CN113269731A - Defect detection method and system - Google Patents
Defect detection method and system Download PDFInfo
- Publication number
- CN113269731A CN113269731A CN202110523070.5A CN202110523070A CN113269731A CN 113269731 A CN113269731 A CN 113269731A CN 202110523070 A CN202110523070 A CN 202110523070A CN 113269731 A CN113269731 A CN 113269731A
- Authority
- CN
- China
- Prior art keywords
- image
- defect
- defect detection
- image acquisition
- internal
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 100
- 238000001514 detection method Methods 0.000 title claims abstract description 83
- 210000003298 dental enamel Anatomy 0.000 claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 21
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 8
- 238000010586 diagram Methods 0.000 claims description 13
- 238000012545 processing Methods 0.000 claims description 13
- 230000008859 change Effects 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 239000000428 dust Substances 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims description 3
- 238000007794 visualization technique Methods 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 description 10
- 230000009467 reduction Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000010892 electric spark Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 239000000843 powder Substances 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000011248 coating agent Substances 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000010304 firing Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
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
- G06T7/0008—Industrial image inspection checking presence/absence
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- 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
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8854—Grading and classifying of flaws
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8883—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges involving the calculation of gauges, generating models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Quality & Reliability (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Analysis (AREA)
Abstract
A defect detection method and system, the method comprising the steps of: acquiring an internal image of a device to be detected by an image acquisition device; carrying out defect identification on the obtained internal image through a preset detection model; judging whether the obtained internal image has defects, if so, giving an alarm and storing the obtained internal image; and if the defects do not exist, continuously acquiring the internal image of the device to be detected until the internal defect detection is finished. The embodiment of the disclosure adopts the FPPN-ALN as the detection network, has high detection speed, higher detection precision and low computation, can reduce the deployment cost, and is very suitable for the detection scene of the enamel inner container of the water heater.
Description
Technical Field
The present disclosure relates to the field of defect detection technologies, and in particular, to a defect detection method and system.
Background
The enamel inner container of the water heater has the characteristics of bearing pressure, durability, corrosion resistance, shriveling prevention and the like, and can effectively ensure that the water quality is not polluted. The inspection of the finished product of the enamel liner is an important link for ensuring the product to be qualified, but the diameter of the inlet of the enamel liner is small due to the uniqueness of the liner structure, the defect of the coating surface of the enamel liner is not obvious inside the liner, and the detection is difficult. The existing detection method for the enamel liner has electric spark detection, belongs to a contact type measurement mode, and is easy to damage the enamel glaze surface and cause damage to the enamel liner. In addition, the traditional manual mode of detecting the enamel inner container by lighting a flashlight has the defects of low recognition efficiency, poor detection effect and the like.
In view of the above, it is desirable to provide a method for accurately detecting and classifying an enamel liner of a water heater.
Disclosure of Invention
In view of the above, the embodiments of the present disclosure provide a defect detection method, which at least partially solves the problems in the prior art.
In a first aspect, an embodiment of the present disclosure provides a defect detection method, which is applied to a device having a liner or a local storage space, and includes the following steps:
acquiring an internal image of a device to be detected by an image acquisition device;
carrying out defect identification on the obtained internal image through a preset detection model;
judging whether the obtained internal image has defects, if so, giving an alarm and storing the obtained internal image; and if the defects do not exist, continuously acquiring the internal image of the device to be detected until the internal defect detection is finished.
According to a specific implementation manner of the embodiment of the present disclosure, the image capturing device in the step of capturing the internal image of the device to be detected by the image capturing device includes: the telescopic mechanical support frame, the image acquisition subassembly is installed to the front end of mechanical support frame, the image acquisition subassembly includes camera and light source.
According to a specific implementation manner of the embodiment of the disclosure, the mechanical support frame comprises a first support rod and a second support rod, the first support rod and the second support rod are telescopically connected, the image acquisition assembly further comprises a camera mounting frame, the camera mounting frame is foldably installed on the second support rod, and the camera is installed on the camera mounting frame.
According to a specific implementation manner of the embodiment of the disclosure, a rotary seat is installed on the camera installation frame, and the camera is installed on the rotary seat so as to realize 360-degree rotation of the camera.
According to a specific implementation manner of the embodiment of the disclosure, the micro water heater enamel liner of the device to be detected.
According to a specific implementation manner of the embodiment of the present disclosure, the preset detection model in the step of performing defect identification on the obtained internal image through the preset detection model is a model obtained by training and optimizing a neural network, and the training method of the preset detection model includes:
selecting a device with a defect in the inner container or the internal storage space, and acquiring an image of the inner container or the internal storage space of the device through an image acquisition device to obtain an image sample with the defect;
the obtained image sample with the defects is transmitted to an FPPN network, and a feature map of the input image sample is obtained by a feature response visualization method through 4-layer convolution operation;
according to the change of the response characteristic diagram, the FPPN network obtains an area with violent change of the characteristic diagram in an input image sample through a sliding window method;
and processing the area with the severely changed characteristic diagram, transmitting the area into an ALN network, and outputting a defect identification result.
According to a specific implementation manner of the embodiment of the present disclosure, the method for processing the area with the drastic change of the feature map in the step of conveying the area with the drastic change of the feature map to the ALN network after processing the area with the drastic change of the feature map comprises: the size of the regions with the violent change of the feature map is unified into 112 x 112, and the regions unified into the preset value size are amplified to 224 x 224 by a bilinear interpolation method.
According to a specific implementation manner of the embodiment of the present disclosure, the processing and transmitting the region with the severely changed feature map to the ALN network, and the step of outputting the defect identification result, in which the defect identification result output by the ALN network includes: the surface of the inner container is damaged, the corners are damaged, and dust is adhered and normal.
In addition, in order to achieve the above object, the embodiments of the present disclosure further provide a defect detection system, where the defect detection system is configured to implement the defect detection method as described above;
the defect detection system comprises an image acquisition device and a defect detection control device, wherein the image acquisition device comprises a telescopic mechanical support frame, the front end of the mechanical support frame is provided with an image acquisition assembly, and the image acquisition assembly comprises a camera and a light source;
the defect detection control device comprises a control module connected with the image acquisition device, an information collection module used for receiving the image acquired by the image acquisition device, and a processing module used for processing the defect information of the acquired image.
The embodiment of the disclosure adopts the FPPN-ALN as the detection network, has the advantages of high detection speed, higher detection precision, low computation amount, and reduced deployment cost, and is very suitable for the detection scene of the enamel inner container of the water heater. And a large number of enamel liner defect images are used for making a data set for neural network training, so that the training of the FPPN-ALN network is completed. In consideration of different conditions such as various visual defect shapes, light reflecting degrees and the like, the images in the data set used by the invention contain different illumination and angles, thereby ensuring that the trained neural network has good robustness. The embodiment of the invention adopts the image-based detection method, and can detect the position and the type of the visual defect of the enamel liner in real time with higher accuracy, so that the system detection means has the advantages of non-contact, anti-interference, real-time, accuracy and the like; meanwhile, mechanical omnibearing scanning detection is adopted, and a visual field blind area does not exist, so that the output result of the detection system is closer to the real condition.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed to be used in the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present disclosure, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a defect detection method according to an embodiment of the present disclosure;
FIG. 2 is a diagram illustrating an overall structure of a defect detection system according to an embodiment of the present disclosure;
fig. 3(a) is a schematic diagram of a position of a camera facing the bottom of an enamel liner in a defect detection system provided in the embodiment of the present disclosure;
fig. 3(b) is a schematic diagram of a position of a camera facing a sidewall of an enamel liner in a defect detection system provided in the embodiment of the present disclosure;
FIG. 3(c) is a schematic diagram illustrating the pivoting of a camera in a defect detection system provided by an embodiment of the present disclosure;
FIG. 4 is a schematic view illustrating a telescopic rotation of a camera in a defect detection system according to an embodiment of the present disclosure;
fig. 5 is a schematic flow chart of a defect detection method according to an embodiment of the present disclosure.
Summary of reference numerals:
10-a device to be detected, 20-a defect detection system, 21-a camera, 22-a light source and 23-a mechanical support frame.
Detailed Description
The embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
The embodiments of the present disclosure are described below with specific examples, and other advantages and effects of the present disclosure will be readily apparent to those skilled in the art from the disclosure in the specification. It is to be understood that the described embodiments are merely illustrative of some, and not restrictive, of the embodiments of the disclosure. The disclosure may be embodied or carried out in various other specific embodiments, and various modifications and changes may be made in the details within the description without departing from the spirit of the disclosure. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
It is noted that various aspects of the embodiments are described below within the scope of the appended claims. It should be apparent that the aspects described herein may be embodied in a wide variety of forms and that any specific structure and/or function described herein is merely illustrative. Based on the disclosure, one skilled in the art should appreciate that one aspect described herein may be implemented independently of any other aspects and that two or more of these aspects may be combined in various ways. For example, an apparatus may be implemented and/or a method practiced using any number of the aspects set forth herein. Additionally, such an apparatus may be implemented and/or such a method may be practiced using other structure and/or functionality in addition to one or more of the aspects set forth herein.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present disclosure, and the drawings only show the components related to the present disclosure rather than the number, shape and size of the components in actual implementation, and the type, amount and ratio of the components in actual implementation may be changed arbitrarily, and the layout of the components may be more complicated.
In addition, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects may be practiced without these specific details.
The embodiment of the disclosure provides a defect detection method.
Step S10, collecting the internal image of the device to be detected through an image collecting device;
step S20, defect recognition is carried out on the obtained internal image through a preset detection model;
step S30, judging whether the obtained internal image has defects, if so, sending an alarm and storing the obtained internal image; and if the defects do not exist, continuously acquiring the internal image of the device to be detected until the internal defect detection is finished.
The embodiment of the disclosure adopts the FPPN-ALN as the detection network, has the advantages of high detection speed, higher detection precision, low computation amount, and reduced deployment cost, and is very suitable for the detection scene of the enamel inner container of the water heater. And a large number of enamel liner defect images are used for making a data set for neural network training, so that the training of the FPPN-ALN network is completed. In consideration of different conditions such as various visual defect shapes, light reflecting degrees and the like, the images in the data set used by the invention contain different illumination and angles, thereby ensuring that the trained neural network has good robustness.
When the enamel inner container of the thermos bottle has visual defects, a plurality of appearances such as cracks, dark spots and the like can appear. Because the problems of visual field blind area and low efficiency exist in manual detection, whether the surface of the liner has visual defects or not can not be well reflected only by manually observing the surface of the liner, the invention firstly utilizes the convolutional neural network to detect the enamel liner, and the detection result is used as the judgment basis for whether the visual defects exist or not. The existing detection methods are relatively inefficient and have disadvantages, and the method is more rapid and convenient and has higher accuracy.
The traditional contact type measuring and detecting method such as electric spark detection and the like has high equipment cost and low detection efficiency, and is easy to cause secondary damage to the enamel liner, but the invention adopts the image-based detecting method, can detect the position and the type of the visual defect of the enamel liner in real time with higher accuracy, so that the system detecting means has the advantages of non-contact, anti-interference, real-time, accuracy and the like; meanwhile, mechanical omnibearing scanning detection is adopted, and a visual field blind area does not exist, so that the output result of the detection system is closer to the real condition.
Referring to fig. 1, a defect detection method provided by the embodiment of the present disclosure is applied to a device having a liner or a local storage space, and specifically:
the input of the system realized by the invention is 8Bit true color pictures acquired by a high-definition industrial camera, the resolution of the input image is 768 multiplied by 576, and the output is defect classification and position information. The deep learning-vision defect detection system is the core of the whole device and completes defect detection and defect classification. The PLC is mainly used for completing motion control of lead screw transmission and control of protection devices such as gratings.
The machine vision system deep learning-visual defect inspection is used for preprocessing and detecting defects of the image, the detection result is stored in a local server, defect classification and position information is given on a display interface, and a detection conclusion is displayed: OK or NG.
The specific defect detection comprises four steps of sample collection, simulated sample generation, model training and model verification:
a. sample collection
A certain number of various defect samples are collected, and defect types and regions are labeled.
b. Sample generation
Because of the large number of types of samples per defect, the number of samples is limited, and it takes a long time to rely on only collected samples. The network generated based on the deep confrontation can be used for generating missing item samples, assisting training and improving the accuracy of the detection model.
c. Model training
Training a deep neural network model based on the collected and generated samples, and detecting the types and the areas of the defects.
d. Processing of detection results
And recording the detected size and displaying and reserving the defect picture and the alarm information on the screen, and storing the defect picture.
In the step a, the liner of the thermos bottle with defects is selected by a manual detection method, the high-definition industrial camera is lifted into the liner of the thermos bottle by the mechanical structure of the invention, and various defect samples are captured in a 360-degree omnibearing manner. The finally captured image data is a three-dimensional tensor (number of channels × width × height) of 3 × 768 × 576.
In the step b, as the actual inner container sample data containing defects is limited, 2000 actual samples are obtained, and in order to guarantee the generalization of the model, more positive samples are generated by adopting a mode of generating the network cycleGAN by deep confrontation, and the number of the defect samples is increased.
In step c, as shown in fig. 5, the input picture is first sent to the FPPN network, and through 4-layer convolution operation, the feature map of the input picture is obtained by using the feature response visualization technology, where the part with severe color change represents that the picture contains more high-frequency components, and the content of the picture changes obviously, and these areas are often just the parts with defects. Therefore, according to the change of the response characteristic diagram, the FPPN network obtains the areas with the characteristic diagram changing seriously in the original input picture by the sliding window method, and the sizes of the areas are unified into 112 multiplied by 112. Secondly, in order to further increase the discrimination of defect types, the original 112 × 112 area is enlarged to 224 × 224 by a bilinear interpolation method, and then is sent to a subsequent ALN network.
The output results of the ALN network include four categories: surface break (Surface break), corner break (cornerbreake), dust adhesion (Powder stuck), normal (OK). The invention defines Surface breakthrough as '1', Cornerbereakage as '2', Powder Sticked as '3' and OK as '4', as shown in Table 1:
TABLE 1 Defect type vs. TAG
| Type (B) | Label label |
| Surfacebreakage | 1 |
| Cornerbreakage | 2 |
| PowderSticked | 3 |
| OK | 4 |
The ALN mainly refers to a GoogleLeNet model, and the structure of the model is as follows:
0. inputting: the original input image is 224 × 224 × 3, and all are subjected to a preprocessing operation of zero averaging (image per pixel minus mean).
1. First layer (coiled layer)
Using a convolution kernel of 7 × 7 (sliding step 2, padding is 3), 64 channels, the output is 112 × 112 × 64, performing the ReLU operation after convolution, and after max posing of 3 × 3 (step 2), the output is ((112-3 +1)/2) +1 ═ 56, i.e. 56 × 56 × 64, and then performing the ReLU operation;
2. second layer (convolution layer)
Using a convolution kernel of 3 × 3 (the sliding step is 1, padding is 1), 192 channels, the output is 56 × 56 × 192, performing the ReLU operation after convolution, and after max firing of 3 × 3 (the step is 2), the output is ((56-3+1)/2) +1 ═ 28, namely 28 × 28 × 192, and then performing the ReLU operation;
3a, third layer (inclusion 3a layer)
The method is divided into four branches, and convolution kernels with different scales are adopted for processing:
(1)64 convolution kernels of 1 × 1, then RuLU, output 28 × 28 × 64;
(2)96 convolution kernels of 1 × 1, which are used as dimensionality reduction before a convolution kernel of 3 × 3, become 28 × 28 × 96, then ReLU calculation is carried out, 128 convolutions of 3 × 3 are carried out (padding is 1), and 28 × 28 × 128 is output;
(3)16 convolution kernels of 1 × 1, which are used as dimensionality reduction before a convolution kernel of 5 × 5, are changed into 28 × 28 × 16, after the ReLU calculation, 32 convolutions of 5 × 5 are performed (padding is 2), and 28 × 28 × 32 is output;
(4) the pool layer outputs 28 × 28 × 192 using a 3 × 3 kernel (1 padding), and then 32 convolutions of 1 × 1 are performed to output 28 × 28 × 32.
The four results are connected, and the third dimension of the four output results is connected in parallel, namely 64+128+32+32 is 256, and finally 28 × 28 × 256 is output.
3b, third layer (inclusion 3b layer)
(1)128 convolution kernels of 1 × 1, then RuLU, output 28 × 28 × 128;
(2)128 convolution kernels of 1 × 1, which are used as dimensionality reduction before the convolution kernel of 3 × 3, become 28 × 28 × 128, perform ReLU, perform 192 convolution of 3 × 3 (padding is 1), and output 28 × 28 × 192;
(3)32 convolution kernels of 1 × 1, which are used as dimensionality reduction before a convolution kernel of 5 × 5, are changed into 28 × 28 × 32, and after the ReLU calculation, 96 convolution of 5 × 5 (padding is 2) are carried out, and 28 × 28 × 96 is output;
(4) the pool layer outputs 28 × 28 × 256 using a 3 × 3 kernel (1 padding), and then 64 convolutions of 1 × 1 are performed to output 28 × 28 × 64.
The four results are connected, and the third dimension of the four output results is connected in parallel, namely 128+192+96+64 is 480, and the final output is 28 × 28 × 480.
Fourth (4a,4b,4c,4d,4e), fifth (5a,5b) … …, similar to 3a, 3b, the final FC layer is modified to have an output of 4.
The above description is only for the specific embodiments of the present disclosure, but the scope of the present disclosure is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present disclosure should be covered within the scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be subject to the protection scope of the claims.
Claims (9)
1. A defect detection method for a device having a liner or a local storage space, the method comprising:
acquiring an internal image of a device to be detected by an image acquisition device;
carrying out defect identification on the obtained internal image through a preset detection model;
judging whether the obtained internal image has defects, if so, giving an alarm and storing the obtained internal image; and if the defects do not exist, continuously acquiring the internal image of the device to be detected until the internal defect detection is finished.
2. The defect inspection method according to claim 1, wherein the image capturing device in the step of capturing the internal image of the device to be inspected by the image capturing device comprises: the telescopic mechanical support frame, the image acquisition subassembly is installed to the front end of mechanical support frame, the image acquisition subassembly includes camera and light source.
3. The defect detection method of claim 2, wherein the mechanical support comprises a first support bar and a second support bar, the first support bar and the second support bar being telescopically coupled, the image capture assembly further comprising a camera mount, the camera mount being foldably mounted on the second support bar, the camera being mounted on the camera mount.
4. The method of claim 3, wherein a swivel is mounted to the camera mount and the camera is mounted to the swivel to enable 360 ° rotation of the camera.
5. The defect detection method of any one of claims 1 to 4, wherein the device to be detected is an enamel liner of the micro-water heater.
6. The method for detecting the defects according to any one of claims 1 to 4, wherein the preset detection model in the step of identifying the defects of the acquired internal images through the preset detection model is a model obtained by training and optimizing a neural network, and the training method of the preset detection model comprises the following steps:
selecting a device with a defect in the inner container or the internal storage space, and acquiring an image of the inner container or the internal storage space of the device through an image acquisition device to obtain an image sample with the defect;
the obtained image sample with the defects is transmitted to an FPPN network, and a feature map of the input image sample is obtained by a feature response visualization method through 4-layer convolution operation;
according to the change of the response characteristic diagram, the FPPN network obtains an area with violent change of the characteristic diagram in an input image sample through a sliding window method;
and processing the area with the severely changed characteristic diagram, transmitting the area into an ALN network, and outputting a defect identification result.
7. The defect detection method according to claim 6, wherein the step of processing the region with the drastic change of the feature map and then transmitting the processed region into the ALN network comprises the following steps: the size of the regions with the violent change of the feature map is unified into 112 x 112, and the regions unified into the preset value size are amplified to 224 x 224 by a bilinear interpolation method.
8. The defect detection method according to claim 6, wherein the processing is performed on the region with the severely changed characteristic diagram, and then the region is transmitted to the ALN network, and the step of outputting the defect identification result includes the step of outputting the defect identification result output by the ALN network: the surface of the inner container is damaged, the corners are damaged, and dust is adhered and normal.
9. A defect detection system, wherein the defect detection system is used for implementing the defect detection method according to any one of claims 1-8;
the defect detection system comprises an image acquisition device and a defect detection control device, wherein the image acquisition device comprises a telescopic mechanical support frame, the front end of the mechanical support frame is provided with an image acquisition assembly, and the image acquisition assembly comprises a camera and a light source;
the defect detection control device comprises a control module connected with the image acquisition device, an information collection module used for receiving the image acquired by the image acquisition device, and a processing module used for processing the defect information of the acquired image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110523070.5A CN113269731A (en) | 2021-05-13 | 2021-05-13 | Defect detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110523070.5A CN113269731A (en) | 2021-05-13 | 2021-05-13 | Defect detection method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113269731A true CN113269731A (en) | 2021-08-17 |
Family
ID=77230638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110523070.5A Pending CN113269731A (en) | 2021-05-13 | 2021-05-13 | Defect detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113269731A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018000731A1 (en) * | 2016-06-28 | 2018-01-04 | 华南理工大学 | Method for automatically detecting curved surface defect and device thereof |
| CN110261407A (en) * | 2019-05-23 | 2019-09-20 | 北京工业大学 | A kind of rotary full scan water heater liner surface defect detection apparatus and method |
| CN110779937A (en) * | 2019-10-11 | 2020-02-11 | 上海航天精密机械研究所 | Casting product internal defect intelligent detection device |
| CN110956619A (en) * | 2019-11-25 | 2020-04-03 | 厦门大学 | Curved glass defect detection method |
-
2021
- 2021-05-13 CN CN202110523070.5A patent/CN113269731A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018000731A1 (en) * | 2016-06-28 | 2018-01-04 | 华南理工大学 | Method for automatically detecting curved surface defect and device thereof |
| CN110261407A (en) * | 2019-05-23 | 2019-09-20 | 北京工业大学 | A kind of rotary full scan water heater liner surface defect detection apparatus and method |
| CN110779937A (en) * | 2019-10-11 | 2020-02-11 | 上海航天精密机械研究所 | Casting product internal defect intelligent detection device |
| CN110956619A (en) * | 2019-11-25 | 2020-04-03 | 厦门大学 | Curved glass defect detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102254773B1 (en) | Automatic decision and classification system for each defects of building components using image information, and method for the same | |
| CN108009515B (en) | Power transmission line positioning and identifying method of unmanned aerial vehicle aerial image based on FCN | |
| CN108765416A (en) | PCB surface defect inspection method and device based on fast geometric alignment | |
| CN109376591B (en) | Ship target detection method for deep learning feature and visual feature combined training | |
| CN111598098B (en) | Water gauge water line detection and effectiveness identification method based on full convolution neural network | |
| CN208207914U (en) | PCB surface defect detecting device based on fast geometric alignment | |
| CN110346699B (en) | Insulator discharge information extraction method and device based on ultraviolet image processing technology | |
| CN110033431B (en) | Non-contact detection device and detection method for detecting corrosion area on surface of steel bridge | |
| CN111667455A (en) | AI detection method for various defects of brush | |
| CN112304960B (en) | High-resolution image object surface defect detection method based on deep learning | |
| CN111398291A (en) | Flat enameled electromagnetic wire surface flaw detection method based on deep learning | |
| CN114719749B (en) | Metal surface crack detection and real size measurement method and system based on machine vision | |
| CN112200808B (en) | Strip steel surface defect detection method based on local Gini coefficient | |
| CN111638218A (en) | Method for detecting surface defects of coating | |
| CN114119591A (en) | Display screen picture quality detection method | |
| CN111157532A (en) | Visual detection device and method for scratches of mobile phone shell | |
| CN111426693A (en) | Quality defect detection system and detection method thereof | |
| CN117455843A (en) | Intelligent cable head defect detection system | |
| Wang et al. | Design of machine vision applications in detection of defects in high-speed bar copper | |
| CN114674830A (en) | Bottle cap flaw detection module on high-speed production line | |
| CN112750113B (en) | Glass bottle defect detection method and device based on deep learning and linear detection | |
| CN111105413B (en) | Intelligent spark plug appearance defect detection system | |
| CN112883969B (en) | Rainfall intensity detection method based on convolutional neural network | |
| CN113269731A (en) | Defect detection method and system | |
| CN103091332B (en) | Detection method and detection system of U-shaped powder pipe 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 | ||
| CB02 | Change of applicant information |
Address after: 215100 First to Second Floors of No.2 Factory Building, No. 99 Xianglu Road, Xiangcheng District Resort (Yangchenghu Town), Suzhou City, Jiangsu Province Applicant after: Suzhou Zhongke Dihong Artificial Intelligence Technology Co.,Ltd. Address before: Room 329, No. 56, Jinzhai Road, Yangchenghu Town, Xiangcheng District, Suzhou, Jiangsu 215000 Applicant before: Suzhou Dihong Artificial Intelligence Technology Co.,Ltd. |
|
| CB02 | Change of applicant information |