CN107122783B - Method for quickly identifying assembly connector based on angular point detection - Google Patents

Method for quickly identifying assembly connector based on angular point detection Download PDF

Info

Publication number
CN107122783B
CN107122783B CN201710256335.3A CN201710256335A CN107122783B CN 107122783 B CN107122783 B CN 107122783B CN 201710256335 A CN201710256335 A CN 201710256335A CN 107122783 B CN107122783 B CN 107122783B
Authority
CN
China
Prior art keywords
corner
image
detection
corners
angular point
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201710256335.3A
Other languages
Chinese (zh)
Other versions
CN107122783A (en
Inventor
刘桂雄
黄坚
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.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
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 South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN201710256335.3A priority Critical patent/CN107122783B/en
Publication of CN107122783A publication Critical patent/CN107122783A/en
Application granted granted Critical
Publication of CN107122783B publication Critical patent/CN107122783B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/06—Recognition of objects for industrial automation

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Other Investigation Or Analysis Of Materials By Electrical Means (AREA)

Abstract

The invention discloses an assembly connector rapid identification method based on angular point detection, which comprises the following steps of sample angular point learning statistics: the method comprises the steps of collecting and detecting Harris corners, Shi-Tomasi corners and FAST corners of a sample, and counting the average value and standard deviation of the number of the corners in the sample; and (3) high-robustness corner detection and rapid identification: setting a reliability requirement, and automatically establishing a calculation formula of the lower limit of the number of the possible positions of the connecting piece; selecting an angular point detection algorithm with the largest number average value and an angular point detection algorithm with the smallest standard deviation in angular point learning statistics, enabling the angular point of the image to be detected by a method meeting conditions, if the number of the angular points in the local area of the image is higher than the threshold requirement, considering that the image is a connecting piece most possibly, extracting the image and inputting the image to a classifier for detection; and (3) detecting and quickly identifying full corner points: the method comprises the steps of using three corner detection methods, namely a Harris corner, a Shi-Tomasi corner and a FAST corner, extracting an image from a position with the corners, namely possibly a connecting piece, and inputting the extracted image to a classifier for detection.

Description

Method for quickly identifying assembly connector based on angular point detection
Technical Field
The invention relates to a method for quickly identifying an assembly connecting piece, in particular to a method for quickly identifying an assembly connecting piece based on angular point detection.
Background
The assembly refers to a process of matching and connecting parts or components according to specified technical requirements to form a semi-finished product or a finished product. The assembly is an important process of a product manufacturing process, and the quality of the assembly plays a decisive role in the quality of the product. The assembly quality is discussed under the precondition that all the assembly parts are qualified and are correctly installed, so that whether the parts are wrongly installed or not and whether the parts are neglected to be installed or not must be checked before the assembly quality of the product is detected. Because the number of assembly parts and connectors is large, the assembly detection objects are diversified, and higher requirements and challenges are provided for the accuracy, consistency and instantaneity of the detection technology. Compared with the whole assembly product, the connector is small in size and is often complete in the image, so that the standard connector can be described by a closed geometric figure, and the closed geometric figure necessarily has corner points.
Generally, the method for identifying the object in the image includes a classification-based identification method and a matching-based identification method. The classification method has large calculation amount, high precision and generalization capability, and has better identification accuracy for the condition of providing samples, for example, an automobile piston assembly quality visual detection system (ZL201010543006.5) can automatically perform visual detection based on digital images on the top surface characters, piston ring assembly, graphite layer integrity and other contents of an automobile piston at one time; the visual detection method (201610496560.X) can improve the accuracy and speed of circle center and radius extraction when detecting a circle on an image of an article to be detected, thereby improving the extraction efficiency; the machine vision detection device (201210348279.3) for the surface quality of the piston of the automobile brake master cylinder can automatically complete the quick detection of the surface quality defect of the piston and give a detection result of the surface quality. The identification method based on the inherent characteristics has good detection effect on specific parts, and if the object changes, the characteristics must be detected again by professional visual detection personnel, so that the flexibility is poor. The identification method based on matching has strong flexibility, can not be detected under the condition that the identification method is not set in advance, and has a narrow application range. The invention aims at the problem that the connecting piece is small in size, is complete in an image, can be described by a closed geometric figure and has the characteristic of an angular point. The quick identification of the assembly connecting piece is realized based on the angular point detection, and the efficiency of the identification method based on classification can be effectively improved.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide a method for quickly identifying an assembly connector based on angular point detection.
The purpose of the invention is realized by the following technical scheme:
an assembly connecting piece rapid identification method based on angular point detection comprises the following steps:
step A, sample corner learning statistics; the method comprises the steps of collecting and detecting Harris corners, Shi-Tomasi corners and FAST corners of a sample, and counting the average value and standard deviation of the number of the corners in the sample;
b, high robustness angular point detection and rapid identification are carried out; setting a reliability requirement, and automatically establishing a calculation formula of the lower limit of the number of the possible positions of the connecting piece; selecting an angular point detection algorithm with the largest number average value and an angular point detection algorithm with the smallest standard deviation in angular point learning statistics, enabling the angular point of the image to be detected by a method meeting conditions, if the number of the angular points in the local area of the image is higher than the threshold requirement, considering that the image is a connecting piece most possibly, extracting the image and inputting the image to a classifier for detection;
c, detecting and quickly identifying all corners; the method comprises the steps of using three corner detection methods, namely a Harris corner, a Shi-Tomasi corner and a FAST corner, extracting an image from a position with the corners, namely possibly a connecting piece, and inputting the extracted image to a classifier for detection.
One or more embodiments of the present invention may have the following advantages over the prior art:
compared with the whole assembly product, the connector is small in size, is often complete in an image, can be described by a closed geometric figure, and has an angular point.
Drawings
FIG. 1 is a flow chart of a method for rapid identification of an assembly connector based on corner detection;
FIG. 2 is a top view of a modular connection;
fig. 3 is a diagram of the effect of rapid identification of the assembly connector based on corner detection.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the following embodiments and accompanying drawings.
As shown in fig. 1, a flow of a method for quickly identifying an assembly connector based on corner detection includes the following steps:
step 1, sample corner learning statistics: the method comprises the steps of collecting and detecting Harris corners, Shi-Tomasi corners and FAST corners of a sample, and counting the average value and standard deviation of the number of the corners in the sample;
step 2, high robustness corner detection and rapid identification: setting a reliability requirement, and automatically establishing a calculation formula of the lower limit of the number of the possible positions of the connecting piece; selecting an angular point detection algorithm with the largest number average value and an angular point detection algorithm with the smallest standard deviation in angular point learning statistics, enabling the angular point of the image to be detected by a method meeting conditions, if the number of the angular points in the local area of the image is higher than the threshold requirement, considering that the image is a connecting piece most possibly, extracting the image and inputting the image to a classifier for detection;
step 3, full-angle point detection and rapid identification: the method comprises the steps of using three corner detection methods, namely a Harris corner, a Shi-Tomasi corner and a FAST corner, extracting an image from a position with the corners, namely possibly a connecting piece, and inputting the extracted image to a classifier for detection.
The sample collection stage in the step 1 specifically includes:
let the set of samples be { R1,R2…Ri…RIIf with R }X、RYRespectively represent images RiFor an arbitrary sample image R, the abscissa and ordinate ranges ofiCan be represented as a set of points p (x, y);
Ri＝{p(x,y)|x∈RX∧y∈RY}
let the image point p (x, y) be along the direction wu,vHas a gray scale change value of Ex,yComprises the following steps:
Figure BDA0001273491840000031
the Harris corner, Shi-Tomasi corner, FAST corner and other detection methods select different w for the point P (x, y)u,vAccording to different Ex,yJudging whether P (x, y) is a corner point;
detecting sample set { R } by using Harris corner, Shi-Tomasi corner and FAST corner 3 methods1,R2…Ri…RIThe vectors of the number of the detected angular points are N respectivelyH、NST、NF：
Figure BDA0001273491840000041
Figure BDA0001273491840000042
Figure BDA0001273491840000043
Respectively calculate NHExpected value and variance E (N)H)、D(NH)；NSTExpected value and variance E (N)ST)、D(NST)；NFExpected value and variance E (N)F)、D(NF)。
In the step 2, in the high robustness angular point detection rapid identification, the confidence requirement is set, and a calculation formula of the lower limit of the possible positions of the connecting pieces is automatically established;
the confidence may be divided into a plurality of stages (c)1,c2…), then under different confidences, the lower limit n of the number of Harris corner pointsTHThe calculation formula is as follows:
Figure BDA0001273491840000044
lower limit n of Shi-Tomasi corner numberTSThe calculation formula is as follows:
Figure BDA0001273491840000045
FAST corner number lower limit nTFThe calculation formula is as follows:
Figure BDA0001273491840000046
step 3, directly setting the lower limit n of the angular point number in the full angular point detection rapid identificationTUsing 3 corner detection methods, Harris corner, Shi-Tomasi corner, FAST corner, for locations with corners, i.e. considered possible connections, the extracted image is input to a classifier for detection, 1.
FIG. 2 is a top view of a modular connection; fig. 3 is a diagram of the effect of rapid identification of the assembly connector based on corner detection.
Although the embodiments of the present invention have been described above, the above descriptions are only for the convenience of understanding the present invention, and are not intended to limit the present invention. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Claims (2)

1. A method for quickly identifying an assembly connector based on corner detection is characterized by comprising the following steps:
step A, sample corner learning statistics: the method comprises the steps of collecting and detecting Harris corners, Shi-Tomasi corners and FAST corners of a sample, and counting the average value and standard deviation of the number of the corners in the sample;
b, high robustness angular point detection and rapid identification: setting a confidence requirement, and automatically establishing a calculation formula of the lower limit of the number of the candidate positions of the connecting piece; selecting an angular point detection algorithm with the largest number average value and an angular point detection algorithm with the smallest standard deviation in angular point learning statistics, enabling the angular points of the image to be detected by a method meeting conditions, if the number of the angular points in the local area of the image is higher than the threshold requirement, considering the image as a candidate connecting piece, and extracting the image and inputting the image to a classifier for detection;
c, full-angle point detection and rapid identification: the method comprises the steps of using three corner detection methods, namely a Harris corner, a Shi-Tomasi corner and a FAST corner, extracting an image for a position with the corners, namely a candidate connecting piece, and inputting the extracted image to a classifier for detection;
in the step B, in the fast identification of the high robustness corner detection, the confidence requirement is set, and the calculation formula for automatically establishing the lower limit of the number of the candidate positions of the connecting piece includes:
the confidence score is divided into a plurality of stages (c)1，c2…), then under different confidences, the lower limit n of the number of Harris corner pointsTHThe calculation formula is as follows:
Figure FDA0002607319260000011
lower limit n of Shi-Tomasi corner numberTSThe calculation formula is as follows:
Figure FDA0002607319260000012
FAST corner number lower limit nTFThe calculation formula is as follows:
Figure FDA0002607319260000021
c, directly setting the lower limit n of the number of the angular points in the full angular point detection rapid identification in the step CTUsing 3 corner detection methods of Harris corner, Shi-Tomasi corner and FAST corner, extracting the image and inputting the extracted image to a classifier for detecting the position with the corner, namely the position considered as a candidate connecting piece.
2. The method for rapidly identifying an assembly connector based on corner point detection according to claim 1, wherein the sample collection in the step a specifically comprises:
let the set of samples be { R1，R2…Ri…RIIf with R }X、RYRespectively represent images RiFor an arbitrary sample image R, the abscissa and ordinate ranges ofiCan be represented as a set of points p (x, y);
Ri＝{p(x,y)|x∈RX∧y∈RY}
let the image point p (x, y) be along the direction wu,vHas a gray scale change value of Ex,yComprises the following steps:
Figure FDA0002607319260000022
the Harris corner, Shi-Tomasi corner and FAST corner detection method selects different points P (x, y)wu,vAccording to different Ex,yJudging whether P (x, y) is a corner point;
three methods of Harris corner point, Shi-Tomasi corner point and FAST corner point are used for detecting sample set { R }1，R2…Ri…RIThe vectors of the number of the detected angular points are N respectivelyH、NS、NF：
Figure FDA0002607319260000023
Figure FDA0002607319260000024
Figure FDA0002607319260000025
Respectively calculate NHExpected value and variance E (N)H)、D(NH)；NSExpected value and variance E (N)S)、D(NS)；NFExpected value and variance E (N)F)、D(NF)。
CN201710256335.3A 2017-04-19 2017-04-19 Method for quickly identifying assembly connector based on angular point detection Active CN107122783B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710256335.3A CN107122783B (en) 2017-04-19 2017-04-19 Method for quickly identifying assembly connector based on angular point detection

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710256335.3A CN107122783B (en) 2017-04-19 2017-04-19 Method for quickly identifying assembly connector based on angular point detection

Publications (2)

Publication Number Publication Date
CN107122783A CN107122783A (en) 2017-09-01
CN107122783B true CN107122783B (en) 2020-11-13

Family

ID=59725704

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710256335.3A Active CN107122783B (en) 2017-04-19 2017-04-19 Method for quickly identifying assembly connector based on angular point detection

Country Status (1)

Country Link
CN (1) CN107122783B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108280430B (en) * 2018-01-24 2021-07-06 陕西科技大学 Flow image identification method
CN112884784B (en) * 2021-03-11 2024-06-04 南通大学 Image-based lens detection and front and back judgment method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009068718A1 (en) * 2007-11-29 2009-06-04 Escribano Gonzalez Jesus Environmentally-friendly device for collecting plastic sheets from cultivated soil
CN102072910A (en) * 2010-11-13 2011-05-25 上海交通大学 Visual inspection system for automobile piston assembly quality
CN106338521A (en) * 2016-09-22 2017-01-18 华中科技大学 Additive manufacturing surface defect, internal defect and shape composite detection method and device

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009068718A1 (en) * 2007-11-29 2009-06-04 Escribano Gonzalez Jesus Environmentally-friendly device for collecting plastic sheets from cultivated soil
CN102072910A (en) * 2010-11-13 2011-05-25 上海交通大学 Visual inspection system for automobile piston assembly quality
CN106338521A (en) * 2016-09-22 2017-01-18 华中科技大学 Additive manufacturing surface defect, internal defect and shape composite detection method and device

Also Published As

Publication number Publication date
CN107122783A (en) 2017-09-01

Similar Documents

Publication Publication Date Title
CN108074231B (en) Magnetic sheet surface defect detection method based on convolutional neural network
CN108562589B (en) Method for detecting surface defects of magnetic circuit material
CN105046196B (en) Front truck information of vehicles structuring output method based on concatenated convolutional neutral net
CN106570511B (en) Keyboard character defect online detection system and method based on computer vision
Hassanin et al. A real-time approach for automatic defect detection from PCBs based on SURF features and morphological operations
CN111833303B (en) Product detection method and device, electronic equipment and storage medium
CN105740899B (en) A kind of detection of machine vision image characteristic point and match compound optimization method
US11030436B2 (en) Object recognition
CN106610969A (en) Multimodal information-based video content auditing system and method
CN107016396B (en) Method for deep learning and identifying image characteristics of assembly connecting piece
CN111652085A (en) Object identification method based on combination of 2D and 3D features
CN105117740B (en) Font identification method and apparatus
CN108154151B (en) Rapid multi-direction text line detection method
CN111161213B (en) Industrial product defect image classification method based on knowledge graph
CN103729631A (en) Vision-based connector surface feature automatically-identifying method
CN113706490B (en) Wafer defect detection method
CN108073940B (en) Method for detecting 3D target example object in unstructured environment
CN110619336B (en) Goods identification algorithm based on image processing
CN113516619B (en) Product surface flaw identification method based on image processing technology
CN112529901B (en) Crack identification method in complex environment
CN115272204A (en) Bearing surface scratch detection method based on machine vision
CN114863464B (en) Second-order identification method for PID drawing picture information
CN111667475A (en) Machine vision-based Chinese date grading detection method
CN107122783B (en) Method for quickly identifying assembly connector based on angular point detection
CN115294109A (en) Real wood board production defect identification system based on artificial intelligence, and electronic equipment

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