CN113680695A - Robot-based machine vision garbage sorting system - Google Patents

Robot-based machine vision garbage sorting system Download PDF

Info

Publication number
CN113680695A
CN113680695A CN202110976991.7A CN202110976991A CN113680695A CN 113680695 A CN113680695 A CN 113680695A CN 202110976991 A CN202110976991 A CN 202110976991A CN 113680695 A CN113680695 A CN 113680695A
Authority
CN
China
Prior art keywords
garbage
image
robot
joint
gray
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110976991.7A
Other languages
Chinese (zh)
Inventor
梅志敏
江晓明
张融
舒慧
毕思
程书阳
陈则璇
崔月阳
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.)
Wuchang Institute of Technology
Original Assignee
Wuchang Institute of Technology
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 Wuchang Institute of Technology filed Critical Wuchang Institute of Technology
Priority to CN202110976991.7A priority Critical patent/CN113680695A/en
Publication of CN113680695A publication Critical patent/CN113680695A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/34—Sorting according to other particular properties
    • B07C5/3412—Sorting according to other particular properties according to a code applied to the object which indicates a property of the object, e.g. quality class, contents or incorrect indication
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/34—Sorting according to other particular properties
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/36—Sorting apparatus characterised by the means used for distribution
    • B07C5/361—Processing or control devices therefor, e.g. escort memory
    • B07C5/362—Separating or distributor mechanisms
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
    • B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
    • B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
    • B07C5/36—Sorting apparatus characterised by the means used for distribution
    • B07C5/38—Collecting or arranging articles in groups
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/2431—Multiple classes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
    • G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
    • G06K7/14—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation using light without selection of wavelength, e.g. sensing reflected white light
    • G06K7/1404—Methods for optical code recognition
    • G06K7/1408—Methods for optical code recognition the method being specifically adapted for the type of code
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20172—Image enhancement details
    • G06T2207/20192—Edge enhancement; Edge preservation

Abstract

The invention discloses a robot-based machine vision garbage sorting system, which comprises an IRB1410 robot workstation, a to-be-sorted garbage conveying device, a vision system, an MCGS human-computer interface and three types of garbage collection boxes, wherein the three types of garbage collection boxes are respectively a harmful garbage box, a recyclable garbage box and a dry garbage box; the sorting rate of the system reaches more than 97.0 percent, which shows that the system can reach higher accuracy and feasibility and has certain popularization value in practical engineering.

Description

Robot-based machine vision garbage sorting system
Technical Field
The invention belongs to the field of garbage classification.
Background
In recent years, the application of robot technology and equipment taking machine vision as a carrier in the intelligent manufacturing industry is more and more extensive, the working process is more automatic and intelligent, the working environment is improved, and the product precision is also improved; meanwhile, garbage classification is a comprehensive issue which puzzles people's lives and environmental governance for a long time, and needs to be solved through scientific and technical means. Most of garbage is caused after the consumption of circulation commodities, the outer package contains a bar code approved and issued by a Chinese coding center, the product name is obtained through visual recognition of the bar code, the product name is matched with a training set, a garbage classification result is output, data are transmitted to the tail end of a robot execution unit to realize sorting, and an important theory and application basis are provided for intelligent research of garbage classification.
Disclosure of Invention
The purpose of the invention is as follows: in order to overcome the defects in the prior art, the invention provides the robot-based machine vision garbage sorting system with high precision.
The technical scheme is as follows: in order to achieve the purpose, the robot-based machine vision garbage sorting system comprises an IRB1410 robot workstation, a to-be-sorted garbage conveying device, a vision system, an MCGS human-computer interface and three types of garbage collection boxes, wherein the three types of garbage collection boxes are respectively a harmful garbage box, a recyclable garbage box and a dry garbage box;
when the to-be-sorted garbage conveying device conveys the to-be-sorted commodity garbage to a visual detection platform acquisition area, a sensor sends an acquisition signal, a visual system acquires an outline image, a bar code image and position information, and the identified bar code number is matched with a commodity barcode database and a garbage classification training set; then, the pose and classification result information is transmitted to a robot controller, and then the pose information of the image is sent to the tail end of the robot through the controller; and solving the position and the posture of each joint according to the joint transformation matrix, solving parameters of each joint by adopting an inverse kinematics method, driving motors of each joint, calling a PROC program to grab the garbage to be sorted, putting the garbage to a corresponding garbage can, and adapting to the beat of the garbage to be sorted by setting the running speed of the robot.
Further, the inverse kinematics equation of the inverse kinematics method is as follows:
Figure BDA0003227695700000011
in the above formula:
Figure BDA0003227695700000021
representing a robot end pose matrix;
Figure BDA0003227695700000022
representing a pose transformation matrix from the joint i to the joint i + 1; n represents a vector in the x-axis direction; o represents a vector in the y-axis direction; a represents a vector in the z-axis direction; it is known that
Figure BDA0003227695700000023
Matrix and each link parameter, and solving the rotation angle theta 1-theta 6 of each joint link:
Figure BDA0003227695700000024
the joint angles were determined by analytical methods: θ 1, θ 2, θ 3, θ 4, θ 5, θ 6; after the pose of the garbage is obtained through visual positioning, the execution parameters of each joint realize sorting; in the above formula: a. theiRepresenting a pose transformation matrix of the joint i; rot represents a rotation operator; trans denotes the translation operator; the joint i represents a robot joint number; a isiIndicating a link length; alpha is alphaiRepresenting a connecting rod torsion angle; diIndicating a link offset; thetaiIndicating the link angle.
Furthermore, the vision system consists of an image acquisition device, an image processing system, a robot sorting workstation and a control communication transmission system; the image acquisition device consists of a light source, a lens and a camera, wherein the light source adopts annular white light, and the number of camera pixels is 1.3 ten thousand; the image processing system adopts a Halcon machine vision processing platform, converts the processing result to a Winform window of VS2010 through C # hybrid programming, and transmits the processing result to an MCGS interface through TCP/IP communication for displaying and counting; when the system works, when different types of circulating commodities pass through a photoelectric sensor, a conveyer belt triggers sensor signal output, a camera shoots in real time, an image is sent to a Halcon processing platform, appearance and bar code information on a package is extracted through processes of RGB gray level conversion, HSV conversion, threshold segmentation, connected domain closing operation, feature extraction, database matching decoding, display and the like, a solving algorithm is output to a VS2010 window in a C # format, the type is displayed and counted on a human-computer interface synchronously, and a robot finishes sorting;
before image acquisition, a light source is required to be adjusted to be consistent with a camera target area, the photographing effect due to the interference of a conveying belt is eliminated, and the photographing frame number is greater than the movement frequency of the conveying belt; converting the collected RGB image gray level into a gray level image and an HSV image, and displaying color discrimination and extracting characteristics; the characteristic extraction is to select an obvious gray image from the obtained HSV image, extract a contour and a bar code area after threshold segmentation and connected domain operation, compare the obtained numerical value with a known database and match the optimal value of a training set;
the grayscale conversion mechanism is as follows:
g(x，y)＝T[f(x,y)]
in the formula: f (x, y) represents a digital image to be processed; g (x, y) represents the processed digital image; t defines the operation of f;
threshold segmentation:
defining the pixel with the gray value i epsilon [0, L-1] as ni, then:
Figure BDA0003227695700000031
the probability of each gray value occurring is:
Figure BDA0003227695700000032
for piIs provided with
Figure BDA0003227695700000033
The image pixels are divided into two types of A and B by using a threshold value T, wherein A is composed of gray values in [0, T-1] pixels, B is composed in [ T, L-1] pixels, and the probability is respectively
Figure BDA0003227695700000034
Figure BDA0003227695700000035
The average gray scale of the whole image is represented by u, and the average gray scale of the region A, B to which the average gray scale belongs
Figure BDA0003227695700000036
Figure BDA0003227695700000037
Figure BDA0003227695700000038
The total variance of the regions is:
Figure BDA0003227695700000039
let T take the value in the interval L-1, make
Figure BDA00032276957000000310
The maximum T is the optimal region threshold value, and the image processing effect is optimal; the first-order differential property of the Gaussian function is utilized to convert the problem of edge detection into the problem of maximum value of the detection function, and compromise is obtained between noise suppression and edge detection;
Figure BDA0003227695700000041
G(x,y)＝f(x,y)*H(x，y)
in the formula: l-1 represents the maximum gray value; n represents a gray value; p represents a probability; u represents an average gray;
Figure BDA0003227695700000042
representing the total variance of the region; f denotes image data, H denotes a gaussian function with coefficients omitted, G denotes a filtered smoothed image, (x, y) denotes pixel point position coordinate values, and σ denotes standard deviation.
Has the advantages that: the invention optimizes the technical process of visual system image processing, including image acquisition, gray level conversion, threshold segmentation, flow domain, matching decoding and the like, adopts the Canny algorithm to improve the edge information processing effect, designs a Winform window and an MCGS interface, outputs the analysis result to the robot in a mode of upper computer serial port communication, and executes the tail end to realize accurate sorting. Tests prove that the system has a sorting rate of over 97.0 percent, can achieve higher accuracy and feasibility, and has certain popularization value in practical engineering.
Drawings
FIG. 1 is a schematic diagram of a system architecture;
FIG. 2 is a system operating schematic;
FIG. 3 is a schematic diagram of the movement of an IRB1410 mechanism and coordinate systems of joints;
FIG. 4 is a visual processing flow;
FIG. 5 is an image processing flow;
FIG. 6 is a communication debug interface;
in the attached figure 1: IRB1410 is robot control ware, and 2 is pending rubbish, and 3 is the transfer chain, and 4 are the body, and 5 are the clamping jaw, and 6 are the camera, and 7 are the light source, and 8 are the communication line, and 9 are PC, and 10 are harmful dustbin, and 11 are recoverable dustbin, and 12 are dry dustbin.
Detailed Description
The design mainly comprises an IRB1410 robot workstation, a garbage conveying device to be sorted, a vision acquisition-recognition-communication platform, an MCGS human-computer interface and three types of garbage collection boxes (recoverable garbage, dry garbage and harmful garbage), wherein the robot workstation comprises a body, a controller and a demonstrator, and is shown in figure 1.
When the conveying device conveys the to-be-sorted circulation commodity garbage to a visual detection platform acquisition area, the sensor sends an acquisition signal, the visual system acquires an outline image, a bar code image and position information, and the identified bar code number is matched with a circulation commodity bar code database and a garbage classification training set. The pose (position and attitude) and classification result information are transmitted to a robot controller, parameters of each joint are calculated according to inverse kinematics, a PROC program is called to grab garbage to be sorted and put the garbage to a corresponding garbage can, and the robot can set the running speed to adapt to the beat of the garbage to be sorted without pause; the working principle diagram of the system is shown in figure 2.
IRB1410 robot inverse kinematics solution
The design is that the pose information of the image is sent to the tail end of the robot through a controller, and the position and the posture of each joint are solved according to a joint transformation matrix, so that an inverse kinematics method is adopted for solving, motors of each joint are driven, and a mechanism motion diagram and a coordinate system of each joint are shown in figure 3; IRB1410 robot link parameters are as follows:
Figure BDA0003227695700000051
in the table: joint i-robot joint number; ai-link length; α i-link torsion angle; di-link offset; θ i — link angle.
The inverse kinematics equation is as follows:
Figure BDA0003227695700000052
in the formula:
Figure BDA0003227695700000053
-a robot end pose matrix;
Figure BDA0003227695700000054
-a pose transformation matrix of joint i to joint i + 1; a vector in the direction of the n-X axis; a vector in the o-Y direction; a vector in the direction of the a-Z axis.
It is known that
Figure BDA0003227695700000055
And (3) solving the rotation angles theta 1-theta 6 of each joint by using the matrix and the parameters of each connecting rod:
Figure BDA0003227695700000056
in the formula: a. thei-a pose transformation matrix for joint i; rot-rotation operator; trans-translation operator; X-X coordinate axis, Z-Z coordinate axis.
The joint angles were determined by analytical methods:
Figure BDA0003227695700000061
θ2＝arctan(c1px+s1py)/px
θ3＝s2(c1px+s1py)+c2pz
Figure BDA0003227695700000062
Figure BDA0003227695700000063
θ6＝arctans6/c6
in the formula: p is a radical ofx/py/pz-representing the component of the link length in the X/Y/Z direction, respectively; c. Ci-represents the cosine function cos θi；siRepresenting a sine function sin θi(ii) a In conclusion, the pose of the garbage is acquired through visual positioning and then transmitted to R6, and the sorting is realized through the execution parameters of each joint.
Visual system and working principle
The vision system is composed of an image acquisition device, an image processing system, a robot sorting workstation and a control communication transmission system. The image acquisition comprises a light source, a lens and a camera, wherein the light source adopts annular white light, and the number of camera pixels is 1.3 ten thousand; the image processing system adopts a Halcon machine vision processing platform, converts the processing result to a Winform window of VS2010 through C # hybrid programming, and communicates to an MCGS interface through TCP/IP for displaying and counting, as shown in FIG. 4.
When the automatic sorting machine works, when different types of circulating commodities pass through the photoelectric sensor, a conveying belt triggers sensor signal output, a camera shoots in real time, an image is sent to a Halcon processing platform, appearance and bar code information on a package is extracted through processes of RGB gray level conversion, HSV conversion, threshold segmentation, connected domain closing operation, feature extraction, database matching decoding, display and the like, a solving algorithm is output to a VS2010 window in a C # format, the type is displayed and counted on a human-computer interface synchronously, and a robot finishes sorting.
Image processing
In the links, the light source is required to be adjusted to be consistent with the target area of the camera before image acquisition, the photographing effect is eliminated due to the interference of the conveying belt, and the photographing frame number is greater than the movement frequency of the conveying belt. And converting the collected RGB image gray level into a gray level image and an HSV image, and displaying color discrimination and extracting characteristics. The key point of feature extraction is to select a relatively obvious gray scale image from the obtained HSV image, extract the outline and the bar code region after threshold segmentation and connected domain operation, and the difficulty is that the outline is fitted and EAN bar code decoding is carried out, and the obtained numerical value is compared with a known database to match the optimal value of a training set. The mechanism is as follows:
gray scale conversion
g(x，y)＝T[f(x,y)]
In the formula: f (x, y) -digital image to be processed; g (x, y) -processed digital image; t-defines the operation of f.
Threshold segmentation (variance method between maximum classes)
Defining the pixel with the gray value i epsilon [0, L-1] as ni, then:
Figure BDA0003227695700000071
the probability of each gray value occurring is:
Figure BDA0003227695700000072
for pi have
Figure BDA0003227695700000073
The image pixels are divided into two types of A and B by using a threshold value T, wherein A is composed of gray values in [0, T-1] pixels, B is composed in [ T, L-1] pixels, and the probability is respectively
Figure BDA0003227695700000074
Figure BDA0003227695700000075
The average gray scale of the whole image is represented by u, and the average gray scale of the region A, B to which the average gray scale belongs is:
Figure BDA0003227695700000076
Figure BDA0003227695700000077
Figure BDA0003227695700000078
the total variance of the regions is:
Figure BDA0003227695700000079
in the formula: l-1 represents the maximum gray value; n-sum of gray values; p-probability; u-average gray;
Figure BDA0003227695700000081
-the area total variance.
Let T take the value in the interval L-1, make
Figure BDA0003227695700000082
The maximum T is the optimal region threshold value, and the image processing effect is optimal.
Canny edge
The Canny edge detection operator is an operator with better edge detection performance, converts the edge detection problem into the problem of maximum value of a detection function by utilizing the first-order differential property of a Gaussian function, and can obtain better compromise between noise suppression and edge detection[8]。
Figure BDA0003227695700000083
G(x,y)＝f(x,y)*H(x，y)
In the formula: f-image data, H-Gaussian function of omission factor, G-filtered smooth image, (x, y) pixel point position coordinate value, and sigma standard deviation.
The barcode image processing process of the garbage external package is shown in fig. 5, and the result text can be correctly read and displayed.
MCGS interface design
To facilitate result visualization, the image analysis results are displayed in real time on the MCGS human-computer interface[9]. Exporting an algorithm on a Halcon platform in a C # format, synchronizing the algorithm to a Winform window of VS2010, and downloading the algorithm to an MCGS human-computer interface through a TCP/IP communication protocol to realize visualization of garbage types and quantity, as shown in FIG. 6;
test materials and environments
The test sample is from a circulation commodity sold in a supermarket, bar codes are arranged on the surface of an outer package, 4765 kinds of garbage information in a circulation commodity database is obtained through tests according to garbage classification standards, 50 kinds of recoverable garbage, 30 kinds of dry garbage and 20 kinds of harmful garbage are respectively selected according to large base number difference of various types of garbage, the length, the width, the height or the diameter are smaller than 100mm, and the recyclable garbage, the dry garbage and the harmful garbage are randomly mixed and loaded in a test tray;
robot system initialization
Before debugging, the robot is initialized. Setting a calibration position of the robot to enable signals of the Hall sensors to be 0, setting other positions to be 1, newly building a Tool coordinate system Tool 1 according to a right-hand rule, enabling an original point to be located at the center of the clamping jaw, enabling a Z axis to be perpendicular to the end face of the flange plate, and calibrating scales for all axes of the robot on the lower table
Figure BDA0003227695700000084
Test results
The conveyer belt was set at a speed of 100mm/s, and the results of the random sorting test of 100 kinds of 3 kinds of garbage were shown in table 3, indicating that the robot sorting rate in the test was 97.0% or more. The average matching time of the collected bar code and the database is 0.002s, which shows that the design has higher accuracy and stability; the following table shows the test results
Figure BDA0003227695700000091
And an ABB robot motion control platform taking Halcon machine vision as a carrier sorts the circulated commodity outer package garbage in real time. The system composition and the working principle are introduced, the inverse kinematics equation of the IRB1410 robot body is solved, the technical process of visual system image processing is optimized, the technical process comprises image acquisition, gray level conversion, threshold segmentation, flow domain matching decoding and the like, the Canny algorithm is adopted to improve the edge information processing effect, a Winform window and an MCGS interface are designed, the analysis result is output to the robot in a host computer serial port communication mode, and the tail end is executed to realize accurate sorting. Tests prove that the system has a sorting rate of over 97.0 percent, can achieve higher accuracy and feasibility, and has certain popularization value in practical engineering.
The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.

Claims (3)

1. The robot-based machine vision garbage sorting system comprises an IRB1410 robot workstation, a to-be-sorted garbage conveying device, a vision system and three types of garbage collection boxes, wherein the three types of garbage collection boxes are respectively a harmful garbage box, a recyclable garbage box and a dry garbage box;
the method is characterized in that: when the to-be-sorted garbage conveying device conveys the to-be-sorted commodity garbage to a visual detection platform acquisition area, a sensor sends an acquisition signal, a visual system acquires an outline image, a bar code image and position information, and the identified bar code number is matched with a commodity barcode database and a garbage classification training set; then, the pose and classification result information is transmitted to a robot controller, and then the pose information of the image is sent to the tail end of the robot through the controller; and solving the position and the posture of each joint according to the joint transformation matrix, solving parameters of each joint by adopting an inverse kinematics method, driving motors of each joint, calling a PROC program to grab the garbage to be sorted, putting the garbage to a corresponding garbage can, and adapting to the beat of the garbage to be sorted by setting the running speed of the robot.
2. The robot-based machine vision garbage sorting system of claim 1, wherein: the inverse kinematics equation of the inverse kinematics method is as follows:
Figure FDA0003227695690000011
in the above formula:
Figure FDA0003227695690000012
representing a robot end pose matrix;
Figure FDA0003227695690000013
representing a pose transformation matrix from the joint i to the joint i + 1; n represents a vector in the x-axis direction; o represents a vector in the y-axis direction; a represents a vector in the z-axis direction; it is known that
Figure FDA0003227695690000014
Matrix and each link parameter, and solving the rotation angle theta 1-theta 6 of each joint link:
Figure FDA0003227695690000015
the joint angles were determined by analytical methods: θ 1, θ 2, θ 3, θ 4, θ 5, θ 6; after the pose of the garbage is obtained through visual positioning, the execution parameters of each joint realize sorting; in the above formula: a. theiRepresenting a pose transformation matrix of the joint i; rot represents a rotation operator; trans denotes the translation operator; the joint i represents a robot joint number; a isiIndicating a link length; alpha is alphaiRepresenting a connecting rod torsion angle; diIndicating a link offset; thetaiIndicating the link angle.
3. The robot-based machine vision garbage sorting system of claim 1, wherein: the vision system consists of an image acquisition device, an image processing system, a robot sorting workstation and a control communication transmission system; the image acquisition device consists of a light source, a lens and a camera, wherein the light source adopts annular white light, and the number of camera pixels is 1.3 ten thousand; the image processing system adopts a Halcon machine vision processing platform, converts the processing result to a Winform window of VS2010 through C # hybrid programming, and transmits the processing result to an MCGS interface through TCP/IP communication for displaying and counting; when the system works, when different types of circulating commodities pass through a photoelectric sensor, a conveyer belt triggers sensor signal output, a camera shoots in real time, an image is sent to a Halcon processing platform, appearance and bar code information on a package is extracted through processes of RGB gray level conversion, HSV conversion, threshold segmentation, connected domain closing operation, feature extraction, database matching decoding, display and the like, a solving algorithm is output to a VS2010 window in a C # format, the type is displayed and counted on a human-computer interface synchronously, and a robot finishes sorting;
before image acquisition, a light source is required to be adjusted to be consistent with a camera target area, the photographing effect due to the interference of a conveying belt is eliminated, and the photographing frame number is greater than the movement frequency of the conveying belt; converting the collected RGB image gray level into a gray level image and an HSV image, and displaying color discrimination and extracting characteristics; the characteristic extraction is to select an obvious gray image from the obtained HSV image, extract a contour and a bar code area after threshold segmentation and connected domain operation, compare the obtained numerical value with a known database and match the optimal value of a training set;
the grayscale conversion mechanism is as follows:
g(x，y)＝T[f(x,y)]
in the formula: f (x, y) represents a digital image to be processed; g (x, y) represents the processed digital image; t defines the operation of f;
threshold segmentation:
let the gray value be i E [0, L-1]Is defined as niAnd then:
Figure FDA0003227695690000021
the probability of each gray value occurring is:
Figure FDA0003227695690000022
for piIs provided with
Figure FDA0003227695690000023
The image pixels are divided into two types of A and B by using a threshold value T, wherein A is composed of gray values in [0, T-1] pixels, B is composed in [ T, L-1] pixels, and the probability is respectively
Figure FDA0003227695690000031
Figure FDA0003227695690000032
The average gray scale of the whole image is represented by u, and the average gray scale of the region A, B to which the average gray scale belongs
Figure FDA0003227695690000033
Figure FDA0003227695690000034
Figure FDA0003227695690000035
The total variance of the regions is:
Figure FDA0003227695690000036
let T take the value in the interval L-1, make
Figure FDA0003227695690000037
The maximum T is the optimal region threshold value, and the image processing effect is optimal; the first-order differential property of the Gaussian function is utilized to convert the problem of edge detection into the problem of maximum value of the detection function, and compromise is obtained between noise suppression and edge detection;
Figure FDA0003227695690000038
G(x,y)＝f(x,y)*H(x，y)
in the formula: l-1 represents the maximum gray value; n represents a gray value; p represents a probability; u represents an average gray;
Figure FDA0003227695690000039
representing the total variance of the region; f denotes image data, H denotes a gaussian function with coefficients omitted, G denotes a filtered smoothed image, (x, y) denotes pixel point position coordinate values, and σ denotes standard deviation.
CN202110976991.7A 2021-08-24 2021-08-24 Robot-based machine vision garbage sorting system Pending CN113680695A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110976991.7A CN113680695A (en) 2021-08-24 2021-08-24 Robot-based machine vision garbage sorting system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110976991.7A CN113680695A (en) 2021-08-24 2021-08-24 Robot-based machine vision garbage sorting system

Publications (1)

Publication Number Publication Date
CN113680695A true CN113680695A (en) 2021-11-23

Family

ID=78582228

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110976991.7A Pending CN113680695A (en) 2021-08-24 2021-08-24 Robot-based machine vision garbage sorting system

Country Status (1)

Country Link
CN (1) CN113680695A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115488876A (en) * 2022-06-22 2022-12-20 湖北商贸学院 Robot sorting method and device based on machine vision

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103347661A (en) * 2010-10-21 2013-10-09 泽恩机器人技术有限公司 Method for filtering of target object image in robot system
CN105772403A (en) * 2016-01-26 2016-07-20 耿春茂 Waste classification detecting system
CN109344894A (en) * 2018-09-28 2019-02-15 广州大学 Garbage classification recognition methods and device based on Multi-sensor Fusion and deep learning
CN110125036A (en) * 2019-04-25 2019-08-16 广东工业大学 A kind of self-identifying sorting system and method based on template matching
CN110976338A (en) * 2019-11-11 2020-04-10 浙江大学 Test paper sorting system and method based on machine vision
CN111077150A (en) * 2019-12-30 2020-04-28 重庆医科大学附属第一医院 Intelligent excrement analysis method based on computer vision and neural network
CN111260616A (en) * 2020-01-13 2020-06-09 三峡大学 Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization
CN112149573A (en) * 2020-09-24 2020-12-29 湖南大学 Garbage classification and picking robot based on deep learning
CN112278636A (en) * 2020-10-15 2021-01-29 张良虎 Garbage classification recycling method, device and system and storage medium
US20210178432A1 (en) * 2017-06-30 2021-06-17 Boe Technology Group Co., Ltd. Trash sorting and recycling method, trash sorting device, and trash sorting and recycling system

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103347661A (en) * 2010-10-21 2013-10-09 泽恩机器人技术有限公司 Method for filtering of target object image in robot system
CN105772403A (en) * 2016-01-26 2016-07-20 耿春茂 Waste classification detecting system
US20210178432A1 (en) * 2017-06-30 2021-06-17 Boe Technology Group Co., Ltd. Trash sorting and recycling method, trash sorting device, and trash sorting and recycling system
CN109344894A (en) * 2018-09-28 2019-02-15 广州大学 Garbage classification recognition methods and device based on Multi-sensor Fusion and deep learning
CN110125036A (en) * 2019-04-25 2019-08-16 广东工业大学 A kind of self-identifying sorting system and method based on template matching
CN110976338A (en) * 2019-11-11 2020-04-10 浙江大学 Test paper sorting system and method based on machine vision
CN111077150A (en) * 2019-12-30 2020-04-28 重庆医科大学附属第一医院 Intelligent excrement analysis method based on computer vision and neural network
CN111260616A (en) * 2020-01-13 2020-06-09 三峡大学 Insulator crack detection method based on Canny operator two-dimensional threshold segmentation optimization
CN112149573A (en) * 2020-09-24 2020-12-29 湖南大学 Garbage classification and picking robot based on deep learning
CN112278636A (en) * 2020-10-15 2021-01-29 张良虎 Garbage classification recycling method, device and system and storage medium

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
冷玉珊等: "基于MATLAB六自由度串联机器人运动学分析", 《制造业自动化》 *
冷玉珊等: "基于MATLAB六自由度串联机器人运动学分析", 《制造业自动化》, vol. 42, no. 9, 30 September 2020 (2020-09-30), pages 56 *
冷玉珊等: "基于MATLAB六自由度串联机器人运动学分析", 制造业自动化, vol. 42, no. 9, pages 56 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115488876A (en) * 2022-06-22 2022-12-20 湖北商贸学院 Robot sorting method and device based on machine vision

Similar Documents

Publication Publication Date Title
CN110314854B (en) Workpiece detecting and sorting device and method based on visual robot
CN109785317B (en) Automatic pile up neatly truss robot's vision system
CN110580725A (en) Box sorting method and system based on RGB-D camera
CN105930854A (en) Manipulator visual system
CN111368852A (en) Article identification and pre-sorting system and method based on deep learning and robot
CN111604909A (en) Visual system of four-axis industrial stacking robot
CN105701476A (en) Machine vision-based automatic identification system and method for production line products
CN111266304A (en) Coal briquette identification, detection and sorting system
CN208574964U (en) A kind of automatic identification and sorting cigarette case device based on machine vision
CN203196911U (en) Industrial CCD (Charge Coupled Device) based visual identification sorting system
CN103308524A (en) PCB automatic optical inspection system
WO2022100145A1 (en) Infusion bag foreign matter identification method and detection method based on artificial intelligence visual technology
CN110640741A (en) Grabbing industrial robot with regular-shaped workpiece matching function
CN112497219A (en) Columnar workpiece classification positioning method based on target detection and machine vision
CN113680695A (en) Robot-based machine vision garbage sorting system
CN105154988A (en) Apparatus automatically extracting down feather and extracting method
CN208092786U (en) A kind of the System of Sorting Components based on convolutional neural networks by depth
CN114758236A (en) Non-specific shape object identification, positioning and manipulator grabbing system and method
CN111487192A (en) Machine vision surface defect detection device and method based on artificial intelligence
CN111060518A (en) Stamping part defect identification method based on instance segmentation
CN210233036U (en) Binocular artificial intelligence arm teaching device
Yang et al. Research on an Automatic Sorting System Based on Machine Vision
CN102200780A (en) Method for realizing 3H charge coupled device (CCD) visual industrial robot
CN112156992A (en) Machine vision teaching innovation platform
Patil et al. Blob detection technique using image processing for identification of machine printed characters

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