CN113674242A - Rotor defect detection method based on generation countermeasure network - Google Patents
Rotor defect detection method based on generation countermeasure network Download PDFInfo
- Publication number
- CN113674242A CN113674242A CN202110955047.3A CN202110955047A CN113674242A CN 113674242 A CN113674242 A CN 113674242A CN 202110955047 A CN202110955047 A CN 202110955047A CN 113674242 A CN113674242 A CN 113674242A
- Authority
- CN
- China
- Prior art keywords
- rotor
- image
- welding
- countermeasure network
- welding spot
- 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.)
- Withdrawn
Links
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
- G06T7/001—Industrial image inspection using an image reference approach
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30152—Solder
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)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to the technical field of product defect detection, in particular to a rotor defect detection method based on a generated countermeasure network, which comprises the following steps of collecting data, classifying templates, extracting features, and verifying the features.
Description
Technical Field
The invention relates to the technical field of product defect detection, in particular to a rotor defect detection method based on a generation countermeasure network.
Background
With the development of society and rapid development of science and technology, the application of the micro motor in the fields of mobile phones, sound box products and the like is more and more extensive, and the performance requirements of people on the motor are more and more strict, which puts forward higher requirements on the rotor production detection requirements. Due to unstable factors of production equipment and manpower, unstable factors of the clamp state and the assembly process can influence the wrapping performance of the welding spot of the motor rotor, and different defects such as more welding, less welding, welding deviation, missing welding, insufficient welding and the like are generated. This requires that the electrode be baked at a specified temperature and time and that the weld zone be cleaned, the proper welding current during welding, and the welding speed be reduced to allow the gas in the weld pool to escape completely. Whether the product is a good product or not is difficult to judge in real time by naked eyes alone, and the traditional manual detection method is difficult to meet the requirements of industrial online detection in the aspects of precision and automation degree.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a rotor defect detection method based on a generation countermeasure network for detecting the welding spot detection efficiency of a motor rotor.
(II) technical scheme
In order to achieve the purpose, the invention provides the following technical scheme: a rotor defect detection method based on a generation countermeasure network comprises the following steps:
acquiring data, namely acquiring images through an industrial area-array camera and a linear array camera, wherein the images comprise three directions of a side surface, a top surface and a bottom surface;
secondly, classifying templates, collecting different welding spot defect images of the motor rotor, using the images as training samples of a motor rotor defect detection unit, and storing a plurality of detection templates in a classified manner, so that the defect detection unit can identify different types of welding spot defects;
extracting characteristics, analyzing the sample, comparing the image shape of the welding spot of the motor rotor with a preset detection template image, obtaining the matching degree of the welding spot image and a standard image, identifying different types of welding spot characteristics, and classifying the characteristics;
step four, verifying the characteristics, namely verifying the effectiveness of the characteristic shapes of the welding spots of the motor, judging whether the classified characteristic shape effects reach the standard or not, judging whether the characteristic shapes among the welding spots are obvious and reliable or not, and if the verification is available, reserving the welding spots as samples for standby
In order to improve the detection effect of the welding spot, the invention has the improvement that the welding spot defect in the step two and the characteristic in the step three are the obvious characteristics of qualified, multi-welding, few-welding, welding deviation, missing welding and insufficient welding of the extracted welding spot.
In order to improve the accurate effect of generating the characteristic images, the invention improves that the countermeasure network carries out countermeasure training with the generator and the reconstruction encoder, promotes the training of the generator and the reconstruction encoder, and optimizes the generator to generate the characteristic images of the welding spot images.
In order to improve the image generation effect, the invention improves that the image preprocessing unit comprises a gray processing unit, an edge detection unit, a morphology processing unit and an extraction unit.
In order to improve the rotor defect detection effect, the invention has the improvement that the rotor defect detection device comprises a rotor magnetic field intensity detection unit and a rotor rotating speed detection unit, wherein the rotor magnetic field intensity detection unit is a magnetic field intensity sensor, and the rotor rotating speed detection unit is a rotary encoder.
(III) advantageous effects
Compared with the prior art, the invention provides a rotor defect detection method based on a generated countermeasure network, which has the following beneficial effects:
the method comprises the steps of firstly collecting data of welding spot characteristics, generating a welding spot characteristic template library, classifying according to remarkable characteristics of qualified welding spots, excessive welding spots, insufficient welding spots, welding deviation, missed welding spots and insufficient welding spots, analyzing samples, comparing the image shape of the welding spot of the motor rotor with a preset detection template image to obtain the matching degree of the welding spot image and a standard image, identifying different types of welding spot characteristics, finally carrying out primary verification, and reserving the sample as a characteristic template for standby after verification is available, so that the efficiency of detecting the welding spot defects of the rotor is improved.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention relates to a rotor defect detection method based on a generation countermeasure network, which comprises the following steps:
acquiring data, namely acquiring images through an industrial area-array camera and a linear array camera, wherein the images comprise three directions of a side surface, a top surface and a bottom surface;
secondly, classifying templates, collecting different welding spot defect images of the motor rotor, using the images as training samples of a motor rotor defect detection unit, and storing a plurality of detection templates in a classified manner, so that the defect detection unit can identify different types of welding spot defects;
extracting characteristics, analyzing the sample, comparing the image shape of the welding spot of the motor rotor with a preset detection template image, obtaining the matching degree of the welding spot image and a standard image, identifying different types of welding spot characteristics, and classifying the characteristics;
and step four, verifying the characteristics, namely verifying the effectiveness of the characteristic shapes of the welding spots of the motor, judging whether the classified characteristic shape effects reach the standard or not, judging whether the characteristic shapes among the welding spots are obvious and reliable or not, and if the verification is available, reserving the welding spots as samples for standby.
In this embodiment, the solder joint defect in the second step and the feature in the third step are significant features of qualified solder joints, excessive solder joints, insufficient solder joints, solder offset, missing solder joints and insufficient solder joints, and corresponding categories are established to improve the solder joint detection effect.
In this embodiment, the countermeasure network, the generator and the reconstruction encoder perform countermeasure training, training of the generator and the reconstruction encoder is promoted, a feature image of a welding spot image generated by the generator is optimized, a difference between a one-dimensional vector obtained by encoding the feature image generated by the preprocessor through an encoder sub-module in the generator and a one-dimensional vector obtained by encoding the feature image generated by the generator through the reconstruction encoder is reduced, meanwhile, in the design of the countermeasure network, spectrum normalization is added to the last layer of network, the countermeasure network is stabilized in the training process, and the accurate effect of generating the feature image is improved.
In the embodiment, the image preprocessing unit comprises a gray processing unit, an edge detection unit, a morphology processing unit and an extraction unit, the gray processing unit enables image details to be rich and to be distinguished more easily, the edge detection unit selects a reasonable threshold value and adopts a Canny edge detection algorithm to extract edges, and then the characteristic image is further repaired through morphology processing and then the welding spot area is segmented by the extraction unit.
In this embodiment, including rotor magnetic field intensity detecting element and rotor speed detecting element, rotor magnetic field intensity detecting element is magnetic field intensity sensor, rotor speed detecting element is rotary encoder, realizes improving rotor defect detection effect to the detection of rotor magnetic induction intensity.
In summary, according to the rotor defect detection method based on the generation countermeasure network, when the method is used, firstly, data collection of welding spot features is carried out, a welding spot feature template library is generated, the remarkable features of qualified welding spots, excessive welding spots, few welding spots, welding deviation, missing welding spots and insufficient welding spots are classified, samples are analyzed, the image shape of the welding spots of the motor rotor is compared with a preset detection template image, the matching degree of the welding spot image and a standard image is obtained, different types of welding spot features are identified, finally, verification is carried out, and if the verification is available, the samples are reserved as feature templates for standby, so that the efficiency of detecting the welding spot defects of the rotor is improved.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (5)
1. A rotor defect detection method based on a generation countermeasure network is characterized by comprising the following steps:
acquiring data, wherein an industrial area-array camera and a linear array camera acquire images, and the images comprise three directions of a side surface, a top surface and a bottom surface;
secondly, classifying templates, collecting different welding spot defect images of the motor rotor, using the images as training samples of a motor rotor defect detection unit, and storing a plurality of detection templates in a classified manner, so that the defect detection unit can identify different types of welding spot defects;
extracting characteristics, analyzing the sample, comparing the image shape of the welding spot of the motor rotor with a preset detection template image, obtaining the matching degree of the welding spot image and a standard image, identifying different types of welding spot characteristics, and classifying the characteristics;
and step four, verifying the characteristics, namely verifying the effectiveness of the characteristic shapes of the welding spots of the motor, judging whether the classified characteristic shape effects reach the standard or not, judging whether the characteristic shapes among the welding spots are obvious and reliable or not, and if the verification is available, reserving the welding spots as samples for standby.
2. The method for detecting defects of a rotor based on a generated countermeasure network as claimed in claim 1, wherein the defects of the solder joints in the second step and the features in the third step are significant features of qualified solder joints, multiple solder joints, few solder joints, solder offsets, missing solder joints and cold solder joints.
3. The method for detecting the rotor defect based on the generation countermeasure network of claim 1, wherein the countermeasure network performs countermeasure training with the generator and the reconstruction encoder, facilitates the training of the generator and the reconstruction encoder, and optimizes the generation of the feature image of the welding spot image by the generator.
4. The rotor defect detection method based on the generation countermeasure network is characterized by comprising an image preprocessing unit, wherein the image preprocessing unit comprises a gray processing unit, an edge detection unit, a morphology processing unit and an extraction unit.
5. The method for detecting the rotor defect based on the generation countermeasure network of claim 1, comprising a rotor magnetic field strength detection unit and a rotor rotation speed detection unit, wherein the rotor magnetic field strength detection unit is a magnetic field strength sensor, and the rotor rotation speed detection unit is a rotary encoder.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110955047.3A CN113674242A (en) | 2021-08-19 | 2021-08-19 | Rotor defect detection method based on generation countermeasure network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110955047.3A CN113674242A (en) | 2021-08-19 | 2021-08-19 | Rotor defect detection method based on generation countermeasure network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113674242A true CN113674242A (en) | 2021-11-19 |
Family
ID=78544011
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110955047.3A Withdrawn CN113674242A (en) | 2021-08-19 | 2021-08-19 | Rotor defect detection method based on generation countermeasure network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113674242A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115753832A (en) * | 2022-11-10 | 2023-03-07 | 合肥华焠新能源科技有限公司 | Detection method for new energy automobile integrated circuit board |
| CN116309319A (en) * | 2023-01-29 | 2023-06-23 | 凌波微步半导体设备(常熟)有限公司 | Welding spot detection method |
| CN116551263A (en) * | 2023-07-11 | 2023-08-08 | 苏州松德激光科技有限公司 | Visual control method and system for welding position selection |
-
2021
- 2021-08-19 CN CN202110955047.3A patent/CN113674242A/en not_active Withdrawn
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115753832A (en) * | 2022-11-10 | 2023-03-07 | 合肥华焠新能源科技有限公司 | Detection method for new energy automobile integrated circuit board |
| CN115753832B (en) * | 2022-11-10 | 2023-07-11 | 合肥华焠新能源科技有限公司 | Detection method for integrated circuit board of new energy automobile |
| CN116309319A (en) * | 2023-01-29 | 2023-06-23 | 凌波微步半导体设备(常熟)有限公司 | Welding spot detection method |
| CN116551263A (en) * | 2023-07-11 | 2023-08-08 | 苏州松德激光科技有限公司 | Visual control method and system for welding position selection |
| CN116551263B (en) * | 2023-07-11 | 2023-10-31 | 苏州松德激光科技有限公司 | Visual control method and system for welding position selection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113674242A (en) | Rotor defect detection method based on generation countermeasure network | |
| CN113989279B (en) | Plastic film quality detection method based on artificial intelligence and image processing | |
| Lin et al. | Detection of a casting defect tracked by deep convolution neural network | |
| CN111369516B (en) | Transformer bushing heating defect detection method based on infrared image recognition | |
| CN107389701A (en) | A kind of PCB visual defects automatic checkout system and method based on image | |
| CN110044909B (en) | Motor rotor welding spot defect detection device and method based on image processing | |
| CN114821114A (en) | Groove cutting robot image processing method based on visual system | |
| CN107016394B (en) | Cross fiber feature point matching method | |
| CN111429423B (en) | Multi-spectral welding image identification method based on quaternion multi-degree-of-freedom neurons | |
| CN115308222B (en) | System and method for identifying poor chip appearance based on machine vision | |
| CN106599890A (en) | Transformer substation patrol robot digital type instrument identification algorithm | |
| CN115035092A (en) | Image-based bottle detection method, device, equipment and storage medium | |
| CN108109154A (en) | A kind of new positioning of workpiece and data capture method | |
| CN114594114A (en) | Full-automatic online nondestructive detection method for lithium battery cell | |
| CN112258444A (en) | Elevator steel wire rope detection method | |
| CN112734698A (en) | Cable terminal abnormity diagnosis method and device based on infrared image | |
| CN105374029B (en) | The dividing method and system of substation equipment infrared image area-of-interest | |
| CN115184380A (en) | Printed circuit board welding spot abnormity detection method based on machine vision | |
| CN110826571B (en) | Image traversal algorithm for rapid image identification and feature matching | |
| CN115601369A (en) | Quality evaluation method for power transformation equipment support for power transmission and transformation engineering | |
| CN116363122A (en) | Steel weld crack detection method and system based on artificial intelligence | |
| CN113744269B (en) | Method and device for detecting welding quality of cylindrical battery cell, electronic equipment and storage medium | |
| CN117114402A (en) | Power equipment risk assessment method based on big data | |
| CN111429411B (en) | X-ray defect image sample generation method for carbon fiber composite core wire | |
| Qian et al. | The autonomous detection of sheet metal parts using imageprocessing |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211119 |