CN112669253A - Building material strength grade detection method, device, equipment and readable storage medium - Google Patents
Building material strength grade detection method, device, equipment and readable storage medium Download PDFInfo
- Publication number
- CN112669253A CN112669253A CN201910984098.1A CN201910984098A CN112669253A CN 112669253 A CN112669253 A CN 112669253A CN 201910984098 A CN201910984098 A CN 201910984098A CN 112669253 A CN112669253 A CN 112669253A
- Authority
- CN
- China
- Prior art keywords
- strength grade
- building material
- image
- model
- material strength
- 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
Abstract
The invention discloses a method, a device, equipment and a readable storage medium for detecting the strength grade of a building material, wherein the method comprises the following steps: building a building material strength grade detection model; acquiring a structural image of a building material to be detected; inputting the obtained structural image of the building material to be detected into a constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected; the device comprises the following modules: a model pre-construction module; an image acquisition module; an intensity level calculation module; a model verification module; the invention can greatly improve the detection precision and the detection efficiency, is suitable for detecting the strength grade of various material structures, has stronger universality, and solves the problems that the rebound method has low detection precision, the core drilling method and the pulling-out method are easy to cause internal damage of building materials, and the detection precision of the ultrasonic detection method is greatly influenced by signal frequency, member size and the like.
Description
Technical Field
The invention relates to the field of image recognition, in particular to a method, a device and equipment for detecting strength grade of a building material and a readable storage medium.
Background
The building industry of today needs to apply the building material widely, however, there is a phenomenon that more building materials are sub-fully good in the actual engineering project, and the intensity level thereof cannot meet the building material intensity standard required by the state, so that the detection of the intensity level of the building material is very necessary, and the intensity detection technology thereof is widely applied to the aspects of building material construction quality control, acceptance, identification, evaluation and the like.
The traditional method for detecting the strength grade of the building material is a destructive detection method or a nondestructive detection method. Wherein, the destructive detection method comprises a core drilling method and a drawing method; non-destructive testing includes springback and ultrasonography.
However, the destructive detection method is not only cumbersome and complex to operate, but also can damage the structure; the rebound method requires no defect in the building material and low precision, and the detection precision of the ultrasonic detection method is greatly influenced by signal frequency, member size and the like.
Disclosure of Invention
In view of the above-mentioned shortcomings of the prior art, the invention provides a building material strength grade detection method, device, equipment and readable storage medium, which can solve the problems that the rebound method is low in detection precision, the core drilling method and the pulling-out method are easy to cause internal damage of the building material, and the detection precision of the ultrasonic detection method is greatly influenced by signal frequency, component size and the like. And the detection precision is high, the detection efficiency is high, the operation is simple, and the structure is not damaged.
In order to achieve the above purpose, the embodiment of the invention adopts the following technical scheme:
a building material strength grade detection method based on image recognition comprises the following steps:
building a building material strength grade detection model;
acquiring a structural image of a building material to be detected;
inputting the obtained structural image of the building material to be detected into a constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected;
the method for constructing the building material strength grade detection model comprises the following steps: training the neural network model by using a training sample set to obtain a building material strength grade detection model;
wherein the training sample set comprises a plurality of sample images of a test member of at least one material at different intensity levels, each sample image being pre-labeled with a respective intensity level.
According to one aspect of the present invention, before inputting the structural image to be tested into the pre-constructed building material strength grade detection model, the method further comprises the following steps: and verifying the building material strength grade detection model.
According to one aspect of the invention, the verifying the building material strength grade detection model comprises the following steps:
obtaining a model verification data set, wherein the model verification data set comprises a plurality of verification sample images, and each verification sample image is an image of a test component with a known actual strength grade and the same as the structural material to be tested;
inputting each verification sample image into the building material strength grade detection model to obtain the predicted strength grade of each verification sample image;
calculating the accuracy of the building material strength grade detection model based on the predicted strength grade, the known actual strength grade and the total number of the verification sample images of each verification sample image;
judging whether the accuracy of the building material strength grade detection model is not less than a preset threshold value or not;
if the accuracy of the building material strength grade detection model is not smaller than a preset threshold value, using the building material strength grade detection model for subsequently calculating the strength grade of the structure to be detected;
and if the accuracy of the building material strength grade detection model is smaller than a preset threshold value, adding the sample images in the training sample set, and retraining the building material strength grade detection model until the accuracy is not smaller than the threshold value.
According to one aspect of the invention, the calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images comprises the following steps:
counting the unqualified number of the verification sample images of which the difference value between the predicted intensity level and the known actual intensity level is greater than a preset deviation value;
and calculating the ratio of the number of the disqualification to the total number to serve as the accuracy of the building material strength grade detection model.
According to one aspect of the invention, the method for obtaining the building material strength grade detection model by training the neural network model by using the training sample set specifically comprises the following steps: training a Faster-RCNN model by using the training sample set based on a transfer learning method to obtain a building material strength grade detection model; the training sample set comprises a plurality of sample images of test members made of the same materials as the building materials to be tested and in different strength grades, and each sample image is labeled with a corresponding strength grade in advance.
According to one aspect of the present invention, the training of the fast-RCNN model using the training sample set further comprises the following steps: and converting each sample image in the training sample set into a data set in a format convenient for deep learning.
A building material strength grade detection device based on image recognition comprises the following modules:
the model pre-construction module is used for training the neural network model to obtain a building material strength grade detection model;
the image acquisition module is used for acquiring an image of the structure to be detected;
and the strength grade calculation module is used for inputting the structural image to be detected into the building material strength grade detection model to obtain the strength grade of the structure to be detected.
According to an aspect of the present invention, the apparatus for detecting strength grade of construction material based on image recognition further comprises a model verification module, the model verification module comprising:
the model verification data set acquisition sub-module is used for acquiring a model verification data set, the model verification data set comprises a plurality of verification sample images, and each verification sample image is an image of a test component with a known actual strength grade and the same as the structural material to be tested;
the prediction strength grade calculation submodule is used for inputting each verification sample image into the building material strength grade detection model to obtain the prediction strength grade of each verification sample image;
the accuracy calculation submodule is used for calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images;
and the model retraining submodule is used for increasing the sample images in the training sample set when the accuracy of the building material strength grade detection model is smaller than a preset threshold value, and retraining the building material strength grade detection model until the accuracy is not smaller than the threshold value.
A building material strength grade detection device based on image recognition comprises an image acquisition device and a processor, wherein the processor is used for realizing the steps of the building material strength grade detection method based on image recognition when executing a computer program stored in a memory.
A computer-readable storage medium on which the image recognition-based construction material strength level detection program is stored, the image recognition-based construction material strength level detection program, when executed by a processor, implementing the steps of the image recognition-based construction material strength level detection method.
The implementation of the invention has the advantages that: the building material strength grade detection method based on image recognition comprises the following steps of: building a building material strength grade detection model; acquiring a structural image of a building material to be detected; inputting the obtained structural image of the building material to be detected into a constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected; the building material strength grade detection device based on image recognition comprises the following modules: a model pre-construction module; an image acquisition module; an intensity level calculation module; the invention can greatly improve the detection precision and the detection efficiency, is suitable for detecting the strength grade of various material structures, has stronger universality, and solves the problems that the rebound method has low detection precision, the core drilling method and the pulling-out method are easy to cause internal damage of building materials, and the detection precision of the ultrasonic detection method is greatly influenced by signal frequency, member size and the like. In addition, the invention also provides a corresponding implementation device, equipment and a computer readable storage medium for the method for detecting the concrete strength grade based on image recognition, so that the method has higher practicability, and the device, the equipment and the computer readable storage medium have corresponding advantages.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic diagram of a building material strength grade detection method based on image recognition according to a first embodiment of the present invention;
fig. 2 is a schematic diagram of a building material strength grade detection method based on image recognition according to a second embodiment of the present invention;
fig. 3 is a schematic diagram of a building material strength grade detection apparatus based on image recognition according to a third embodiment of the present invention;
fig. 4 is a schematic diagram of an apparatus for detecting strength grade of building material based on image recognition according to a fourth embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in 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.
Example one
A building material strength grade detection method based on image recognition comprises the following steps:
s101: building a building material strength grade detection model;
in practical application, the building material strength grade detection model is constructed by the following steps: and training the neural network model by utilizing the training sample set to obtain a building material strength grade detection model.
In practical applications, the training sample set comprises a plurality of sample images of a test member of at least one material at different intensity levels, and each sample image is pre-labeled with a corresponding intensity level.
In this embodiment, the test member may adopt a material having the same structure as that of the building material to be tested, and the sample image of the test member labeled with the strength level is input to the neural network model for training, so as to establish a correspondence between the strength level of the building material to be tested and the picture characteristics.
In this embodiment, the test member may also be made of any one of the building materials, the corresponding relationship between the strength level of the test member material and the image feature is established first, and the corresponding relationship between the strength level of the building material to be tested and the image feature is established by deep learning migration after analysis and comparison.
In practical applications, the intensity level of the sample image can be obtained by any related technology, or by referring to the mixing ratio data referred to when the building material is processed.
In practical application, a large number of test components at least comprising one material can be manufactured in advance, and images of the test components under different strength grade marks are acquired by using image acquisition equipment to serve as sample images; the richer the strength grade and the more the sample images are, the higher the detection accuracy and precision of the building material strength grade detection model obtained by training is.
In this embodiment, the image capturing device may be a Digital Microscope (DM) which can quantitatively magnify the shooting times, effectively capture the local features of the material, accurately represent the brightness and color range of the structure, and output a high-quality picture.
In practical application, a Faster-RCNN model can be designed in advance, and the weights of the fast-RCNN model which is designed in advance are migrated to a used neural network through adjustment and verification by adopting a migration learning method.
In practical application, the training sample set can be used for training the Faster-RCNN model and then migrating the fast-RCNN model to the neural network to obtain the building material strength grade detection model.
In practical applications, each sample image in the training sample set may be converted into a data set in a format convenient for deep learning, specifically, in a voc 2007 format, so as to serve as an intensity level image feature of each sample image.
S102: acquiring an image of a structure to be detected;
in practical application, any image acquisition equipment can be adopted to acquire an image of a structure to be detected, and then the acquired image is sent to a system.
In practical application, the acquisition of the sample image and the structural image to be detected can be realized by arranging an image acquisition device and an image acquisition card; the image acquisition card is used for sampling and quantizing image signals into digital signals of images, then the digital video signals are sent to a frame memory or a computer memory for processing, and the mode of the image acquisition card can realize high sampling and transmission speed, thereby achieving high resolution and real-time performance.
S103: and inputting the obtained structural image of the building material to be detected into the constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected.
In practical application, the structural image to be detected collected in real time is transmitted to the trained building material strength grade detection model through wired or wireless connection, and the result output by the building material strength grade detection model is the strength grade of the structure to be detected, so that the detection of the true strength grade of the structure is realized.
In practical application, as shown in fig. 1, the specific implementation of this embodiment is as follows:
in practical applications, the building material described in this embodiment is described by taking concrete as an example, and other building materials can be identified by the method steps described in this embodiment, only the training sample is the corresponding building material to be tested.
Step S101: training a neural network model by utilizing a training sample set in advance to obtain a concrete strength grade detection model;
step S102: acquiring an image of a structure to be detected;
step S103: and inputting the structural image to be detected into the concrete strength grade detection model to obtain the strength grade of the concrete.
The embodiment provides a building material strength grade detection method based on image recognition, which comprises the following steps: building a building material strength grade detection model; acquiring an image of a structure to be detected; and inputting the obtained structural image of the building material to be detected into the constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected. The strength grade detection model of the building material is trained by using the sample set, and the strength grade of the structure to be detected can be accurately obtained by using the strength grade detection model of the building material.
Example two
A building material strength grade detection method based on image recognition comprises the following steps:
s201: building a building material strength grade detection model;
in practical applications, the same technical solutions are adopted in this embodiment and the first embodiment.
S202: obtaining a model verification dataset;
in practical application, the model verification data set comprises a plurality of verification sample images, each verification sample image is an image of a test component with a known actual strength grade and the same as the structural material to be tested, and the verification sample images and the sample images in the training sample plate set are images generated by the same method.
In practical application, the sample images in the training sample set can be divided into two parts, one part is used for training the neural network model, and the other part is used as the verification sample image.
S203: inputting each verification sample image into a building material strength grade detection model to obtain the predicted strength grade of each verification sample image;
in practical application, the predicted strength grade is the strength grade of each verification sample image predicted by the building material strength grade detection model.
S204: and calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images.
In practical application, for some application scenes with low requirement precision, when the difference value between the predicted intensity level and the actual intensity level is in allowable deviation, the predicted intensity level and the actual intensity level can be considered to be equivalent, namely, a default error does not exist; therefore, in a specific embodiment, the unqualified number of the verification sample images with the difference value between the predicted intensity level and the actual intensity level larger than the preset deviation value can be counted; and calculating the ratio of the number of disqualified products to the total number of the disqualified products to serve as the accuracy of the building material strength grade detection model.
S205: judging whether the accuracy of the building material strength grade detection model is not less than a preset threshold value, if not, executing a step S206; if yes, go to step S207.
In practical applications, the preset deviation value and the preset threshold value may be set according to the requirement of the detection precision of a practical application scene, for example, in a high-precision detection scene, the preset threshold value is set to 0, that is, if the difference between the predicted intensity level and the actual intensity level of one verification sample image is greater than the preset deviation value, it is determined that the accuracy of the building material intensity level detection model does not pass, and the building material intensity level detection model needs to be retrained.
S206: the sample images in the training sample set are added, and the process returns to step S201.
In practical application, when the building material strength grade detection model fails to be verified, a plurality of sample images can be added on the basis of a pre-training sample set, and the building material strength grade detection model is retrained until the accuracy is not less than the preset threshold.
In practical application, the number of the increased sample images can be determined according to the accuracy of the building material strength grade detection model and the required detection precision.
S207: acquiring an image of a structure to be detected;
in practical applications, the same technical solutions are adopted in this embodiment and the first embodiment.
S208: and inputting the obtained structural image of the building material to be detected into the constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected.
In practical applications, the same technical solutions are adopted in this embodiment and the first embodiment.
In practical application, as shown in fig. 2, the specific implementation of this embodiment is as follows:
in practical applications, the building material described in this embodiment is described by taking concrete as an example, and other building materials can be identified by the method steps described in this embodiment, only the training sample is the corresponding building material to be tested.
S201: training a neural network model by utilizing a training sample set in advance to obtain a concrete strength grade detection model;
s202: obtaining a model verification dataset;
s203: inputting each verification sample image into a concrete strength grade detection model to obtain the predicted strength grade of each verification sample image;
s204: calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images;
s205: judging whether the accuracy of the concrete strength grade detection model is not less than a preset threshold value, if not, executing a step S206; if yes, go to step S207;
s206: adding sample images in the training sample set, and returning to the step S201;
s207: acquiring an image of a structure to be detected;
s208: and inputting the structural image to be detected into the concrete strength grade detection model to obtain the strength grade of the structure to be detected.
The embodiment provides a building material strength grade detection method based on image recognition, which comprises the following steps: building a building material strength grade detection model; obtaining a model verification dataset; inputting each verification sample image into a building material strength grade detection model to obtain the predicted strength grade of each verification sample image; calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images; judging whether the accuracy of the building material strength grade detection model is not less than a preset threshold value or not; adding sample images in a training sample set; acquiring an image of a structure to be detected; and inputting the obtained structural image of the building material to be detected into the constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected. The strength grade detection model is trained by using the sample set, the accuracy of the strength grade detection model is ensured by using the model verification data set, and the strength grade of the structure to be detected can be accurately obtained by using the strength grade detection model.
EXAMPLE III
As shown in fig. 3, an apparatus for detecting strength grade of building material based on image recognition comprises the following modules:
the model pre-construction module 301 is used for training a neural network model to obtain a building material strength grade detection model;
an image obtaining module 302, configured to obtain an image of a structure to be detected;
and the strength grade calculation module 303 is configured to input the structural image to be detected to the building material strength grade detection model to obtain the strength grade of the structure to be detected.
In practical application, the model pre-construction module 301 may be a module that trains a neural network model to obtain a building material strength level detection model by using a training sample set based on a transfer learning method.
The embodiment provides a building material strength grade detection device based on image recognition, which comprises the following modules: a model pre-construction module; an image acquisition module; an intensity level calculation module; the model can be established quickly and effectively, and the strength grade of the structure to be measured can be obtained.
Example four
As shown in fig. 4, an apparatus for detecting strength grade of building material based on image recognition includes the following modules:
the model pre-construction module 301 is used for training a neural network model to obtain a building material strength grade detection model;
an image obtaining module 302, configured to obtain an image of a structure to be detected;
and the strength grade calculation module 303 is configured to input the structural image to be detected to the building material strength grade detection model to obtain the strength grade of the structure to be detected.
In practical application, the model pre-construction module 301 may be a module that trains a neural network model to obtain a building material strength level detection model by using a training sample set based on a transfer learning method.
In practical applications, the apparatus for detecting strength level of building material based on image recognition may further include a model verification module 304, and the model verification module 304 may include:
the model verification data set acquisition sub-module is used for acquiring a model verification data set, the model verification data set comprises a plurality of verification sample images, and each verification sample image is an image of a test component with a known actual strength grade and the same as the structural material to be tested;
the prediction strength grade calculation submodule is used for inputting each verification sample image into the building material strength grade detection model to obtain the prediction strength grade of each verification sample image;
the accuracy calculation submodule is used for calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images;
and the model retraining submodule is used for increasing the sample images in the training sample set when the accuracy of the building material strength grade detection model is smaller than a preset threshold value, and retraining the building material strength grade detection model until the accuracy is not smaller than the threshold value.
In practical application, the accuracy calculation sub-module can specifically count the number of unqualified verification sample images of which the difference value between the predicted intensity level and the actual intensity level is greater than a preset deviation value; and calculating the ratio of the number of disqualified products to the total number of the disqualified products to be used as a module for detecting the accuracy of the model of the strength grade of the building material.
The embodiment provides a building material strength grade detection device based on image recognition, which comprises the following modules: a model pre-construction module; an image acquisition module; an intensity level calculation module; a model verification module; the model can be quickly and effectively established, the strength grade of the structure to be tested can be obtained, and meanwhile, the accuracy of the model is ensured through model verification.
EXAMPLE five
The building material strength grade detection equipment based on image recognition specifically comprises:
the image acquisition equipment is used for acquiring an image of the structure to be detected and sending the image of the structure to be detected to the processor;
a memory for storing a computer program;
and a processor for executing a computer program to implement the steps of the building material strength grade detection method based on image recognition as described in the first and second embodiments.
In practical applications, various forms of image information collected by the image collecting device may be input into the processor through the scanner.
The embodiment provides a building material intensity grade detection equipment based on image recognition, specifically includes: an image acquisition device; a memory; a processor; the embodiment can be used for storing and executing computer programs to realize the steps of the method for detecting the strength grade of the building material by image recognition in the embodiment.
EXAMPLE six
A computer readable storage medium, on which a building material strength grade detection program based on image recognition is stored, wherein the building material strength grade detection program based on image recognition realizes the steps of the building material strength grade detection method based on image recognition according to the first embodiment and the second embodiment when being executed by a processor.
The implementation of the invention has the advantages that: the building material strength grade detection method based on image recognition comprises the following steps of: building a building material strength grade detection model; acquiring a structural image of a building material to be detected; inputting the obtained structural image of the building material to be detected into a constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected; the building material strength grade detection device based on image recognition comprises the following modules: a model pre-construction module; an image acquisition module; an intensity level calculation module; the invention can greatly improve the detection precision and the detection efficiency, is suitable for detecting the strength grade of various material structures, has stronger universality, and solves the problems that the rebound method has low detection precision, the core drilling method and the pulling-out method are easy to cause internal damage of building materials, and the detection precision of the ultrasonic detection method is greatly influenced by signal frequency, member size and the like. In addition, the invention also provides a corresponding implementation device, equipment and a computer readable storage medium for the method for detecting the concrete strength grade based on image recognition, so that the method has higher practicability, and the device, the equipment and the computer readable storage medium have corresponding advantages.
The above description is only an embodiment of the present invention, but the scope of the present invention 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 invention disclosed herein are intended to be covered by the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (10)
1. A building material strength grade detection method based on image recognition is characterized by comprising the following steps:
building a building material strength grade detection model;
acquiring a structural image of a building material to be detected;
inputting the obtained structural image of the building material to be detected into a constructed building material strength grade detection model for detection to obtain the strength grade of the building material to be detected;
the method for constructing the building material strength grade detection model comprises the following steps: training the neural network model by using a training sample set to obtain a building material strength grade detection model;
wherein the training sample set comprises a plurality of sample images of a test member of at least one material at different intensity levels, each sample image being pre-labeled with a respective intensity level.
2. The building material strength grade detection method based on image recognition as claimed in claim 1, further comprising the following steps before inputting the structural image to be detected into a pre-constructed building material strength grade detection model: and verifying the building material strength grade detection model.
3. The building material strength level detection method based on image recognition as claimed in claim 2, wherein the verifying the building material strength level detection model comprises the steps of:
obtaining a model verification data set, wherein the model verification data set comprises a plurality of verification sample images, and each verification sample image is an image of a test component with a known actual strength grade and the same as the structural material to be tested;
inputting each verification sample image into the building material strength grade detection model to obtain the predicted strength grade of each verification sample image;
calculating the accuracy of the building material strength grade detection model based on the predicted strength grade, the known actual strength grade and the total number of the verification sample images of each verification sample image;
judging whether the accuracy of the building material strength grade detection model is not less than a preset threshold value or not;
if the accuracy of the building material strength grade detection model is not smaller than a preset threshold value, using the building material strength grade detection model for subsequently calculating the strength grade of the structure to be detected;
and if the accuracy of the building material strength grade detection model is smaller than a preset threshold value, adding the sample images in the training sample set, and retraining the building material strength grade detection model until the accuracy is not smaller than the threshold value.
4. The building material strength grade detection method based on image recognition as claimed in claim 3, wherein the calculating of the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images comprises the following steps:
counting the unqualified number of the verification sample images of which the difference value between the predicted intensity level and the known actual intensity level is greater than a preset deviation value;
and calculating the ratio of the number of the disqualification to the total number to serve as the accuracy of the building material strength grade detection model.
5. The method for detecting the strength grade of the building material based on the image recognition according to one of claims 1 to 4, wherein the training of the neural network model by using the training sample set to obtain the strength grade detection model of the building material is specifically as follows: training a Faster-RCNN model by using the training sample set based on a transfer learning method to obtain a building material strength grade detection model; the training sample set comprises a plurality of sample images of test members made of the same materials as the building materials to be tested and in different strength grades, and each sample image is labeled with a corresponding strength grade in advance.
6. The method for building material strength grade detection based on image recognition according to claim 5, wherein the training of the fast-RCNN model by using the training sample set further comprises the following steps: and converting each sample image in the training sample set into a data set in a format convenient for deep learning.
7. The building material strength grade detection device based on image recognition is characterized by comprising the following modules:
the model pre-construction module is used for training the neural network model to obtain a building material strength grade detection model;
the image acquisition module is used for acquiring an image of the structure to be detected;
and the strength grade calculation module is used for inputting the structural image to be detected into the building material strength grade detection model to obtain the strength grade of the structure to be detected.
8. The image recognition-based construction material strength level detection apparatus according to claim 7, further comprising a model verification module, the model verification module including:
the model verification data set acquisition sub-module is used for acquiring a model verification data set, the model verification data set comprises a plurality of verification sample images, and each verification sample image is an image of a test component with a known actual strength grade and the same as the structural material to be tested;
the prediction strength grade calculation submodule is used for inputting each verification sample image into the building material strength grade detection model to obtain the prediction strength grade of each verification sample image;
the accuracy calculation submodule is used for calculating the accuracy of the building material strength grade detection model based on the predicted strength grade and the actual strength grade of each verification sample image and the total number of the verification sample images;
and the model retraining submodule is used for increasing the sample images in the training sample set when the accuracy of the building material strength grade detection model is smaller than a preset threshold value, and retraining the building material strength grade detection model until the accuracy is not smaller than the threshold value.
9. An image recognition-based building material strength level detection device, characterized by comprising an image acquisition device and a processor, wherein the processor is used for implementing the steps of the image recognition-based building material strength level detection method according to any one of claims 1 to 6 when executing a computer program stored in a memory.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a construction material strength level detection program based on image recognition, which when executed by a processor implements the steps of the construction material strength level detection method based on image recognition according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910984098.1A CN112669253A (en) | 2019-10-16 | 2019-10-16 | Building material strength grade detection method, device, equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910984098.1A CN112669253A (en) | 2019-10-16 | 2019-10-16 | Building material strength grade detection method, device, equipment and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112669253A true CN112669253A (en) | 2021-04-16 |
Family
ID=75400547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910984098.1A Pending CN112669253A (en) | 2019-10-16 | 2019-10-16 | Building material strength grade detection method, device, equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112669253A (en) |
-
2019
- 2019-10-16 CN CN201910984098.1A patent/CN112669253A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114359283B (en) | Defect detection method based on Transformer and electronic equipment | |
| CN112149356A (en) | Method, device, equipment and medium for predicting structural crack propagation path | |
| CN109858551A (en) | Method, apparatus, equipment and medium based on image recognition detection structure stress | |
| CN108711148B (en) | Tire defect intelligent detection method based on deep learning | |
| CN110006896B (en) | Virtual detector and detection method for concrete performance | |
| CN114913150A (en) | Intelligent identification method for concrete dam defect time sequence image | |
| CN112818969A (en) | Knowledge distillation-based face pose estimation method and system | |
| CN111062423B (en) | Point cloud classification method of point cloud graph neural network based on self-adaptive feature fusion | |
| CN109284700B (en) | Method, storage medium, device and system for detecting multiple faces in image | |
| CN114360030A (en) | Face recognition method based on convolutional neural network | |
| CN113255590A (en) | Defect detection model training method, defect detection method, device and system | |
| Murao et al. | Concrete crack detection using uav and deep learning | |
| CN112669253A (en) | Building material strength grade detection method, device, equipment and readable storage medium | |
| CN116823793A (en) | Device defect detection method, device, electronic device and readable storage medium | |
| CN116030292A (en) | Concrete surface roughness detection method based on improved ResNext | |
| CN115019252B (en) | Concrete quality detection method and device and monitoring equipment | |
| CN114677670B (en) | Method for automatically identifying and positioning identity card tampering | |
| CN115311526A (en) | Defect sample generation method and system based on improved Cycle GAN network | |
| CN115760720A (en) | Crack online detection method and system based on mobile device and super-resolution reconstruction segmentation network | |
| CN114186784A (en) | Electrical examination scoring method, system, medium and device based on edge calculation | |
| CN106655170A (en) | Power system adverse data identification method | |
| CN112434717B (en) | Model training method and device | |
| CN114299421A (en) | Method for detecting bolt looseness, convolutional neural network, training method and system | |
| CN111932642A (en) | Method, device and equipment for measuring and calculating volume of structural crack and storage medium | |
| CN111948286A (en) | Stress detection method, device and equipment based on ultrasonic waves and deep learning |
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 |