CN112184622A - Pipe pile appearance quality inspection method and system, computer equipment and storage medium - Google Patents
Pipe pile appearance quality inspection method and system, computer equipment and storage medium Download PDFInfo
- Publication number
- CN112184622A CN112184622A CN202010894761.1A CN202010894761A CN112184622A CN 112184622 A CN112184622 A CN 112184622A CN 202010894761 A CN202010894761 A CN 202010894761A CN 112184622 A CN112184622 A CN 112184622A
- Authority
- CN
- China
- Prior art keywords
- appearance
- layer
- layers
- convolution layers
- residual
- 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
- 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
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- 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
- G06T2207/30132—Masonry; Concrete
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses a method for inspecting the appearance quality of a tubular pile, which comprises the following steps: acquiring an appearance image of a pipe pile to be detected; acquiring a target image to be inspected according to the appearance image; inputting the target image into a pre-constructed quality inspection model to obtain an inspection value; and when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be inspected is qualified. The invention also discloses a system for inspecting the appearance quality of the tubular pile, computer equipment and a storage medium. By adopting the invention, the appearance quality of the pipe pile can be automatically inspected, the working intensity can be reduced, the number of workers can be effectively reduced, the safety risk in the manual inspection process can be eliminated, and the production efficiency and the product quality can be improved.
Description
Technical Field
The invention relates to production of pipe piles, in particular to a method and a system for inspecting the appearance quality of pipe piles, computer equipment and a storage medium.
Background
In present tubular pile production process, when product output is big, the inspection project is many, the inspector can cause visual fatigue owing to use the eye for a long time on the production line of high strength repeatability work, and the inspector is difficult to avoid appearing safety problems such as colliding with under high strength operational environment, and the phenomenon of lou examining and false retrieval can appear in the inspector when visual fatigue.
Disclosure of Invention
The invention aims to solve the technical problems that a method, a system, computer equipment and a storage medium for inspecting the appearance quality of a tubular pile are provided, and the defects of high labor intensity, poor safety, low efficiency and the like of the conventional manual inspection method for the prestressed concrete tubular pile can be overcome.
In order to solve the technical problems, the invention provides a method for inspecting the appearance quality of a tubular pile, which comprises the following steps: acquiring an appearance image of a pipe pile to be detected; acquiring a target image to be inspected according to the appearance image; inputting the target image into a pre-constructed quality inspection model to obtain an inspection value; and when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be inspected is qualified.
Preferably, the method for constructing the quality inspection model includes: the method comprises the steps of inputting a picture of a tubular pile appearance sample with qualified appearance quality into a quality inspection model for training, wherein the quality inspection model comprises a first coiling layer, a first primary pooling layer, a second coiling layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are sequentially connected.
Preferably, the step of obtaining the target image to be inspected according to the appearance image comprises: inputting the appearance image into a target model to obtain a target image, wherein the target model comprises a first convolution layer, a first pooling layer, a first residual block, a second residual block, a third residual block, a fourth residual block and a global pooling layer which are sequentially connected, the target model is trained by adopting a random gradient descent algorithm to obtain model weights, and the target image is output by combining a loss function and a classification function, wherein: the channel of the first convolutional layer is 16, kernel is 7 and stride is 2; the kernel of the first pooling layer is 3, the stride is 2, and the first pooling layer is maximally pooled; the first residual block comprises two convolution layers and a residual connection, wherein channels, kernels and strides of the two convolution layers are respectively 32, 3 and 1, and the two convolution layers are sequentially connected; the second residual block comprises two convolution layers and a residual connection, wherein channels of the two convolution layers are 64, kernels of the two convolution layers are 3, and strides of the two convolution layers are 2 and 1 respectively, the two convolution layers are connected in sequence, and the residual connection comprises the convolution layers with the channels of 64, the kernels of 1 and the strides of 2; the third residual block comprises two convolutional layers and a residual connection, wherein channels of the two convolutional layers are both 128, kernel is both 3, and stride is respectively 2 and 1, the two convolutional layers are sequentially connected, and the residual connection comprises one convolutional layer with a channel of 128, a kernel of 1, and a stride of 2; the fourth residual block comprises two convolutional layers and a residual connection, wherein the channels of the two convolutional layers are 256, the kernel is 3, and the stride is 2 and 1 respectively, the two convolutional layers are connected in sequence, and the residual connection comprises the convolutional layers with the channels of 256, the kernel of 1, and the stride of 2; all the convolution layers are connected with the activation function; the batch size of the stochastic gradient descent algorithm is 32, a total of 40 epochs are trained, the initial learning rate is 0.01, and every 30 epochs learning rate is divided by 0.
Preferably, after the step of acquiring the appearance image of the pipe pile to be inspected, the method further comprises the following steps: and preprocessing the appearance image.
The invention also discloses a system for inspecting the appearance quality of the tubular pile, which comprises the following components: the system comprises an appearance image acquisition module, a target image acquisition module, an appearance quality inspection module and a judgment module; the image acquisition module is used for acquiring an appearance image of the tubular pile to be detected; the target image acquisition module is used for inputting the appearance image into a target model to obtain a target image; the appearance quality inspection module is used for inputting the target image into a pre-constructed quality inspection model and obtaining an inspection value; the judging module is used for judging whether the appearance quality of the tubular pile to be detected is qualified or not, wherein when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be detected is qualified.
Preferably, the quality inspection model includes a first convolution layer, a first sub-pooling layer, a second convolution layer, a second pooling layer, a first full-link layer and a second full-link layer, which are connected in sequence.
Preferably, the target model includes a first convolution layer, a first pooling layer, a first residual block, a second residual block, a third residual block, a fourth residual block, and a global pooling layer, which are connected in sequence, and the target model is trained by using a random gradient descent algorithm to obtain a model weight, and outputs a target image by combining a loss function and a classification function, where: the channel of the first convolutional layer is 16, kernel is 7 and stride is 2; the kernel of the first pooling layer is 3, the stride is 2, and the first pooling layer is maximally pooled; the first residual block comprises two convolution layers and a residual connection, wherein channels, kernels and strides of the two convolution layers are respectively 32, 3 and 1, and the two convolution layers are sequentially connected; the second residual block comprises two convolution layers and a residual connection, wherein channels of the two convolution layers are 64, kernels of the two convolution layers are 3, and strides of the two convolution layers are 2 and 1 respectively, the two convolution layers are connected in sequence, and the residual connection comprises the convolution layers with the channels of 64, the kernels of 1 and the strides of 2; the third residual block comprises two convolutional layers and a residual connection, wherein channels of the two convolutional layers are both 128, kernel is both 3, and stride is respectively 2 and 1, the two convolutional layers are sequentially connected, and the residual connection comprises one convolutional layer with a channel of 128, a kernel of 1, and a stride of 2; the fourth residual block comprises two convolutional layers and a residual connection, wherein the channels of the two convolutional layers are 256, the kernel is 3, and the stride is 2 and 1 respectively, the two convolutional layers are connected in sequence, and the residual connection comprises the convolutional layers with the channels of 256, the kernel of 1, and the stride of 2; all the convolution layers are connected with the activation function; the batch size of the stochastic gradient descent algorithm is 32, a total of 40 epochs are trained, the initial learning rate is 0.01, and every 30 epochs learning rate is divided by 0.
Preferably, the system for inspecting the appearance quality of the tubular pile further comprises an image preprocessing module, wherein the image preprocessing module is used for preprocessing the appearance image.
The invention also provides a computer device comprising a memory, a processor and computer instructions stored on the memory and executable on the processor, the processor implementing any of the above method steps when executing the instructions.
The present invention also provides a storage medium storing computer instructions which, when executed by a processor, implement the steps of any one of the above-described methods.
The beneficial effects of the implementation of the invention are as follows:
the invention provides a method and a system for inspecting the appearance quality of a tubular pile, computer equipment and a storage medium. The invention can automatically inspect the appearance quality of the pipe pile, thereby reducing the working intensity, effectively reducing the number of workers, eliminating the safety risk in the manual inspection process and improving the production efficiency and the product quality.
Drawings
FIG. 1 is a flow chart of a method for inspecting the appearance quality of a tubular pile provided by the present invention;
fig. 2 is a schematic diagram of a system for inspecting the appearance quality of a pipe pile provided by the invention.
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 accompanying drawings. It is only noted that the invention is intended to be limited to the specific forms set forth herein, including any reference to the drawings, as well as any other specific forms of embodiments of the invention.
As shown in fig. 1, the invention provides a method for inspecting the appearance quality of a tubular pile, which comprises the following steps:
s101, obtaining an appearance image of the pipe pile to be detected.
And S102, acquiring a target image to be checked according to the appearance image.
Since the acquired appearance image includes not only the tube stake, it is necessary to further determine the image of the tube stake that needs to be inspected.
S103, inputting the target image into a pre-constructed quality inspection model to obtain an inspection value.
And S104, when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be inspected is qualified.
According to the invention, the appearance image of the tubular pile is automatically acquired, the target image to be inspected is further determined, then the target image is input into the constructed quality inspection model for inspection, and when the inspection value is smaller than the preset threshold value, the appearance quality of the tubular pile to be inspected is qualified. The invention can automatically inspect the appearance quality of the pipe pile, thereby reducing the working intensity, effectively reducing the number of workers, eliminating the safety risk in the manual inspection process and improving the production efficiency and the product quality.
Preferably, the method for constructing the quality inspection model includes:
the method comprises the steps of inputting a picture of a tubular pile appearance sample with qualified appearance quality into a quality inspection model for training, wherein the quality inspection model comprises a first coiling layer, a first primary pooling layer, a second coiling layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are sequentially connected. According to the invention, the picture of the tubular pile appearance sample with qualified appearance quality is input into the quality inspection model for training, so that the quality inspection model capable of inspecting the appearance quality of the tubular pile can be obtained.
Preferably, the step of obtaining the target image to be inspected according to the appearance image comprises: inputting the appearance image into a target model to obtain a target image, wherein the target model comprises a first convolution layer, a first pooling layer, a first residual block, a second residual block, a third residual block, a fourth residual block and a global pooling layer which are sequentially connected, the target model is trained by adopting a random gradient descent algorithm to obtain model weights, and the target image is output by combining a loss function and a classification function, wherein: the channel of the first convolutional layer is 16, kernel is 7 and stride is 2; the kernel of the first pooling layer is 3, the stride is 2, and the first pooling layer is maximally pooled; the first residual block comprises two convolution layers and a residual connection, wherein channels, kernels and strides of the two convolution layers are respectively 32, 3 and 1, and the two convolution layers are sequentially connected; the second residual block comprises two convolution layers and a residual connection, wherein channels of the two convolution layers are 64, kernels of the two convolution layers are 3, and strides of the two convolution layers are 2 and 1 respectively, the two convolution layers are connected in sequence, and the residual connection comprises the convolution layers with the channels of 64, the kernels of 1 and the strides of 2; the third residual block comprises two convolutional layers and a residual connection, wherein channels of the two convolutional layers are both 128, kernel is both 3, and stride is respectively 2 and 1, the two convolutional layers are sequentially connected, and the residual connection comprises one convolutional layer with a channel of 128, a kernel of 1, and a stride of 2; the fourth residual block comprises two convolutional layers and a residual connection, wherein the channels of the two convolutional layers are 256, the kernel is 3, and the stride is 2 and 1 respectively, the two convolutional layers are connected in sequence, and the residual connection comprises the convolutional layers with the channels of 256, the kernel of 1, and the stride of 2; all the convolution layers are connected with the activation function; the batch size of the stochastic gradient descent algorithm is 32, a total of 40 epochs are trained, the initial learning rate is 0.01, and every 30 epochs learning rate is divided by 0.
Further, still include after the step of obtaining the outward appearance image of the tubular pile of waiting to examine: and preprocessing the appearance image. The invention can improve the quality of the appearance image by carrying out a series of image preprocessing on the appearance image.
As shown in fig. 2, the invention also discloses a system for inspecting the appearance quality of a tubular pile, which comprises: the system comprises an appearance image acquisition module 1, a target image acquisition module 2, an appearance quality inspection module 3 and a judgment module 4; the image acquisition module 1 is used for acquiring an appearance image of the tubular pile to be detected; the target image obtaining module 2 is used for inputting the appearance image into a target model to obtain a target image; the appearance quality inspection module 3 is used for inputting the target image into a pre-constructed quality inspection model and obtaining an inspection value; the judging module 4 is used for judging whether the appearance quality of the tubular pile to be inspected is qualified or not, wherein when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be inspected is qualified.
The pipe pile appearance quality inspection system provided by the invention can automatically acquire the appearance image of the pipe pile, further determine the target image to be inspected, input the target image into the constructed quality inspection model for inspection, and when the inspection value is smaller than the preset threshold value, the appearance quality of the pipe pile to be inspected is qualified. The invention can automatically inspect the appearance quality of the pipe pile, thereby reducing the working intensity, effectively reducing the number of workers, eliminating the safety risk in the manual inspection process and improving the production efficiency and the product quality.
Preferably, the quality inspection model includes a first convolution layer, a first sub-pooling layer, a second convolution layer, a second pooling layer, a first full-link layer and a second full-link layer, which are connected in sequence. According to the invention, the picture of the tubular pile appearance sample with qualified appearance quality is input into the quality inspection model for training, so that the quality inspection model capable of inspecting the appearance quality of the tubular pile can be obtained.
Preferably, the target model includes a first convolution layer, a first pooling layer, a first residual block, a second residual block, a third residual block, a fourth residual block, and a global pooling layer, which are connected in sequence, and the target model is trained by using a random gradient descent algorithm to obtain a model weight, and outputs a target image by combining a loss function and a classification function, where: the channel of the first convolutional layer is 16, kernel is 7 and stride is 2; the kernel of the first pooling layer is 3, the stride is 2, and the first pooling layer is maximally pooled; the first residual block comprises two convolution layers and a residual connection, wherein channels, kernels and strides of the two convolution layers are respectively 32, 3 and 1, and the two convolution layers are sequentially connected; the second residual block comprises two convolution layers and a residual connection, wherein channels of the two convolution layers are 64, kernels of the two convolution layers are 3, and strides of the two convolution layers are 2 and 1 respectively, the two convolution layers are connected in sequence, and the residual connection comprises the convolution layers with the channels of 64, the kernels of 1 and the strides of 2; the third residual block comprises two convolutional layers and a residual connection, wherein channels of the two convolutional layers are both 128, kernel is both 3, and stride is respectively 2 and 1, the two convolutional layers are sequentially connected, and the residual connection comprises one convolutional layer with a channel of 128, a kernel of 1, and a stride of 2; the fourth residual block comprises two convolutional layers and a residual connection, wherein the channels of the two convolutional layers are 256, the kernel is 3, and the stride is 2 and 1 respectively, the two convolutional layers are connected in sequence, and the residual connection comprises the convolutional layers with the channels of 256, the kernel of 1, and the stride of 2; all the convolution layers are connected with the activation function; the batch size of the stochastic gradient descent algorithm is 32, a total of 40 epochs are trained, the initial learning rate is 0.01, and every 30 epochs learning rate is divided by 0.
Further, the device also comprises an image preprocessing module 5, wherein the image preprocessing module is used for preprocessing the appearance image. The invention can improve the quality of the appearance image by carrying out a series of image preprocessing on the appearance image.
The invention also provides a computer device comprising a memory, a processor and computer instructions stored on the memory and executable on the processor, the processor implementing any of the above method steps when executing the instructions. The present invention also provides a storage medium storing computer instructions which, when executed by a processor, implement the steps of any one of the above-described methods.
While the present disclosure has been described in considerable detail and with particular reference to a few illustrative embodiments thereof, it is not intended to be limited to any such details or embodiments or any particular embodiments, but it is to be construed as effectively covering the intended scope of the disclosure by providing a broad, potential interpretation of such claims in view of the prior art with reference to the appended claims. Furthermore, the foregoing describes the disclosure in terms of embodiments foreseen by the inventor for which an enabling description was available, notwithstanding that insubstantial modifications of the disclosure, not presently foreseen, may nonetheless represent equivalent modifications thereto.
Claims (10)
1. The method for inspecting the appearance quality of the tubular pile is characterized by comprising the following steps:
acquiring an appearance image of a pipe pile to be detected;
acquiring a target image to be inspected according to the appearance image;
inputting the target image into a pre-constructed quality inspection model to obtain an inspection value;
and when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be inspected is qualified.
2. The method for inspecting the appearance quality of the pipe pile according to claim 1, wherein the method for constructing the quality inspection model comprises the following steps:
the method comprises the steps of inputting a picture of a tubular pile appearance sample with qualified appearance quality into a quality inspection model for training, wherein the quality inspection model comprises a first coiling layer, a first primary pooling layer, a second coiling layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are sequentially connected.
3. The method for inspecting the appearance quality of the pipe pile according to claim 2, wherein the step of obtaining the target image to be inspected according to the appearance image comprises the following steps:
inputting the appearance image into a target model to obtain a target image, wherein the target model comprises a first convolution layer, a first pooling layer, a first residual block, a second residual block, a third residual block, a fourth residual block and a global pooling layer which are sequentially connected, the target model is trained by adopting a random gradient descent algorithm to obtain model weights, and the target image is output by combining a loss function and a classification function, wherein:
the channel of the first convolutional layer is 16, kernel is 7 and stride is 2;
the kernel of the first pooling layer is 3, the stride is 2, and the first pooling layer is maximally pooled;
the first residual block comprises two convolution layers and a residual connection, wherein channels, kernels and strides of the two convolution layers are respectively 32, 3 and 1, and the two convolution layers are sequentially connected;
the second residual block comprises two convolution layers and a residual connection, wherein channels of the two convolution layers are 64, kernels of the two convolution layers are 3, and strides of the two convolution layers are 2 and 1 respectively, the two convolution layers are connected in sequence, and the residual connection comprises the convolution layers with the channels of 64, the kernels of 1 and the strides of 2;
the third residual block comprises two convolutional layers and a residual connection, wherein channels of the two convolutional layers are both 128, kernel is both 3, and stride is respectively 2 and 1, the two convolutional layers are sequentially connected, and the residual connection comprises one convolutional layer with a channel of 128, a kernel of 1, and a stride of 2;
the fourth residual block comprises two convolutional layers and a residual connection, wherein the channels of the two convolutional layers are 256, the kernel is 3, and the stride is 2 and 1 respectively, the two convolutional layers are connected in sequence, and the residual connection comprises the convolutional layers with the channels of 256, the kernel of 1, and the stride of 2;
all the convolution layers are connected with the activation function;
the batch size of the stochastic gradient descent algorithm is 32, a total of 40 epochs are trained, the initial learning rate is 0.01, and every 30 epochs learning rate is divided by 0.
4. The method for inspecting the appearance quality of the tubular pile according to claim 1, further comprising, after the step of obtaining the appearance image of the tubular pile to be inspected:
and preprocessing the appearance image.
5. Tubular pile appearance quality inspection system, its characterized in that includes: the system comprises an appearance image acquisition module, a target image acquisition module, an appearance quality inspection module and a judgment module;
the image acquisition module is used for acquiring an appearance image of the tubular pile to be detected;
the target image acquisition module is used for inputting the appearance image into a target model to obtain a target image;
the appearance quality inspection module is used for inputting the target image into a pre-constructed quality inspection model and obtaining an inspection value;
the judging module is used for judging whether the appearance quality of the tubular pile to be detected is qualified or not, wherein when the inspection value is smaller than a preset threshold value, the appearance quality of the tubular pile to be detected is qualified.
6. The system for inspecting the appearance and quality of the tubular pile according to claim 5, wherein the quality inspection model comprises a first coiling layer, a first secondary pooling layer, a second coiling layer, a second pooling layer, a first full-connection layer and a second full-connection layer which are connected in sequence.
7. The system for inspecting the appearance and quality of the tubular pile according to claim 6, wherein the target model comprises a first convolution layer, a first pooling layer, a first residual block, a second residual block, a third residual block, a fourth residual block and a global pooling layer, which are connected in sequence, the target model is trained by a stochastic gradient descent algorithm to obtain model weights, and a target image is output by combining a loss function and a classification function, wherein:
the channel of the first convolutional layer is 16, kernel is 7 and stride is 2;
the kernel of the first pooling layer is 3, the stride is 2, and the first pooling layer is maximally pooled;
the first residual block comprises two convolution layers and a residual connection, wherein channels, kernels and strides of the two convolution layers are respectively 32, 3 and 1, and the two convolution layers are sequentially connected;
the second residual block comprises two convolution layers and a residual connection, wherein channels of the two convolution layers are 64, kernels of the two convolution layers are 3, and strides of the two convolution layers are 2 and 1 respectively, the two convolution layers are connected in sequence, and the residual connection comprises the convolution layers with the channels of 64, the kernels of 1 and the strides of 2;
the third residual block comprises two convolutional layers and a residual connection, wherein channels of the two convolutional layers are both 128, kernel is both 3, and stride is respectively 2 and 1, the two convolutional layers are sequentially connected, and the residual connection comprises one convolutional layer with a channel of 128, a kernel of 1, and a stride of 2;
the fourth residual block comprises two convolutional layers and a residual connection, wherein the channels of the two convolutional layers are 256, the kernel is 3, and the stride is 2 and 1 respectively, the two convolutional layers are connected in sequence, and the residual connection comprises the convolutional layers with the channels of 256, the kernel of 1, and the stride of 2;
all the convolution layers are connected with the activation function;
the batch size of the stochastic gradient descent algorithm is 32, a total of 40 epochs are trained, the initial learning rate is 0.01, and every 30 epochs learning rate is divided by 0.
8. The tube pile appearance quality inspection system of claim 6, further comprising an image preprocessing module for preprocessing the appearance image.
9. Computer device comprising a memory, a processor and computer instructions stored on the memory and executable on the processor, characterized in that the steps of the method according to any of claims 1-4 are implemented when the processor executes the instructions.
10. Storage medium storing computer instructions, characterized in that the program is adapted to carry out the steps of the method according to any one of claims 1-4 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010894761.1A CN112184622A (en) | 2020-08-31 | 2020-08-31 | Pipe pile appearance quality inspection method and system, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010894761.1A CN112184622A (en) | 2020-08-31 | 2020-08-31 | Pipe pile appearance quality inspection method and system, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112184622A true CN112184622A (en) | 2021-01-05 |
Family
ID=73925181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010894761.1A Pending CN112184622A (en) | 2020-08-31 | 2020-08-31 | Pipe pile appearance quality inspection method and system, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112184622A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154498A (en) * | 2017-12-06 | 2018-06-12 | 深圳市智能机器人研究院 | A kind of rift defect detecting system and its implementation |
| CN108921839A (en) * | 2018-07-02 | 2018-11-30 | 北京百度网讯科技有限公司 | Continuous casting billet quality detection method, device, electronic equipment and storage medium |
| CN110363072A (en) * | 2019-05-31 | 2019-10-22 | 正和智能网络科技(广州)有限公司 | Tongue image recognition method, apparatus, computer equipment and computer readable storage medium |
| CN110517259A (en) * | 2019-08-30 | 2019-11-29 | 北京百度网讯科技有限公司 | A kind of detection method, device, equipment and the medium of product surface state |
-
2020
- 2020-08-31 CN CN202010894761.1A patent/CN112184622A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108154498A (en) * | 2017-12-06 | 2018-06-12 | 深圳市智能机器人研究院 | A kind of rift defect detecting system and its implementation |
| CN108921839A (en) * | 2018-07-02 | 2018-11-30 | 北京百度网讯科技有限公司 | Continuous casting billet quality detection method, device, electronic equipment and storage medium |
| CN110363072A (en) * | 2019-05-31 | 2019-10-22 | 正和智能网络科技(广州)有限公司 | Tongue image recognition method, apparatus, computer equipment and computer readable storage medium |
| CN110517259A (en) * | 2019-08-30 | 2019-11-29 | 北京百度网讯科技有限公司 | A kind of detection method, device, equipment and the medium of product surface state |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105389814B (en) | A kind of bubble detecting method for air-tight test | |
| CN105938563A (en) | Weld surface defect identification method based on image texture | |
| WO2018177739A1 (en) | Defect detection using ultrasound scan data | |
| CN110363337B (en) | Oil measuring method and system of oil pumping unit based on data driving | |
| WO2023082760A1 (en) | Defective picture generation method and apparatus applied to industrial quality inspection | |
| CN113177937B (en) | Improved YOLOv 4-tiny-based cloth defect detection method | |
| Rummel | Nondestructive inspection reliability history, status and future path | |
| CN112762362A (en) | Underwater pipeline leakage acoustic emission detection method based on convolutional neural network | |
| CN104181233A (en) | B ultrasound scanning detection method of strain clamp crimping defect based on feature enhancement | |
| CN113971660A (en) | Computer vision method for bridge health diagnosis and intelligent camera system | |
| CN110133016A (en) | A kind of method of welding value auxiliary X-ray diffraction detection residual stress | |
| CN107876530A (en) | One kind experiment house infrastructure intelligence cleaning method | |
| CN114119502A (en) | Method for detecting chromatic aberration and stain defects of colored texture fabric | |
| CN115561176A (en) | Water quality inversion method based on feature adaptive operation and machine learning fusion | |
| CN112184622A (en) | Pipe pile appearance quality inspection method and system, computer equipment and storage medium | |
| CN113570571A (en) | Industrial edge end power battery defect detection method and system | |
| AU2020272117B2 (en) | Method for determining the geometry of a defect and for determining a load limit | |
| CN116342542A (en) | Lightweight neural network-based steel product surface defect detection method | |
| CN116433661A (en) | Method, device, equipment and medium for detecting semiconductor wafer by multitasking | |
| Qi et al. | A CNN-based method for concreate crack detection in underwater environments | |
| CN111238927A (en) | Fatigue durability evaluation method and device, electronic equipment and computer readable medium | |
| Akhyar et al. | Lightning YOLOv4 for a surface defect detection system for sawn lumber | |
| CN108131568B (en) | A kind of artificial intelligence recognition methods of pipeline non-destructive testing | |
| CN110567905A (en) | Water quality detection method and device | |
| CN117288903B (en) | Building block type test method and system for aircraft composite material |
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 |