merging on the central point in the dynamic garbage merging map comprises the following steps:
step SA1: setting a maximum combination radius and a maximum distance between two central points, and marking the maximum distance between the two central points as a distribution distance; optionally selecting one central point from all the central points as a p point;
step SA2: calculating the distances from the point p to all the central points in the dynamic garbage combination graph in real time, and marking as the calculated distances; all center points for which the calculated distance is less than the distribution distance are labeled as p 1 Forming a class, determining the class center of the class, and calculating the class radius L of the class according to the class center 1 ；
Step SA3: computing p in real time 1 Distances from the point to all the remaining center points, and the center points with the distances smaller than the distribution distance are marked as p 2 Point, p 1 Point sum p 2 Forming a new class by the points, determining the class center of the class, and calculating the class radius L of the class according to the class center 2 And so on until the label p i Point, p i The point represents the center point of the ith calculation; obtain the class radius L i Wherein i =1,2,3, \8230;, n;
class radius L i And when the center point is not smaller than the maximum combination radius, the combination of the center points is completed, and the garbage coordinate area replacement is carried out on the center point in the combination area.
The method for completing the center point combination in the step SA3 comprises the following steps:
when class radius L i When the radius is equal to the maximum merging radius, the center is completedPoint merging;
when the class radius L i When the maximum merging radius is larger than the maximum merging radius, removing the p farthest from the center of the category in the category 2 Point, re-calculate the class center and class radius of the class, iterate until the p farthest from the class center is removed 2 And after point counting, when the calculated category radius is not larger than the maximum combination radius, the center point combination is completed.
Exemplary, the method for calculating the shortest path for cleaning all garbage of the unmanned ship based on the ant colony algorithm includes:
step SB1: modeling a map environment by using a grid method, initializing basic parameters of an ant colony algorithm, and initializing an ant colony at an initial node;
step SB2: each ant starts to search and traverse from an initial node, a next node is selected according to the transition probability, each node traveled by the ant is recorded in a taboo table, and when the ant reaches a target node, the length of a path selected by the ant and the number of inflection points on the path are calculated; after all ants finish searching, selecting the optimal path of the current iteration by applying a principle of less inflection points; applying a simplified path principle to the optimal path of the current iteration to obtain an optimized optimal path of the current iteration; applying a pre-sorting rule to all feasible paths of the current iteration, selecting a better path, carrying out a self-adaptive adjustment strategy on pheromone volatilization coefficients on the better path, and updating the pheromone concentration on each better path on the basis;
step SB3: and (5) repeatedly executing the step SB2 until the iteration times reach the maximum iteration times, and calculating the optimal path length of each iteration to obtain a global optimal path.
The application of ant colony algorithm is common, and the technology which is not disclosed in the method is the prior art.
The method for setting the cleaning route extending over the garbage combination area according to the cleaning width, the area starting point and the area ending point comprises the following steps:
the area and the boundary contour of each garbage coordinate area in the dynamic garbage combination graph are identified, the reciprocating times and the corresponding reciprocating time routes of the unmanned ship for cleaning the garbage in the garbage coordinate area are calculated, the starting point and the end point corresponding to the reciprocating times are obtained and marked as the reciprocating starting point and the reciprocating end point, the reciprocating starting point and the reciprocating end point are marked at the corresponding positions in the dynamic garbage combination graph, the area starting point, the area end point, the reciprocating starting point, the reciprocating end point and the reciprocating time routes in the dynamic garbage combination graph are used as the must-pass routes, the shortest path of the unmanned ship passing the must-pass routes is calculated based on an ant colony algorithm and marked as the cleaning routes.
The method for calculating the reciprocating times of the unmanned ship for cleaning the garbage in the garbage coordinate area comprises the following steps:
establishing an analysis model based on a CNN network or a DNN network, wherein a training set comprises an area, a boundary outline, a cleaning width, and reciprocating times and reciprocating time routes which are correspondingly set, and analyzing the reciprocating times, the reciprocating routes and starting points and end points of the corresponding reciprocating times of each garbage coordinate area through the analysis model, wherein the obtained reciprocating times are the minimum times.
The recycling and classifying module is used for classifying the garbage recycled by the unmanned ship, and the specific method comprises the following steps:
acquiring detection images of the recovered garbage, training a deep learning neural network, processing a first detection image through the deep learning neural network to obtain a detection result, generating a first control signal when the detection results are matched, and sending the corresponding recovered garbage into a recovery area; and when the detection result is not matched, generating a second control signal, and sending the corresponding recovered garbage into the non-recovery area.
In one embodiment, the training process of the deep learning neural network comprises the following operations:
acquiring training images capable of recycling garbage in a sample image library; preprocessing the training image, wherein the preprocessing can eliminate irrelevant information in the training image; randomly selecting a training image and initializing the training image, wherein the training image can be converted into a data signal which can be processed by a deep learning neural network by initializing the training image so as to facilitate subsequent operation;
performing convolution and sampling processing on the training image for multiple times, wherein the convolution processing can extract the characteristics of the training image, and the sampling processing can reduce the scale of training data and reduce the calculated amount; the last sampled layer or convolutional layer is connected to one or more fully-connected layers. The full connection layer is configured to synthesize the features of the recyclable garbage extracted after the convolution and sampling processing and output training parameters and feature models of the recyclable garbage. The feature model is an abstract feature representation of the recyclable garbage.
Judging the feature model output by the full connection layer, and outputting the feature model when the feature model is matched with a preset standard feature model;
otherwise, the adjusting weight matrix is propagated reversely; in the training process, if an error exists between the output characteristic model and the standard characteristic model, the error information is reversely transmitted along the original path through reverse propagation so as to correct the training parameters of each layer, wherein the training parameters can comprise weighted values and bias, and then the training images are subjected to convolution and sampling again by using the corrected convolution layer and sampling layer until the characteristic model meets the end condition.
Comparing the characteristic model of the garbage to be classified with the trained characteristic model of the recyclable garbage, and judging whether the garbage to be classified is matched with the recyclable garbage or not; outputting a detection result; the detection results include a match and a mismatch.
In one embodiment, the deep learning neural network may directly use an existing already trained garbage classification model.
Although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the present invention.
Claims (6)
1. An unmanned ship system for intelligently identifying and classifying and collecting river channel garbage is characterized by comprising an area division module, a garbage identification module, a route planning module, a recovery classification module and a server;
dividing a river channel garbage cleaning area through an area dividing module, and identifying garbage in the garbage cleaning area through a garbage identification module; the route planning module plans a cleaning route of the river channel garbage, acquires the central coordinate of a garbage image coordinate area in real time, marks the central coordinate as a central point, establishes a central point dynamic distribution map according to the central point acquired in real time, merges representative areas of the central point in the dynamic distribution map to obtain a dynamic garbage merging map, marks the coordinate of the unmanned ship, calculates the shortest paths of the unmanned ship for cleaning all garbage based on an ant colony algorithm, and marks the corresponding shortest paths as a whole-course planning path; acquiring a region starting point of the unmanned ship entering each garbage combination region and a region ending point of the unmanned ship leaving the corresponding garbage combination region according to the whole-course planning path; acquiring the cleaning width of the unmanned ship, and setting a cleaning route extending over the garbage merging area according to the cleaning width, the area starting point and the area terminal point;
then classifying the garbage recycled by the unmanned ship through a recycling and classifying module;
the method for performing representative region merging on the central point in the dynamic garbage merging graph comprises the following steps:
step SA1: setting a maximum combination radius and a maximum distance between two central points, and marking the maximum distance between the two central points as a distribution distance; optionally selecting one central point from all the central points as a p point;
step SA2: calculating the distances from the point p to all the central points in the dynamic garbage combination graph in real time, and marking as the calculated distances; all center points for which the calculated distance is less than the distribution distance are labeled as p 1 Forming a class, determining the class center of the class, and calculating the class radius L of the class according to the class center 1 ；
Step SA3: computing p in real time 1 Distances from the point to all the remaining center points, and the center points with the distances smaller than the distribution distance are marked as p 2 Point, p 1 Point sum p 2 Forming a new category, determining the category center of the category, and calculating the category radius L of the category according to the category center 2 And so on until the label p i Point, p i The point represents the center point of the ith calculation; obtaining a class radius L i Wherein i =1,2,3, \8230;, n;
class radius L i When the center point is not smaller than the maximum merging radius, merging the center points, and performing garbage coordinate area replacement on the center points in the merging area;
the method for completing the center point combination in the step SA3 comprises the following steps:
when class radius L i When the radius is equal to the maximum merging radius, the merging of the central points is completed;
when class radius L i When the maximum merging radius is larger than the maximum merging radius, removing the p farthest from the center of the category in the category 2 Point, recalculating the category center and the category radius of the category, and iterating until the p farthest from the category center is removed 2 After point counting, when the calculated category radius is not larger than the maximum combination radius, the combination of the central points is completed;
the method for setting the cleaning route extending over the garbage combination area according to the cleaning width, the area starting point and the area ending point comprises the following steps:
the area and the boundary contour of each garbage coordinate area in the dynamic garbage combination graph are identified, the reciprocating times and the corresponding reciprocating time routes of the unmanned ship for cleaning the garbage in the garbage coordinate area are calculated, the starting point and the end point corresponding to the reciprocating times are obtained and marked as the reciprocating starting point and the reciprocating end point, the reciprocating starting point and the reciprocating end point are marked at the corresponding positions in the dynamic garbage combination graph, the area starting point, the area end point, the reciprocating starting point, the reciprocating end point and the reciprocating time routes in the dynamic garbage combination graph are used as the must-pass routes, the shortest path of the unmanned ship passing the must-pass routes is calculated based on an ant colony algorithm and marked as the cleaning routes.
2. The unmanned ship system for intelligently identifying and classifying and collecting river channel garbage according to claim 1, wherein the working method of the region division module comprises the following steps:
acquiring a river channel map needing garbage cleaning, identifying longitude and latitude coordinates of a river channel boundary, acquiring the model of an unmanned ship, setting buffer areas in the river channel map according to the model of the unmanned ship, and marking areas among the buffer areas in the river channel map as action areas;
the method comprises the steps of setting an area unit, obtaining unmanned ship coordinates, updating the obtained unmanned ship coordinates in a river map, obtaining the river map in a range corresponding to the unmanned ship coordinates, framing a garbage cleaning area in the river map by an operator through the area unit, identifying an area boundary framed by the operator, integrating the identified framing area boundary and an action area boundary, and marking the integrated area as a garbage cleaning area.
3. The unmanned ship system for intelligently identifying and classifying and collecting river channel garbage according to claim 1, wherein a working method of the garbage identification module comprises the following steps:
acquiring images of the garbage cleaning areas, establishing a grid graph in the images of the garbage cleaning areas, and numbering the images of the garbage cleaning areas corresponding to each grid; segmenting the garbage cleaning area image according to grids in the grid map to obtain a plurality of segmented images;
acquiring a water surface image when the river channel is free of garbage, carrying out size processing on the acquired water surface image, and marking the processed image as a standard image; calculating the similarity between all the segmented images and the standard image, judging whether the calculated similarity meets the similarity requirement, and marking the corresponding segmented images as normal images when the calculated similarity meets the similarity requirement; and when the calculated similarity value is judged not to meet the similarity requirement, marking the corresponding segmentation image as a garbage image, and acquiring a coordinate area corresponding to the garbage image.
4. The unmanned ship system for intelligently identifying and classifying and collecting river channel garbage according to claim 3, wherein the method for acquiring the garbage clearing area image comprises the following steps:
the method comprises the steps of identifying boundary coordinates of a garbage cleaning area, inputting the acquired boundary coordinates of the garbage cleaning area into an external image acquisition device, carrying out image acquisition on the garbage cleaning area through the external image acquisition device, identifying coordinate boundaries in an acquired image, and segmenting the acquired image according to the identified coordinate boundaries to obtain an image of the garbage cleaning area.
5. The unmanned ship system for intelligently identifying and classifying river channel garbage according to claim 3, wherein each grid in the grid map has the same size.
6. The unmanned ship system for intelligently identifying and classifying and collecting river channel garbage according to claim 1, wherein the working method of the recovery and classification module comprises the following steps:
acquiring detection images of the recovered garbage, training a deep learning neural network, processing a first detection image through the deep learning neural network to obtain a detection result, generating a first control signal when the detection result is matched, and sending the corresponding recovered garbage into a recovery area; and when the detection result is not matched, generating a second control signal, and sending the corresponding recovered garbage into the non-recovery area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210218162.7A CN114565635B (en) | 2022-03-08 | 2022-03-08 | Unmanned ship system capable of intelligently identifying river channel garbage and performing classified collection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210218162.7A CN114565635B (en) | 2022-03-08 | 2022-03-08 | Unmanned ship system capable of intelligently identifying river channel garbage and performing classified collection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114565635A CN114565635A (en) | 2022-05-31 |
| CN114565635B true CN114565635B (en) | 2022-11-11 |
Family
ID=81717582
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210218162.7A Active CN114565635B (en) | 2022-03-08 | 2022-03-08 | Unmanned ship system capable of intelligently identifying river channel garbage and performing classified collection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114565635B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115185280B (en) * | 2022-07-27 | 2023-02-24 | 柳州城市职业学院 | Unmanned ship control system and method based on path planning autonomous navigation |
| CN115291607B (en) * | 2022-08-02 | 2023-03-14 | 柳州城市职业学院 | Autonomous planning generation method and system for unmanned ship path in complex unknown water area |
| CN117577284B (en) * | 2024-01-17 | 2024-04-26 | 湖南洁城环保科技有限公司 | Medical waste intelligent collection vehicle traceability management method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106741683A (en) * | 2017-03-08 | 2017-05-31 | 中国矿业大学 | A kind of water area refuse automated cleaning ship and automatic cleaning method based on machine vision |
| CN107622231A (en) * | 2017-09-08 | 2018-01-23 | 内蒙古大学 | A kind of water day integrally intelligent floating object collection system and its collection method |
| CN109375633A (en) * | 2018-12-18 | 2019-02-22 | 河海大学常州校区 | River course clear up path planning system and method based on global state information |
| CN112102369A (en) * | 2020-09-11 | 2020-12-18 | 陕西欧卡电子智能科技有限公司 | Autonomous inspection method, device and equipment for water surface floating target and storage medium |
| CN112278170A (en) * | 2020-10-06 | 2021-01-29 | 武汉烽火凯卓科技有限公司 | Unmanned aerial vehicle and unmanned ship cooperative water area cleaning method, system, device and storage medium |
| CN113723371A (en) * | 2021-11-01 | 2021-11-30 | 陕西欧卡电子智能科技有限公司 | Unmanned ship cleaning route planning method and device, computer equipment and storage medium |
| WO2022033484A1 (en) * | 2020-08-10 | 2022-02-17 | 北京智行者科技有限公司 | Garbage recognition method and system for intelligent sweeper |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111126116A (en) * | 2018-11-01 | 2020-05-08 | 中国移动通信集团浙江有限公司 | Unmanned ship river channel garbage identification method and system |
| CN111259809B (en) * | 2020-01-17 | 2021-08-17 | 五邑大学 | Unmanned aerial vehicle coastline floating garbage inspection system based on DANet |
-
2022
- 2022-03-08 CN CN202210218162.7A patent/CN114565635B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106741683A (en) * | 2017-03-08 | 2017-05-31 | 中国矿业大学 | A kind of water area refuse automated cleaning ship and automatic cleaning method based on machine vision |
| CN107622231A (en) * | 2017-09-08 | 2018-01-23 | 内蒙古大学 | A kind of water day integrally intelligent floating object collection system and its collection method |
| CN109375633A (en) * | 2018-12-18 | 2019-02-22 | 河海大学常州校区 | River course clear up path planning system and method based on global state information |
| WO2022033484A1 (en) * | 2020-08-10 | 2022-02-17 | 北京智行者科技有限公司 | Garbage recognition method and system for intelligent sweeper |
| CN112102369A (en) * | 2020-09-11 | 2020-12-18 | 陕西欧卡电子智能科技有限公司 | Autonomous inspection method, device and equipment for water surface floating target and storage medium |
| CN112278170A (en) * | 2020-10-06 | 2021-01-29 | 武汉烽火凯卓科技有限公司 | Unmanned aerial vehicle and unmanned ship cooperative water area cleaning method, system, device and storage medium |
| CN113723371A (en) * | 2021-11-01 | 2021-11-30 | 陕西欧卡电子智能科技有限公司 | Unmanned ship cleaning route planning method and device, computer equipment and storage medium |
Non-Patent Citations (4)
| Title |
|---|
| Neurogenetic algorithm applied to Route Planning for Autonomous Mobile Robots;Diego Renan Bruno et al.;《2018 International Joint Conference on Neural Networks (IJCNN)》;20181231;全文 * |
| 基于分位数半径的动态K-means算法;程明畅等;《南京大学学报(自然科学)》;20180130(第01期);全文 * |
| 基于物联网通信技术的水质检测无人船;赖俊委等;《通信电源技术》;20180225(第02期);全文 * |
| 汉中市城市生活垃圾收运路线优化研究;唐欣;《价值工程》;20200208(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114565635A (en) | 2022-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114565635B (en) | Unmanned ship system capable of intelligently identifying river channel garbage and performing classified collection | |
| CN107609525B (en) | Remote sensing image target detection method for constructing convolutional neural network based on pruning strategy | |
| CN111553236B (en) | Road foreground image-based pavement disease target detection and example segmentation method | |
| CN108830188B (en) | Vehicle detection method based on deep learning | |
| CN111553387B (en) | Personnel target detection method based on Yolov3 | |
| CN109977812B (en) | Vehicle-mounted video target detection method based on deep learning | |
| CN110210475B (en) | License plate character image segmentation method based on non-binarization and edge detection | |
| CN112232371B (en) | American license plate recognition method based on YOLOv3 and text recognition | |
| CN111008337B (en) | Deep attention rumor identification method and device based on ternary characteristics | |
| CN108596038A (en) | Erythrocyte Recognition method in the excrement with neural network is cut in a kind of combining form credit | |
| CN109753853A (en) | One kind being completed at the same time pedestrian detection and pedestrian knows method for distinguishing again | |
| CN110032952A (en) | A kind of road boundary point detecting method based on deep learning | |
| CN114170511A (en) | Pavement crack disease identification method based on Cascade RCNN | |
| CN113361533A (en) | Segmentation identification method and system for overlapped and shielded target object | |
| CN113409252A (en) | Obstacle detection method for overhead transmission line inspection robot | |
| CN115719475A (en) | Three-stage trackside equipment fault automatic detection method based on deep learning | |
| CN113903173B (en) | Vehicle track feature extraction method based on directed graph structure and LSTM | |
| CN114511627A (en) | Target fruit positioning and dividing method and system | |
| CN115858840B (en) | Remote sensing image mosaic method based on scene | |
| CN111523453A (en) | Intelligent road surface cleanliness judging method based on machine vision | |
| CN112085101A (en) | High-performance and high-reliability environment fusion sensing method and system | |
| CN115830302A (en) | Multi-scale feature extraction and fusion power distribution network equipment positioning identification method | |
| Rajeswari et al. | Automatic road extraction based on normalized cuts and level set methods | |
| CN111311643B (en) | Video target tracking method using dynamic search | |
| Liu et al. | Peaks Fusion assisted Early-stopping Strategy for Overhead Imagery Segmentation with Noisy Labels |
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 | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: An Intelligent Unmanned Ship System for Identifying and Classifying River Garbage Effective date of registration: 20230821 Granted publication date: 20221111 Pledgee: Agricultural Bank of China Limited Hefei Baohe District Branch Pledgor: Anhui Xinyu Environmental Protection Technology Co.,Ltd. Registration number: Y2023980053078 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |