CN109225941B - Automatic detection and sorting system and method for internal thread tapping condition - Google Patents
Automatic detection and sorting system and method for internal thread tapping condition Download PDFInfo
- Publication number
- CN109225941B CN109225941B CN201811175978.6A CN201811175978A CN109225941B CN 109225941 B CN109225941 B CN 109225941B CN 201811175978 A CN201811175978 A CN 201811175978A CN 109225941 B CN109225941 B CN 109225941B
- Authority
- CN
- China
- Prior art keywords
- region
- screw hole
- light source
- degree
- workpiece
- 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
-
- 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
- B07C5/342—Sorting according to other particular properties according to optical properties, e.g. colour
- B07C5/3422—Sorting according to other particular properties according to optical properties, e.g. colour using video scanning devices, e.g. TV-cameras
-
- 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
- B07C5/02—Measures preceding sorting, e.g. arranging articles in a stream orientating
-
- 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
-
- 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
- B07C2501/00—Sorting according to a characteristic or feature of the articles or material to be sorted
- B07C2501/0009—Sorting of fasteners, e.g. screws, nuts, bolts
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Manipulator (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an automatic detection and sorting system and method for tapping conditions of internal threads, which comprises a workpiece conveying line, a photoelectric sensor, an annular light source, a light source controller, an industrial camera, an industrial personal computer and a multi-axis robot, wherein the workpiece conveying line is connected with the photoelectric sensor; the workpiece conveying line comprises a detection station and a blanking station, and the annular light source is connected with the light source controller; the external trigger interface of the industrial camera is connected with the output of the photoelectric sensor, triggers the acquisition of images and transmits the images to the industrial personal computer through the camera interface; the photoelectric sensor is connected with the industrial camera and the light source controller; the industrial personal computer is connected with the industrial camera and the multi-degree-of-freedom robot controller, receives digital image information of the industrial camera, converts position information into physical coordinates of the motion of the multi-degree-of-freedom robot, and transmits the physical coordinates to the multi-degree-of-freedom robot controller for sorting; the invention adopts the machine vision technology to realize the real-time and rapid detection of the tapping condition of the internal thread of the workpiece, and greatly improves the detection efficiency and the automation degree of the tapping condition of the internal thread.
Description
Technical Field
The invention belongs to the technical field of automatic detection of internal thread tapping conditions, and particularly relates to an automatic detection and sorting device and method for internal thread tapping conditions, which are realized by combining a machine vision technology and a multi-degree-of-freedom robot.
Background
The threaded workpiece is the most important fastening connection and transmission assembly in industrial production and life, threaded connection is widely applied, for internal thread detection, most production enterprises at present adopt a manual visual inspection method to complete the manual detection, the manual detection is greatly susceptible to subjective factors, individual differences and other factors, and due to the fact that the workpiece yield is large, the method not only requires workers to carry out high-intensity labor, but also potential safety hazards can be brought to products with missed defects, and therefore efficient, accurate and intelligent detection of the tapping condition of the internal thread of the workpiece is the basic guarantee for achieving automatic production.
Retrieved, patents CN106226316, CN202823923U and CN107289847 disclose automatic thread detection systems based on machine vision, but their designs are all directed to the external thread detection aspect; in the aspect of internal thread detection, the patent CN106644447 detects various parameters such as torque, stroke, time and the like of a rotating component to detect the tapping condition of the internal thread, and the patent CN204115651 provides a pneumatic internal thread detection device for an automobile hub, wherein the patent uses a mechanical force sensing method to feed back the current tapping condition of the internal thread, and the implementation of the method firstly requires precise positioning of a threaded hole, and cannot meet the condition that the threaded hole is changeable; patent CN201610236353 proposes a gas medium laser doppler type internal thread detector, and the detection device of the method has a complex structure and high cost, and is difficult to be popularized in a large range.
Disclosure of Invention
1. Objects of the invention
Aiming at the defects of the existing internal thread tapping condition automatic detection device, the invention aims to solve the technical problem of providing an internal thread tapping condition automatic detection and sorting device and method which adopt the combination of a machine vision technology and a multi-degree-of-freedom robot.
2. The technical scheme adopted by the invention
The invention discloses an automatic detection and sorting system for tapping conditions of internal threads, which comprises a workpiece conveying line, a photoelectric sensor, an annular light source, a light source controller, an industrial camera, an industrial personal computer and a multi-degree-of-freedom robot, wherein the workpiece conveying line is connected with the photoelectric sensor;
the workpiece conveying line comprises a detection station and a blanking station;
the annular light source is connected with the light source controller, the diameter of the annular light source is larger than that of the thread, the annular light source is arranged on the detection line bracket of the detection station, and the front side of a bright field is used for low-angle illumination, so that a dark field is formed in the through hole area of the screw hole, and the internal wall characteristics of the internal thread are enhanced;
the photoelectric sensor is arranged on a detection station, is connected with the industrial camera and the light source controller, is positioned on the diameter of the annular light source perpendicular to the conveying and running direction of the workpiece, triggers the industrial camera to shoot at the moment when the thread reaches the diameter position of the annular light source, triggers the annular light source to start, and adjusts the brightness of the annular light source to proper intensity through the potentiometer;
the industrial camera is fixed on the detection line bracket with a certain inclination angle, the center of the visual field of the industrial camera is concentric with the center of the annular light source, and the annular light source covers the whole workpiece to be detected to obtain complete screw hole and thread imaging characteristics; the external trigger interface is connected with the output end of the photoelectric sensor, triggers the acquisition of images and transmits the images to the industrial personal computer through the camera interface;
furthermore, the industrial camera transmits digital images to the industrial personal computer, and the automatic detection and sorting of the industrial personal computer comprise average smoothing, image pre-segmentation, screw hole region extraction, thread region extraction, feature extraction and calculation, classifier design and training and intelligent sorting of workpieces:
the average smoothing means that the acquired internal thread image adopts an average template of 3 x 3
Carrying out average smooth related filtering processing for filtering noise points on the surface of the internal thread workpiece in the image;
the image pre-segmentation adopts a single threshold segmentation method to segment a screw hole region to be detected and a background region, adopts eight-neighborhood connected region extraction, and realizes the pre-segmentation of the screw hole region by calculating the roundness, the area and the radius of each connected region and screening according to characteristics;
the screw hole region extraction refers to adopting a region growing method, taking the mass center of the pre-segmentation screw hole region as a seed point, and taking the gray value of contour pixels of the screw hole region and the mean value difference and the region area of the screw hole region as a growing rule, namely, the entry condition of adjacent pixels of the screw hole region is that the difference value between the average value of the gray values of the grown region and the value before entry is smaller than a set threshold value, and the area of the screw hole region after growth is smaller than the set threshold value;
the thread region extraction means that the minimum circumscribed circle of the current screw hole region is obtained by taking the center of mass of the screw hole region obtained after the growth of the region as the center, and algebraic subtraction operation is carried out on the circumscribed circle region and the screw hole region, so that the thread region segmentation can be realized;
the characteristic extraction and calculation means that the color, density, circularity, rectangularity, concavity, sphericity, standard deviation and entropy parameters of each screw hole region and each thread region are calculated, and a rough set algorithm is adopted to process region shape and texture descriptors so as to find out the simplest classification description subset suitable for identifying the threaded holes and the tapping holes, so that the characteristic dimensionality reduction and identification rule extraction are realized;
the classifier design and training are to simplify the redundant fuzzy rules by taking the rule coverage and the rule confidence of the rough set as heuristic information, set the threshold of the obtained rule confidence to be 0.75 and the threshold of the rule coverage to be 0.1, and finally obtain a classification criterion to judge the tapping condition of the internal thread of the workpiece to be detected;
the intelligent sorting of the workpieces refers to that the quality information and the physical position information of the workpieces, which are acquired through machine vision, are converted into physical coordinates of the motion of the multi-degree-of-freedom robot, so that the sorted workpieces are sorted;
the industrial personal computer is connected with the industrial camera and the multi-degree-of-freedom robot controller, receives digital image information of the industrial camera, converts position information into physical coordinates of the motion of the multi-degree-of-freedom robot, and transmits the physical coordinates to the multi-degree-of-freedom robot controller for sorting;
the industrial personal computer is a main processor and is connected with an industrial camera through a camera interface to realize real-time acquisition and analysis of digital images, convert the tapping condition of internal threads of a detected workpiece and the position information of the workpiece into physical coordinates of the motion of the multi-degree-of-freedom robot and transmit the physical coordinates to the multi-degree-of-freedom robot controller through a data interface;
the multi-degree-of-freedom robot is used for sorting workpieces with different qualities, a clamping jaw at the front end of the multi-degree-of-freedom robot can be replaced according to different workpieces, the multi-degree-of-freedom robot is tightly connected with a robot body through a standard flange, and a digital image processing method is adopted to process and analyze acquired images so as to realize identification of the tapping condition of internal threads of the detected workpieces.
Furthermore, the front end clamping jaw of the multi-degree-of-freedom robot (7) can be detached and is fixedly connected with the robot body through a standard flange plate.
The invention discloses an automatic detection and sorting method for tapping conditions of internal threads, which is characterized by comprising the following steps of:
(1) calibrating the industrial camera in an off-line state, correcting lens distortion, and calculating the pixel resolution, the aperture and the focal length of a phase modulator at each pixel position in an image to ensure that the obtained image is the clearest;
(2) in an off-line state, performing hand-eye calibration on the industrial camera and the multi-degree-of-freedom robot, and establishing a mapping relation between the industrial camera and a robot coordinate system;
(3) the detected workpiece is placed on the workpiece conveying line;
(4) when the industrial camera arrives at the detection station, the photoelectric sensor outputs and triggers annular light source illumination and industrial camera image acquisition;
(5) the main controller acquires and processes the real-time image acquired by the current imaging scheme;
(6) through image analysis, the current work piece internal thread tapping condition of examining is discerned to the image position of the work piece of examining is fixed a position, specifically is:
(61) averaging and smoothing, and adopting an average template of 3X 3 for the collected internal thread images
Carrying out average smooth related filtering processing for filtering noise points on the surface of the internal thread workpiece in the image;
(62) pre-dividing the image, namely dividing a screw hole region to be detected and a background region by adopting a single threshold value division method, extracting eight neighborhood connected regions, and screening according to characteristics by calculating the roundness, the area and the radius of each connected region to realize the pre-division of the screw hole region;
(63) extracting a screw hole region, namely adopting a region growing method, taking the mass center of the pre-segmentation screw hole region as a seed point, and taking the gray value of contour pixels of the screw hole region and the mean value difference and the region area of the screw hole region as a growing rule, namely, the entry condition of adjacent pixels of the screw hole region is that the difference value between the average value of the gray values of the grown region and the value before entry is smaller than a set threshold value, and the area of the screw hole region after growth is smaller than the set threshold value;
(64) extracting a thread area, namely acquiring the minimum circumscribed circle of the current screw hole area by taking the center of mass of the screw hole area acquired after the growth of the area as the center, and performing algebraic subtraction operation on the circumscribed circle area and the screw hole area to realize thread area segmentation;
(65) the method comprises the steps of extracting and calculating features, calculating the color, density, circularity, rectangularity, concavity, sphericity, standard deviation and entropy parameters of each screw hole region and each thread region, and processing region shapes and texture descriptors by adopting a rough set algorithm to find out the simplest classification description subset suitable for identifying threaded holes and tapping holes so as to realize feature dimension reduction and identification rule extraction;
(66) designing and training a classifier, simplifying redundant fuzzy rules by taking the rule coverage and the rule confidence of a rough set as heuristic information, setting the threshold of the obtained rule confidence to be 0.75 and the threshold of the rule coverage to be 0.1, and finally judging the tapping condition of the internal thread of the workpiece to be detected according to the obtained classification criterion;
(67) the intelligent sorting of the workpieces is realized, and the quality information and the physical position information of the workpieces acquired through machine vision are converted into physical coordinates of the motion of the multi-degree-of-freedom robot, so that the sorted workpieces are realized;
(7) converting the quality information and the physical position information of the current detected workpiece into a three-dimensional motion coordinate of the multi-degree-of-freedom robot by using the hand-eye calibration result;
(8) and clamping and sorting the detected workpieces, and placing the workpieces at a preset position.
3. The invention has the advantages of
(1) The invention adopts the machine vision technology to realize the real-time and rapid detection of the tapping condition of the internal thread of the workpiece, overcomes the defects of the traditional detection method, and greatly improves the detection efficiency and the automation degree of the tapping condition of the internal thread;
(2) the invention carries out quantization processing on the internal thread image characteristics, improves the accuracy of quality detection of the workpiece to be detected, has the average detection time of about 0.256s for a single workpiece, has the detectable rate of 95.88 percent for the missed tapping workpiece and 100 percent for the tapped workpiece, and can meet the requirement of online detection;
(3) the invention combines the machine vision imaging technology, the machine vision automatic detection algorithm and the multi-degree-of-freedom robot to realize the automatic sorting of the workpieces to be detected.
Drawings
FIG. 1 is a schematic structural view of the present invention;
FIG. 2 is a diagram of workpiece imaging results obtained by the imaging protocol of the present invention;
FIG. 3 is an electrical connection diagram of the present invention patent;
FIG. 4 is a flow chart of the patented image processing of the invention;
FIG. 5 is a diagram illustrating the effect of detecting the tapping condition of the internal thread of the workpiece according to the present invention;
in the figure: 1, a workpiece conveying line; 2 a photosensor; 3, a ring-shaped light source; 4, a light source controller; 5 an industrial camera; 6, an industrial personal computer; 7 multi-degree-of-freedom robots; 8, detecting a station; 9, a blanking station; 10, imaging the inner screw hole; 11, imaging the inner wall of the internal thread; 12 an untapped internal screw hole; 13 tapped internal screw holes.
Detailed Description
The detailed technical scheme of the invention is described below by combining the accompanying drawings, and the detection device of the invention can be applied to automatic detection of the tapping condition of the internal thread of the workpiece.
Example 1
As shown in fig. 1, the invention discloses an automatic detection and sorting system for internal thread tapping conditions, which comprises a workpiece conveying line 1, a photoelectric sensor 2, an annular light source 3, a light source controller 4, an industrial camera 5, an industrial personal computer 6 and a multi-degree-of-freedom robot 7;
the workpiece conveying line 1 comprises a detection station 8 and a blanking station 9;
the annular light source 3 is connected with a light source controller, the diameter of the annular light source is larger than that of the thread, the annular light source is arranged on a detection line bracket of the detection station 8, and the front side of a bright field is used for low-angle illumination, so that a dark field is formed in the through hole area of the screw hole, and the internal wall characteristics of the internal thread are enhanced;
the photoelectric sensor 2 is arranged on a detection station 8, is connected with an industrial camera 5 and a light source controller 4, is positioned on the diameter of the annular light source 3 perpendicular to the conveying and running direction of the workpiece, triggers the industrial camera to shoot when the thread reaches the diameter position of the annular light source, triggers the light source to start, and adjusts the brightness of the light source to proper intensity through a potentiometer;
the industrial camera 5 is fixed on the detection line bracket with a certain inclination angle, the center of the visual field of the industrial camera is concentric with the center of the annular light source 3, and the light source covers the whole workpiece to be detected to obtain complete screw hole and thread imaging characteristics; the external trigger interface is connected with the output of the photoelectric sensor 2, triggers the collection of images and transmits the images to the industrial personal computer 6 through the camera interface;
the industrial personal computer 6 is connected with the industrial camera 5 and the multi-degree-of-freedom robot 7, receives digital image information of the industrial camera 5, converts position information into physical coordinates of the multi-degree-of-freedom robot motion, and transmits the physical coordinates to the multi-degree-of-freedom robot 7 for sorting;
the industrial personal computer is a main processor and is connected with an industrial camera through a camera interface to realize real-time acquisition and analysis of digital images, convert the tapping condition of internal threads of a detected workpiece and the position information of the workpiece into physical coordinates of the motion of the multi-degree-of-freedom robot and transmit the physical coordinates to the multi-degree-of-freedom robot controller through a data interface;
the multi-degree-of-freedom robot is used for sorting workpieces with different qualities, a clamping jaw at the front end of the multi-degree-of-freedom robot can be replaced according to different workpieces of the workpieces, the multi-degree-of-freedom robot is tightly connected with a robot body through a standard flange, and a digital image processing method is adopted to process and analyze acquired images so as to realize identification of the tapping condition of the internal threads of the detected workpieces.
Example 2
On the basis of the embodiment 1, as shown in fig. 2, a workpiece imaging effect diagram is obtained by applying the above scheme, that is, the annular light source 3 is selected, bright field front low-angle direct illumination is adopted, the industrial camera 5 is fixed on the detection line support with a certain inclination angle, and the annular light source 3 and the industrial camera 5 are concentric with the center of the detection station 8, so that the internal screw hole imaging 10 forms a dark field, and the internal screw thread inner wall imaging 11 characteristic is enhanced, thereby achieving the purpose of showing the screw hole and the screw thread imaging characteristic of the workpiece to be detected, and inhibiting other unimportant characteristics.
As shown in fig. 3, in the electrical connection of the automatic detection and sorting device for the tapping condition of the internal thread, the external trigger interfaces of the industrial camera 5 and the light source controller 4 are both connected to the photoelectric sensor; the annular light source 3 is connected with the light source controller 4 through a specific interface; the industrial camera 5 is connected with the industrial personal computer 6 through a camera interface; the industrial personal computer 6 is connected with the multi-degree-of-freedom robot 7 through a data interface.
The automatic detection and sorting method of the workpieces to be detected is shown in FIG. 4 and mainly comprises average smoothing, image pre-segmentation, screw hole region extraction, thread region extraction, feature extraction and calculation, classifier design and training and intelligent sorting of the workpieces;
the average smoothing means that the acquired internal thread image adopts an average template of 3 x 3
And carrying out average smooth related filtering processing for filtering noise points on the surface of the metal workpiece in the image.
The image pre-segmentation adopts a single threshold segmentation method to segment a screw hole region to be detected and a background region, adopts eight-neighborhood connected region extraction, and realizes the pre-segmentation of the screw hole region by calculating the roundness, the area and the radius of each connected region and screening according to characteristics;
the screw hole region extraction means that a region growing method is adopted, the mass center of the pre-segmentation screw hole region is used as a seed point, the difference between the gray value of contour pixels of the screw hole region and the mean value of the screw hole region and the region area are used as a growing rule, namely the entry condition of adjacent pixels of the screw hole region is that the difference between the average value of the gray value of the grown region and the difference before entry is smaller than a set threshold value, the area of the grown screw hole region is smaller than the set threshold value, and when no pixel meets the entry condition, the region growing algorithm is finished, so that the complete internal screw hole imaging 10 is obtained;
the thread region extraction means that the minimum circumscribed circle of the current screw hole region is obtained by taking the screw hole region mass center obtained after the growth of the region as the center, and algebraic subtraction operation is carried out on the circumscribed circle region and the screw hole region, so that the complete internal thread inner wall imaging 11 is obtained.
The feature extraction and calculation means that color, density, circularity, rectangularity, concavity, sphericity, standard deviation and entropy parameters of each internal thread imaging 10 and internal thread inner wall imaging 11 are calculated, and a rough set algorithm is adopted to process the region shape and texture descriptors to find a simplest classification description subset suitable for identifying the threaded holes and the tapped holes, so that feature dimension reduction and identification rule extraction are realized;
the classifier design and training are to simplify the redundant fuzzy rules by taking the rule coverage and the rule confidence of the rough set as heuristic information, set the threshold of the obtained rule confidence to be 0.75 and the threshold of the rule coverage to be 0.1, and finally obtain the classification criterion with the strongest classification capability to judge the tapping condition of the internal thread of the detected workpiece;
as shown in fig. 5, in the detection effect diagram of the tapping condition of the internal thread of the workpiece, the inner threaded hole 12 which is not tapped is marked by a white ring, the inner threaded hole 13 which is not marked is a tapped inner threaded hole, and the discrimination effect of the machine vision algorithm is consistent with the visual effect, so that the algorithm has a good detection effect.
The intelligent sorting of the workpieces refers to that the quality information and the physical position information of the workpieces, which are acquired through machine vision, are converted into physical coordinates of the motion of the multi-degree-of-freedom robot, so that the sorted workpieces are sorted.
Claims (3)
1. The utility model provides an internal thread tapping condition automated inspection and letter sorting system which characterized in that: the system comprises a workpiece conveying line, a photoelectric sensor, an annular light source, a light source controller, an industrial camera, an industrial personal computer and a multi-degree-of-freedom robot;
the workpiece conveying line comprises a detection station and a blanking station;
the annular light source is connected with the light source controller, the diameter of the annular light source is larger than that of the thread, the annular light source is arranged on the detection line bracket of the detection station, and the front side of a bright field is used for low-angle illumination, so that a dark field is formed in the through hole area of the screw hole, and the internal wall characteristics of the internal thread are enhanced;
the photoelectric sensor is arranged on a detection station, is connected with the industrial camera and the light source controller, is positioned on the diameter of the annular light source perpendicular to the conveying and running direction of the workpiece, triggers the industrial camera to shoot at the moment when the thread reaches the diameter position of the annular light source, triggers the annular light source to start, and adjusts the brightness of the annular light source to proper intensity through the potentiometer;
the industrial camera is fixed on the detection line bracket with a certain inclination angle, the center of the visual field of the industrial camera is concentric with the center of the annular light source, and the annular light source covers the whole workpiece to be detected to obtain complete screw hole and thread imaging characteristics; the external trigger interface is connected with the output end of the photoelectric sensor, triggers the acquisition of images and transmits the images to the industrial personal computer through the camera interface;
the industrial camera transmits digital images to the industrial personal computer, and the industrial personal computer automatically detects and sorts the images, including average smoothness, image pre-segmentation, screw hole region extraction, thread region extraction, feature extraction and calculation, classifier design and training and intelligent sorting of workpieces:
the average smoothing means that the acquired internal thread image adopts an average template of 3 x 3
Carrying out average smooth related filtering processing for filtering noise points on the surface of the internal thread workpiece in the image;
the image pre-segmentation adopts a single threshold segmentation method to segment a screw hole region to be detected and a background region, adopts eight-neighborhood connected region extraction, and realizes the pre-segmentation of the screw hole region by calculating the roundness, the area and the radius of each connected region and screening according to characteristics;
the screw hole region extraction refers to adopting a region growing method, taking the mass center of the pre-segmentation screw hole region as a seed point, and taking the gray value of contour pixels of the screw hole region and the mean value difference and the region area of the screw hole region as a growing rule, namely, the entry condition of adjacent pixels of the screw hole region is that the difference value between the average value of the gray values of the grown region and the value before entry is smaller than a set threshold value, and the area of the screw hole region after growth is smaller than the set threshold value;
the thread region extraction means that the minimum circumscribed circle of the current screw hole region is obtained by taking the center of mass of the screw hole region obtained after the growth of the region as the center, and algebraic subtraction operation is carried out on the circumscribed circle region and the screw hole region, so that the thread region segmentation can be realized;
the characteristic extraction and calculation means that the color, density, circularity, rectangularity, concavity, sphericity, standard deviation and entropy parameters of each screw hole region and each thread region are calculated, and a rough set algorithm is adopted to process region shape and texture descriptors so as to find out the simplest classification description subset suitable for identifying the threaded holes and the tapping holes, so that the characteristic dimensionality reduction and identification rule extraction are realized;
the classifier design and training are to simplify the redundant fuzzy rules by taking the rule coverage and the rule confidence of the rough set as heuristic information, set the threshold of the obtained rule confidence to be 0.75 and the threshold of the rule coverage to be 0.1, and finally obtain a classification criterion to judge the tapping condition of the internal thread of the detected workpiece;
the intelligent sorting of the workpieces refers to that the quality information and the physical position information of the workpieces, which are acquired through machine vision, are converted into physical coordinates of the motion of the multi-degree-of-freedom robot, so that the sorted workpieces are sorted;
the industrial personal computer is connected with the industrial camera and the multi-degree-of-freedom robot controller, receives digital image information of the industrial camera, converts position information into physical coordinates of the motion of the multi-degree-of-freedom robot, and transmits the physical coordinates to the multi-degree-of-freedom robot controller for sorting;
the industrial personal computer is a main processor and is connected with an industrial camera through a camera interface to realize real-time acquisition and analysis of digital images, convert the tapping condition of internal threads of a detected workpiece and the position information of the workpiece into physical coordinates of the motion of the multi-degree-of-freedom robot and transmit the physical coordinates to the multi-degree-of-freedom robot controller through a data interface;
the multi-degree-of-freedom robot is used for sorting workpieces with different qualities, a clamping jaw at the front end of the multi-degree-of-freedom robot can be replaced according to different workpieces, the multi-degree-of-freedom robot is tightly connected with a robot body through a standard flange, and a digital image processing method is adopted to process and analyze acquired images so as to realize identification of the tapping condition of internal threads of the detected workpieces.
2. The automatic internal thread tapping condition detection and sorting system according to claim 1, wherein: the front end clamping jaw of the multi-degree-of-freedom robot (7) can be detached and is fixedly connected with the robot body through a standard flange plate.
3. An automatic detection and sorting method for tapping conditions of internal threads is characterized by comprising the following steps:
(1) calibrating the industrial camera in an off-line state, correcting lens distortion, and calculating the pixel resolution, the aperture and the focal length of a phase modulator at each pixel position in an image to ensure that the obtained image is the clearest;
(2) in an off-line state, performing hand-eye calibration on the industrial camera and the multi-degree-of-freedom robot, and establishing a mapping relation between the industrial camera and a robot coordinate system;
(3) the detected workpiece is placed on the workpiece conveying line;
(4) when the industrial camera arrives at the detection station, the photoelectric sensor outputs and triggers annular light source illumination and industrial camera image acquisition;
(5) the main controller acquires and processes the real-time image acquired by the current imaging scheme;
(6) through image analysis, the current work piece internal thread tapping condition of examining is discerned to the image position of the work piece of examining is fixed a position, specifically is:
(61) averaging and smoothing, and acquiring an average template of 3X 3 for the internal thread image
Carrying out average smooth related filtering processing for filtering noise points on the surface of the internal thread workpiece in the image;
(62) pre-dividing the image, namely dividing a screw hole region to be detected and a background region by adopting a single threshold value division method, extracting eight neighborhood connected regions, and screening according to characteristics by calculating the roundness, the area and the radius of each connected region to realize the pre-division of the screw hole region;
(63) extracting a screw hole region, namely adopting a region growing method, taking the mass center of the pre-segmentation screw hole region as a seed point, and taking the gray value of contour pixels of the screw hole region and the mean value difference and the region area of the screw hole region as a growing rule, namely, the entry condition of adjacent pixels of the screw hole region is that the difference value between the average value of the gray values of the grown region and the value before entry is smaller than a set threshold value, and the area of the screw hole region after growth is smaller than the set threshold value;
(64) extracting a thread area, namely acquiring the minimum circumscribed circle of the current screw hole area by taking the center of mass of the screw hole area acquired after the growth of the area as the center, and performing algebraic subtraction operation on the circumscribed circle area and the screw hole area to realize thread area segmentation;
(65) the method comprises the steps of extracting and calculating features, calculating the color, density, circularity, rectangularity, concavity, sphericity, standard deviation and entropy parameters of each screw hole region and each thread region, and processing region shapes and texture descriptors by adopting a rough set algorithm to find out the simplest classification description subset suitable for identifying threaded holes and tapping holes so as to realize feature dimension reduction and identification rule extraction;
(66) designing and training a classifier, simplifying redundant fuzzy rules by taking the rule coverage and the rule confidence of a rough set as heuristic information, setting the threshold of the obtained rule confidence to be 0.75 and the threshold of the rule coverage to be 0.1, and finally judging the tapping condition of the internal thread of the workpiece to be detected according to the obtained classification criterion;
(67) the intelligent sorting of the workpieces is realized, and the quality information and the physical position information of the workpieces acquired through machine vision are converted into physical coordinates of the motion of the multi-degree-of-freedom robot, so that the sorted workpieces are realized;
(7) converting the quality information and the physical position information of the current detected workpiece into a three-dimensional motion coordinate of the multi-degree-of-freedom robot by using the hand-eye calibration result;
(8) and clamping and sorting the detected workpieces, and placing the workpieces at a preset position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811175978.6A CN109225941B (en) | 2018-10-10 | 2018-10-10 | Automatic detection and sorting system and method for internal thread tapping condition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811175978.6A CN109225941B (en) | 2018-10-10 | 2018-10-10 | Automatic detection and sorting system and method for internal thread tapping condition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109225941A CN109225941A (en) | 2019-01-18 |
| CN109225941B true CN109225941B (en) | 2021-03-19 |
Family
ID=65055907
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811175978.6A Active CN109225941B (en) | 2018-10-10 | 2018-10-10 | Automatic detection and sorting system and method for internal thread tapping condition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109225941B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111144426B (en) * | 2019-12-28 | 2023-05-30 | 广东拓斯达科技股份有限公司 | Sorting method, sorting device, sorting equipment and storage medium |
| CN112858312B (en) * | 2021-01-20 | 2022-09-02 | 上海工程技术大学 | Pen cap silicon bead installation visual detection device and method |
| CN113351503B (en) * | 2021-06-23 | 2023-02-24 | 上海宏金设备工程有限公司 | Scaffold detection and sorting system and method |
| CN115880365A (en) * | 2023-03-01 | 2023-03-31 | 超音速人工智能科技股份有限公司 | Double-station automatic screw screwing detection method, system and device |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS5995407A (en) * | 1982-11-24 | 1984-06-01 | Nagoya Rashi Seisakusho:Kk | Detector for defective bolt |
| KR100899361B1 (en) * | 2008-12-10 | 2009-05-26 | (주)진합 | Screw thread test device |
| CN105066890A (en) * | 2015-07-21 | 2015-11-18 | 苏州佳祺仕信息科技有限公司 | Threaded hole detection device |
| CN105004275A (en) * | 2015-07-21 | 2015-10-28 | 苏州佳祺仕信息科技有限公司 | Threaded hole three-CCD camera detection mechanism |
| CN106226316A (en) * | 2016-08-31 | 2016-12-14 | 江苏大学 | A kind of single camera wide visual field vision thread detecting device and detection method thereof |
| CN106409711B (en) * | 2016-09-12 | 2019-03-12 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of solar energy silicon crystal chip defect detecting system and method |
| CN206235577U (en) * | 2016-11-03 | 2017-06-09 | 上海戈冉泊精模科技有限公司 | A kind of nut thread detection means |
| CN106925529A (en) * | 2017-03-21 | 2017-07-07 | 柳州欧维姆机械股份有限公司 | With the method that internal thread hemicone tooth form and profile automatic checkout system and internal threads hemicone tooth form are detected with profile |
-
2018
- 2018-10-10 CN CN201811175978.6A patent/CN109225941B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109225941A (en) | 2019-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109225941B (en) | Automatic detection and sorting system and method for internal thread tapping condition | |
| CN107804514B (en) | Toothbrush sorting method based on image recognition | |
| CN102539443B (en) | Bottle body defect automatic detection method based on machine vision | |
| CN111230593B (en) | Milling cutter abrasion loss visual measurement method based on dynamic image sequence | |
| CN107154039B (en) | Rubber tube online defect detection method | |
| CN109993154B (en) | Intelligent identification method for single-pointer sulfur hexafluoride instrument of transformer substation | |
| CN106709529B (en) | Visual detection method for photovoltaic cell color difference classification | |
| CN110108712A (en) | Multifunctional visual sense defect detecting system | |
| CN107610085A (en) | A kind of welding point defect detecting system based on computer vision | |
| CN109693140B (en) | Intelligent flexible production line and working method thereof | |
| CN114279357A (en) | Die casting burr size measurement method and system based on machine vision | |
| CN106097323B (en) | Engine cylinder block casting positioning method based on machine vision | |
| CN110728657A (en) | Annular bearing outer surface defect detection method based on deep learning | |
| CN114155301A (en) | Robot target positioning and grabbing method based on Mask R-CNN and binocular camera | |
| CN113822810A (en) | Method for positioning workpiece in three-dimensional space based on machine vision | |
| CN109877437B (en) | Centering detection device and method for resistance welding gun | |
| CN202471610U (en) | Automatic bottle body defect detecting device based on machine vision | |
| CN208155260U (en) | A kind of vision detection system for bulb lamp stem position detection | |
| CN111267094A (en) | Workpiece positioning and grabbing method based on binocular vision | |
| CN114022441A (en) | Defect detection method for irregular hardware | |
| CN111210413B (en) | Pose detection method in movement process of wire feeding mechanism | |
| CN108332665A (en) | A kind of vision detection system and detection method for bulb lamp stem position detection | |
| CN111323425A (en) | Multi-camera visual detection device and method | |
| CN110969357A (en) | Visual detection method for holes of aluminum alloy machined part | |
| CN107121063A (en) | The method for detecting workpiece |
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 |