CN111223081A - Part hole opening recognition and detection method and system based on deep learning - Google Patents
Part hole opening recognition and detection method and system based on deep learning Download PDFInfo
- Publication number
- CN111223081A CN111223081A CN202010001491.7A CN202010001491A CN111223081A CN 111223081 A CN111223081 A CN 111223081A CN 202010001491 A CN202010001491 A CN 202010001491A CN 111223081 A CN111223081 A CN 111223081A
- Authority
- CN
- China
- Prior art keywords
- hole
- countersunk
- black
- deep learning
- black covering
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a part open hole identification and detection method and system based on deep learning, which comprises the following steps: extracting front images of various template parts; classifying and marking the countersunk holes and the black covering holes in the front images of the template parts; carrying out species analysis training on the marked countersunk head holes and the marked black covering holes by using a convolutional neural network, extracting coordinates of the countersunk head holes and the black covering holes, and generating a recognition model; scanning a front image of a part to be detected; loading an identification model, identifying and detecting a counter bore and a black covering hole in a front image of a part to be detected; compared with the existing manual observation, the method can realize real-time online automatic identification and detection of the sizes and the opening positions of the counter bores and the black covering holes in the part to be detected, is not limited by the sizes of airplane parts, ensures the reliability of the detection result, and effectively improves the detection efficiency of defective airplane parts.
Description
Technical Field
The invention relates to the technical field of part visual inspection, in particular to a part open hole identification and detection method and system based on deep learning.
Background
The existing large civil aircraft is usually assembled by tens of thousands of parts, and in order to ensure the flight safety of the large civil aircraft, the requirements on various parameters of the parts are very high. The countersunk holes and the black covering holes are one of the most common structural characteristics in aircraft manufacturing engineering, exist on various aircraft parts such as frames, ribs, wall plates and skins in large quantity, are important factors influencing the service life and flight safety of the aircraft, and form great hidden danger to the safety of the aircraft due to assembly stress generated by forced assembly when the quality defects such as geometric deviation exist. Because aircraft accessories size is great, it is more complicated to include the curved surface in the appearance, and traditional mode through manual observation detects counter bore and black hole processing quality that covers, can not in time observe effectively and judge, can't detect in batches to the trompil quality in the aircraft accessories, hardly satisfies the detection requirement completely.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a part open hole identification and detection method and system based on deep learning.
The invention is realized by the following technical scheme:
a part hole opening recognition and detection method based on deep learning is characterized by comprising the following steps: a. extracting front images of various template parts; b. classifying and marking the countersunk holes and the black covering holes in the front images of the template parts; c. carrying out species analysis training on the marked countersunk head hole and the marked black covering hole by using a convolutional neural network, extracting coordinates of the countersunk head hole and the black covering hole, and generating a recognition model; d. scanning a front image of a part to be detected; e. loading the identification model, and identifying and detecting a counter bore and a black covering hole in the front image of the part to be detected; f. and outputting a detection result.
According to the above technical solution, preferably, step a includes: and placing the template part on a workbench and aligning the template part with the origin, and scanning the front image of the template part.
According to the above technical solution, preferably, step c further includes: and recording the information of each countersunk hole and the black covering hole according to the coordinates of the countersunk hole and the black covering hole.
According to the above technical solution, preferably, the information of the countersunk hole and the black covering hole includes a position of the countersunk hole, a size of the countersunk hole, a distance between adjacent countersunk holes, a position of the black covering hole, and a size of the black covering hole.
According to the above technical solution, preferably, step d further includes: and correcting the front image of the part to be detected to obtain the offset of the part.
The invention also discloses a part open hole identification and detection system based on deep learning, which is characterized by comprising the following components: the extraction unit is used for extracting front images of various template parts; the marking unit is used for classifying and marking the countersunk holes and the black covering holes in the front image of the template part; the training unit is used for carrying out type analysis training on the marked countersunk head hole and the marked black covering hole by using a convolutional neural network, extracting the coordinates of the countersunk head hole and the black covering hole and generating a recognition model; the scanning unit is used for scanning the front image of the part to be detected; the recognition detection unit is used for loading the recognition model, recognizing and detecting the countersunk hole and the black covering hole in the front image of the part to be detected; and an output unit which outputs the detection result.
According to the above technical solution, preferably, the extracting unit includes a positioning module, which is used for placing the template part on the workbench and aligning with the origin, and scanning the front image of the template part.
According to the above technical solution, preferably, the training unit includes an information extraction module, configured to record information of each of the countersunk hole and the black covered hole according to coordinates of the countersunk hole and the black covered hole.
According to the above technical solution, preferably, the scanning unit includes a correction module, which is used for correcting the front image of the part to be detected to obtain the offset of the part.
The invention has the beneficial effects that:
the sizes and the opening positions of the counter bores and the black covering holes in the part to be detected are automatically identified and detected on line in real time through the identification model.
Drawings
FIG. 1 is a schematic workflow diagram of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood by those skilled in the art, the present invention will be further described in detail with reference to the accompanying drawings and preferred embodiments.
As shown in the figure, the invention comprises the following steps: a. extracting front images of various template parts through a camera; b. classifying and marking the countersunk holes and the black covering holes in the front images of the template parts; c. the marked counter bores and the black covering holes are subjected to type analysis training by using a convolutional neural network, in the embodiment, networks such as frcnn, yolo and the like can be used for training, contour extraction and classification are realized at the same time, coordinates of the counter bores and the black covering holes are extracted, a recognition model is generated, in the process of training the counter bores, multiple deformation holes need to be collected in advance for training, the problem that some counter bores are difficult to recognize due to different photographing angles is avoided, and the detection precision is effectively improved; d. scanning a front image of a part to be detected through a camera; e. loading the identification model, and identifying and detecting a counter bore and a black covering hole in the front image of the part to be detected; f. and outputting a detection result, namely, using green marks for the counter bores and the black covering holes with correct sizes and positions and using red marks for errors, storing detection data of the opening quality of the parts, such as detection photos, part numbers, work order numbers, detection time, detection results and other information, and automatically generating an electronic report. The sizes and the opening positions of the counter bores and the black covering holes in the part to be detected are automatically identified and detected on line in real time through the identification model.
According to the above embodiment, preferably, step a includes: the template part is placed on the workbench and aligned with the original point, the front image of the template part is scanned through the camera, one corner of the part is aligned with the original point of the workbench, and the coordinates of the countersunk hole and the black covering hole are determined according to the coordinate system relation between the most basic physical coordinates of the construction platform and the pixel coordinates of the image.
According to the above embodiment, preferably, step c further includes: and recording the information of each countersunk hole and the black covering hole according to the coordinates of the countersunk hole and the black covering hole, and deducing the position and the size of each countersunk hole and the black covering hole on the part through the coordinates of the countersunk hole and the black covering hole.
According to the above embodiment, preferably, the information of the countersunk hole and the black covering hole includes a countersunk hole position, a countersunk hole size, an adjacent countersunk hole distance, a black covering hole position, and a black covering hole size.
According to the above embodiment, preferably, step d further includes: correcting the front image of the part to be detected to obtain the offset of the part, correcting the part to be detected according to the corresponding template part through an algorithm to reach the position almost matched with the template part on the image, so as to accurately determine the position coordinates of the counter bore and the black covering hole.
The invention also discloses a part open pore recognition and detection system based on deep learning, which uses the part open pore recognition and detection method based on deep learning of claim 5, and is characterized by comprising the following steps: the extraction unit is used for extracting front images of various template parts; the marking unit is used for classifying and marking the countersunk holes and the black covering holes in the front image of the template part; the training unit is used for carrying out type analysis training on the marked countersunk head hole and the marked black covering hole by using a convolutional neural network, extracting the coordinates of the countersunk head hole and the black covering hole and generating a recognition model; the scanning unit is used for scanning the front image of the part to be detected; the recognition detection unit is used for loading the recognition model, recognizing and detecting the countersunk hole and the black covering hole in the front image of the part to be detected; and an output unit which outputs the detection result.
According to the above embodiment, preferably, the extracting unit includes a positioning module, which is used for placing the template part on the workbench and aligning with the origin, and scanning the front image of the template part.
According to the above embodiment, preferably, the training unit includes an information extraction module, configured to record information of each of the countersunk hole and the black covered hole according to coordinates of the countersunk hole and the black covered hole.
According to the above embodiment, preferably, the scanning unit includes a correction module, which is configured to correct the front image of the part to be detected to obtain the offset of the part.
The sizes and the opening positions of the counter bores and the black covering holes in the part to be detected are automatically identified and detected on line in real time through the identification model.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (9)
1. A part hole opening recognition and detection method based on deep learning is characterized by comprising the following steps: a. extracting front images of various template parts; b. classifying and marking the countersunk holes and the black covering holes in the front images of the template parts; c. carrying out species analysis training on the marked countersunk head hole and the marked black covering hole by using a convolutional neural network, extracting coordinates of the countersunk head hole and the black covering hole, and generating a recognition model; d. scanning a front image of a part to be detected; e. loading the identification model, and identifying and detecting a counter bore and a black covering hole in the front image of the part to be detected; f. and outputting a detection result.
2. The part hole opening identification and detection method based on deep learning of claim 1, wherein the step a comprises: and placing the template part on a workbench and aligning the template part with the origin, and scanning the front image of the template part.
3. The deep learning-based part hole opening identification and detection method as claimed in claim 2, wherein the step c further comprises: and recording the information of each countersunk hole and the black covering hole according to the coordinates of the countersunk hole and the black covering hole.
4. The part open hole identification and detection method based on deep learning as claimed in claim 3, wherein the information of the countersunk hole and the black covered hole comprises a countersunk hole position, a countersunk hole size, an adjacent countersunk hole distance, a black covered hole position and a black covered hole size.
5. The deep learning based part hole opening identification and detection method according to any one of claims 1 to 4, wherein the step d further comprises: and correcting the front image of the part to be detected to obtain the offset of the part.
6. A deep learning based part hole recognition and detection system using the deep learning based part hole recognition and detection method of claim 5, comprising:
the extraction unit is used for extracting front images of various template parts;
the marking unit is used for classifying and marking the countersunk holes and the black covering holes in the front image of the template part;
the training unit is used for carrying out type analysis training on the marked countersunk head hole and the marked black covering hole by using a convolutional neural network, extracting the coordinates of the countersunk head hole and the black covering hole and generating a recognition model;
the scanning unit is used for scanning the front image of the part to be detected;
the recognition detection unit is used for loading the recognition model, recognizing and detecting the countersunk hole and the black covering hole in the front image of the part to be detected;
and an output unit which outputs the detection result.
7. The deep learning based part hole opening recognition and detection system as claimed in claim 6, wherein the extraction unit comprises a positioning module for placing the template part on the workbench and aligning with the origin, and scanning the front image of the template part.
8. The deep learning based part hole opening recognition and detection system as claimed in claim 7, wherein the training unit comprises an information extraction module for recording information of each of the countersunk hole and the black covered hole according to the coordinates of the countersunk hole and the black covered hole.
9. The part opening recognition and detection system based on deep learning as claimed in any one of claims 6 to 8, wherein the scanning unit comprises a correction module for correcting the front image of the part to be detected to obtain the part offset.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010001491.7A CN111223081A (en) | 2020-01-02 | 2020-01-02 | Part hole opening recognition and detection method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010001491.7A CN111223081A (en) | 2020-01-02 | 2020-01-02 | Part hole opening recognition and detection method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111223081A true CN111223081A (en) | 2020-06-02 |
Family
ID=70830959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010001491.7A Pending CN111223081A (en) | 2020-01-02 | 2020-01-02 | Part hole opening recognition and detection method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111223081A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113160162A (en) * | 2021-04-14 | 2021-07-23 | 深圳远荣智能制造股份有限公司 | Hole recognition method and device applied to workpiece and hole processing equipment |
| CN113343355A (en) * | 2021-06-08 | 2021-09-03 | 四川大学 | Aircraft skin profile detection path planning method based on deep learning |
| CN115131553A (en) * | 2022-08-30 | 2022-09-30 | 季华实验室 | Shielding hole positioning method and device, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101619964A (en) * | 2009-05-26 | 2010-01-06 | 北京理工大学 | Quick microscopic detection method and quick microscopic detection device of micro accessory based on process match |
| CN103699055A (en) * | 2013-12-24 | 2014-04-02 | 沈阳飞机工业（集团）有限公司 | Intelligent numerical control machining programming system and intelligent numerical control machining programming method for aircraft structural parts |
| CN107463533A (en) * | 2017-08-17 | 2017-12-12 | 广东工业大学 | A kind of three-dimensional CAD physical model manufacturing feature recognition methods based on PCA and CNN |
| CN110009617A (en) * | 2019-04-02 | 2019-07-12 | 浙江大学 | A kind of shaft forgings online quality control method |
| CN110385466A (en) * | 2019-08-07 | 2019-10-29 | 沈阳飞机工业（集团）有限公司 | A kind of numerical-control processing method of titanium alloy ultra-large type complex thin-wall component |
-
2020
- 2020-01-02 CN CN202010001491.7A patent/CN111223081A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101619964A (en) * | 2009-05-26 | 2010-01-06 | 北京理工大学 | Quick microscopic detection method and quick microscopic detection device of micro accessory based on process match |
| CN103699055A (en) * | 2013-12-24 | 2014-04-02 | 沈阳飞机工业（集团）有限公司 | Intelligent numerical control machining programming system and intelligent numerical control machining programming method for aircraft structural parts |
| CN107463533A (en) * | 2017-08-17 | 2017-12-12 | 广东工业大学 | A kind of three-dimensional CAD physical model manufacturing feature recognition methods based on PCA and CNN |
| CN110009617A (en) * | 2019-04-02 | 2019-07-12 | 浙江大学 | A kind of shaft forgings online quality control method |
| CN110385466A (en) * | 2019-08-07 | 2019-10-29 | 沈阳飞机工业（集团）有限公司 | A kind of numerical-control processing method of titanium alloy ultra-large type complex thin-wall component |
Non-Patent Citations (1)
| Title |
|---|
| 徐超 等: "基于视觉测量的沉头孔垂直度检测方法", 《浙江大学学报(工学刊)》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113160162A (en) * | 2021-04-14 | 2021-07-23 | 深圳远荣智能制造股份有限公司 | Hole recognition method and device applied to workpiece and hole processing equipment |
| CN113343355A (en) * | 2021-06-08 | 2021-09-03 | 四川大学 | Aircraft skin profile detection path planning method based on deep learning |
| CN113343355B (en) * | 2021-06-08 | 2022-10-18 | 四川大学 | Aircraft skin profile detection path planning method based on deep learning |
| CN115131553A (en) * | 2022-08-30 | 2022-09-30 | 季华实验室 | Shielding hole positioning method and device, electronic equipment and storage medium |
| CN115131553B (en) * | 2022-08-30 | 2022-11-08 | 季华实验室 | Shielding hole positioning method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111951237B (en) | Visual appearance detection method | |
| CN111223081A (en) | Part hole opening recognition and detection method and system based on deep learning | |
| CN108917593B (en) | Intelligent measurement system and method based on element configuration of workpiece to be measured | |
| CN109142383B (en) | Character defect detection method based on morphology | |
| TWI669519B (en) | Board defect filtering method and device thereof and computer-readabel recording medium | |
| CN108710876A (en) | A kind of battery surface mark defect inspection method and system based on machine vision | |
| CN111242902A (en) | Method, system and equipment for identifying and detecting parts based on convolutional neural network | |
| CN107895362A (en) | A kind of machine vision method of miniature binding post quality testing | |
| CN109946304A (en) | Surface defects of parts on-line detecting system and detection method based on characteristic matching | |
| EP4009038A1 (en) | Method and device for detecting mechanical equipment parts | |
| CN115131268A (en) | Automatic welding system based on image feature extraction and three-dimensional model matching | |
| CN106599890A (en) | Transformer substation patrol robot digital type instrument identification algorithm | |
| CN113688817A (en) | Instrument identification method and system for automatic inspection | |
| CN108268841A (en) | A kind of rolled steel plate thermal jet character string identification and verification system and method | |
| CN105023018A (en) | Jet code detection method and system | |
| CN111626995B (en) | Intelligent insert detection method and device for workpiece | |
| CN113469991B (en) | Visual online detection method for laser welding spot of lithium battery tab | |
| CN113269234B (en) | Connecting piece assembly detection method and system based on target detection | |
| US20220284699A1 (en) | System and method of object detection using ai deep learning models | |
| CN117647527A (en) | Method and apparatus for detecting defect of glass plate, and storage medium | |
| CN109870141A (en) | The polar method and system of electrolytic capacitor are judged using image | |
| CN116486177A (en) | Underwater target identification and classification method based on deep learning | |
| CN110414470B (en) | Inspection method based on terahertz and visible light | |
| CN114392940A (en) | Stitch detection method and device for special-shaped component | |
| CN114354491A (en) | DCB ceramic substrate defect detection method based on machine vision |
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 |