CN117952981B - Intelligent indoor lamp detection device and method based on CNN convolutional neural network - Google Patents
Intelligent indoor lamp detection device and method based on CNN convolutional neural network Download PDFInfo
- Publication number
- CN117952981B CN117952981B CN202410353973.7A CN202410353973A CN117952981B CN 117952981 B CN117952981 B CN 117952981B CN 202410353973 A CN202410353973 A CN 202410353973A CN 117952981 B CN117952981 B CN 117952981B
- Authority
- CN
- China
- Prior art keywords
- neural network
- cnn
- intelligent indoor
- indoor lamp
- image
- 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.)
- Active
Links
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 129
- 238000001514 detection method Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000003062 neural network model Methods 0.000 claims abstract description 43
- 238000012545 processing Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 230000007547 defect Effects 0.000 claims description 29
- 238000012549 training Methods 0.000 claims description 28
- 238000011176 pooling Methods 0.000 claims description 25
- 238000012360 testing method Methods 0.000 claims description 13
- 238000003860 storage Methods 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 9
- 230000007246 mechanism Effects 0.000 claims description 8
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000013500 data storage Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 claims description 4
- 238000007670 refining Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 230000035945 sensitivity Effects 0.000 claims description 4
- 239000003086 colorant Substances 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000001965 increasing effect Effects 0.000 description 2
- 241000899771 Arenga undulatifolia Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000005034 decoration Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- 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
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B20/00—Energy efficient lighting technologies, e.g. halogen lamps or gas discharge lamps
- Y02B20/40—Control techniques providing energy savings, e.g. smart controller or presence detection
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses an intelligent indoor lamp detection device and method based on CNN, the intelligent indoor lamp detection device based on CNN convolutional neural network comprises: the robot is used for performing multi-point position touch control on the intelligent indoor lamp to be detected; the machine vision equipment is used for acquiring an image of the surface of the intelligent indoor lamp to be detected; the image processing module is used for receiving the image data transmitted by the machine vision equipment and preprocessing the image; extracting features through a CNN neural network model to obtain feature information in an image; and acquiring the light intensity of the intelligent indoor lamp and the position information of the lighting area according to the characteristic information. The invention provides an intelligent indoor lamp detection device and method based on a CNN convolutional neural network, which solve the problems that the existing intelligent indoor lamp detection method is low in efficiency and is easy to cause non-uniform judgment standards due to artificial subjective factors, and achieve accurate and stable detection of the light intensity and position information of the intelligent indoor lamp.
Description
Technical Field
The invention relates to an intelligent indoor lamp detection device and method based on a CNN convolutional neural network, and belongs to the field of lamp detection.
Background
At present, an intelligent indoor lamp is an automotive interior part integrating decoration and functionality, and can wake up and activate through touch sensing, gestures or voice commands to illuminate different positions in the automobile to different degrees when a user needs. The intelligent indoor lamp adopts induction design, and has the advantages of high operation accuracy, real touch feeling, long service life and the like, and can prevent dust, water and false touch while ensuring attractive appearance.
The traditional intelligent indoor lamp detection method generally depends on manual operation and experience judgment, so that the efficiency is low, and the judgment standard is not unified due to artificial subjective factors. Along with the continuous development of science and technology, the development of the robot industry has an increasing influence on the testing industry, and the testing becomes simpler and more efficient due to the advantages of automation, high precision, consistency, acceleration of testing period, diversity testing and the like. The application of the CNN network technology in the fields of computer vision, natural language processing and the like is more and more widespread, and how to apply the CNN network technology to intelligent indoor lamp detection is a problem which needs to be solved currently and urgently.
Disclosure of Invention
The invention aims to solve the technical problems of overcoming the defects of the prior art, providing an intelligent indoor lamp detection device and method based on a CNN convolutional neural network, solving the problems that the existing intelligent indoor lamp detection method is low in efficiency and is easy to cause non-uniform judgment standard due to artificial subjective factors, and achieving accurate and stable detection of the light intensity and position information of the intelligent indoor lamp.
In order to solve the technical problems, the technical scheme of the invention is as follows:
the invention provides an intelligent indoor lamp detection device based on a CNN convolutional neural network, which comprises:
The robot is used for performing multi-point position touch control on the intelligent indoor lamp to be detected;
the machine vision equipment is used for acquiring an image of the surface of the intelligent indoor lamp to be detected;
The image processing module is used for receiving the image data transmitted by the machine vision equipment and preprocessing the image; extracting features through a CNN neural network model to obtain feature information in an image; acquiring the light intensity of the intelligent indoor lamp and the position information of the lighting area according to the characteristic information; judging whether the intelligent indoor lamp meets the standard requirements or not according to the acquired intelligent indoor lamp light intensity and the position information of the lighting area, and transmitting the detection result to a display storage module;
And the display storage module is used for storing and displaying the detection result.
Further, the machine vision device is an industrial camera.
Further, the display storage module comprises a data storage device and a display, wherein the data storage device is used for storing the detection report, and the display is used for displaying the detection result.
The invention further provides a detection method of the intelligent indoor lamp detection device based on the CNN convolutional neural network, which comprises the following steps:
S1, collecting an intelligent indoor lamp surface image by a machine vision device;
S2, preprocessing the acquired image data;
step S3, training is carried out based on the preprocessed image data, a CNN neural network model is constructed, the intelligent indoor lamp light intensity and the lighting area position information are identified and positioned through the CNN neural network model, and whether the standard requirement is met or not is judged;
And S4, sending the judging result to a display storage module.
Further, the step S2 specifically includes the following steps:
Step S21, gray scale is adopted The weighted average method of (2) carries out graying treatment on the image data, and the calculation formula of the graying treatment is as follows:
;
Wherein 0.299, 0.587 and 0.114 are based on the human eye to red Green/>Blue/>Weights obtained by calculating the sensitivity of the three colors;
step S22, a canny edge detector is adopted to extract the ROI area in the gray level image, then a gradient amplitude sharpening image is used, f (x, y) is made to represent an input image, G (x, y) is made to represent a Gaussian function, and a calculation formula of the Gaussian function G (x, y) is as follows:
;
wherein delta represents standard deviation, x represents horizontal coordinates of the image, and y represents vertical coordinates of the image;
Convolving G and f to form a smoothed image :
;
Calculating image gradient amplitudeAnd direction/>:
;
;
Wherein,Representing the gradient in the x-direction; /(I)Representing the gradient in the y-direction.
Further, in the step S3, training is performed based on the preprocessed image data, and a CNN neural network model is constructed, which specifically includes the following steps:
The CNN neural network model comprises a plurality of convolution layers, an average pooling layer and a full connection layer;
Defining a training data set, a test data set and a verification data set, dividing the preprocessed image data into various label sets, and completing the manufacture of the training data set, the test data set and the verification data set;
And training and optimizing the CNN convolutional neural network by adopting the CNN convolutional neural network to obtain a CNN neural network model after training.
Further, the training optimization is performed on the CNN convolutional neural network by adopting the CNN convolutional neural network to obtain a trained CNN neural network model, which specifically comprises the following steps:
setting a maximum pooling layer between a first convolution layer and a second convolution layer of a CNN convolution neural network, adopting maximum pooling operation, and selecting the maximum value in a local area as output;
the third to fifth convolution layers of the CNN convolution neural network adopt a downsampling residual error module and a lightweight residual error module;
outputting the data trained by the multi-layer convolution layer to an average pooling layer, and selecting an average value in a local area as output;
Performing verification and parameter adjustment, testing and deployment, repeating the processes of image acquisition and image preprocessing, and training a CNN convolutional neural network;
and obtaining the CNN neural network model after training.
Further, each layer in the CNN convolutional neural network adopts a channel attention mechanism, and a channel attention mechanism is adopted between the average pooling layer and the full-connection layer.
Further, in the step S3, the light intensity and the position information of the lighting area of the intelligent indoor lamp are identified and positioned through the CNN neural network model, and whether the light intensity and the position information of the lighting area meet the specification requirements is judged, which specifically comprises the following steps:
Acquiring the light intensity and the position information of a lighting area of the intelligent indoor lamp, analyzing image data through a CNN neural network model, inputting the preprocessed intelligent indoor lamp surface image to a first layer convolution layer of the CNN neural network model, and extracting initial defect edge characteristics;
Then stacking residual blocks, extracting further defect edge characteristics, and further refining and screening initial defect edge characteristics;
then global average pooling is carried out, and the integral characteristics of the defect characteristic diagram are reserved to obtain identification, positioning and classification information of the defects;
And processing and predicting through the CNN neural network model, and finally carrying out defect identification and classification on the intelligent indoor lamp surface image.
By adopting the technical scheme, the intelligent indoor lamp detection device is connected with the CNN neural network model by using the machine vision equipment to collect and detect the image of the surface of the intelligent indoor lamp, and has the advantages of improving the automation degree and the robustness of the intelligent indoor lamp detection. By using the CNN neural network technology, the interference caused by equipment and external factors can be self-adapted, the cost and error of manual detection are reduced, and a new solution is provided for intelligent indoor lamp detection.
Drawings
FIG. 1 is a schematic block diagram of an intelligent indoor lamp detection device based on a CNN convolutional neural network;
FIG. 2 is a flow chart of a detection method of the intelligent indoor lamp detection device based on the CNN convolutional neural network;
Fig. 3 is a flowchart of a training process of the CNN neural network model of the present invention.
Detailed Description
In order that the invention may be more readily understood, a more particular description of the invention will be rendered by reference to specific embodiments that are illustrated in the appended drawings.
Example 1
As shown in fig. 1, this embodiment provides an intelligent indoor lamp detection device based on CNN convolutional neural network, which includes:
the robot is used for carrying out multi-point position touch control on the intelligent indoor lamp to be detected, and JAKA cooperative robots are adopted in the robot of the embodiment.
The machine vision equipment is used for acquiring the image of the surface of the intelligent indoor lamp to be detected, and the machine vision equipment of the embodiment adopts an industrial camera.
The image processing module is used for receiving the image data transmitted by the machine vision equipment and preprocessing the image; extracting features through a CNN neural network model to obtain feature information in an image; acquiring the light intensity of the intelligent indoor lamp and the position information of the lighting area according to the characteristic information; and judging whether the intelligent indoor lamp meets the standard requirement or not according to the acquired intelligent indoor lamp light intensity and the position information of the lighting area, and transmitting the detection result to the display storage module.
And the display storage module is used for storing and displaying the detection result. The display storage module of the embodiment comprises a data storage device and a display, wherein the data storage device is used for storing a detection report, and the display is used for displaying a detection result.
Example two
As shown in fig. 2, the present embodiment provides a detection method of an intelligent indoor lamp detection device based on a CNN convolutional neural network, which includes:
S1, collecting an intelligent indoor lamp surface image by a machine vision device;
S2, preprocessing the acquired image data;
step S3, training is carried out based on the preprocessed image data, a CNN neural network model is constructed, the intelligent indoor lamp light intensity and the lighting area position information are identified and positioned through the CNN neural network model, and whether the standard requirement is met or not is judged;
And S4, sending the judging result to a display storage module.
Step S2 of the present embodiment specifically includes the following steps:
Step S21, gray scale is adopted The weighted average method of (2) is used for carrying out graying treatment on the image data, and the purpose of the graying treatment is to convert a color image into a gray image, so that an image matrix can be simplified, and the operation speed of a computer for processing the image can be improved. Because the gray level image only contains brightness information and does not contain color information, compared with a color image, the gray level image has lower information content, and is convenient for subsequent image processing, such as graph segmentation, feature extraction and the like. In addition, graying can also increase the contrast between images, highlighting the target area when processing. The calculation formula of the graying treatment is as follows:
;
Wherein 0.299, 0.587 and 0.114 are based on the human eye to red Green/>Blue/>And calculating weights of the sensitivity of the three colors, and adding the weights to obtain a gray value.
Step S22, a canny edge detector which is most widely applied in the digital image processing technology is adopted to extract the ROI (region of interest) in the gray level image, so that the image processing time is reduced and the accuracy is increased by selecting a specific ROI. The image is then sharpened using gradient magnitude, which can enhance the defect and eliminate features in the slowly varying background. Let f (x, y) denote the input image, G (x, y) denote a gaussian function, and the calculation formula of the gaussian function G (x, y) is:
;
wherein delta represents standard deviation, x represents horizontal coordinates of the image, and y represents vertical coordinates of the image;
Convolving G and f to form a smoothed image :
;
Calculating image gradient amplitudeAnd direction/>:
;
;
Wherein,Representing the gradient in the x-direction; /(I)Representing the gradient in the y-direction.
In step S3 of the present embodiment, training is performed based on the preprocessed image data, and a CNN neural network model is constructed, which specifically includes the following steps:
the CNN neural network model comprises a plurality of convolution layers, an average pooling layer and a full connection layer, wherein the convolution layers are used for extracting the characteristics of images, the pooling layer is used for reducing the dimension of data, and the full connection layer is used for tasks such as classification or regression;
multilayer convolution layer: the convolution layer is the core of the CNN, which extracts features of the image through a convolution operation. Each convolution layer typically contains a plurality of convolution kernels, each of which may extract a feature of the image. Feature extraction: once the CNN neural network model training is complete, it can be used to extract features of the image by inputting the image into the model and then taking the output of the convolutional layer. These outputs are characteristic representations of the image and can be used for subsequent classification, identification, etc.
Average pooling layer: the pooling layer is typically located after the convolution layer to reduce the dimensionality of the data and prevent overfitting. Common pooling operations include maximum pooling and average pooling.
Full tie layer: the full connection layer is typically located at the last few layers of the network for mapping the previously extracted features to the signature space of the sample.
And (3) subsequent treatment: the extracted features may be used for a variety of tasks such as image classification, object detection, image segmentation, etc. You can choose the appropriate algorithm and model for the subsequent processing according to the specific task.
As shown in fig. 3, the training process of the CNN neural network model is as follows:
1. and defining a training data set, a test data set and a verification data set, dividing the preprocessed image data into each label set, and completing the manufacture of the training data set, the test data set and the verification data set.
2. Training and optimizing the CNN convolutional neural network by adopting the CNN convolutional neural network to obtain a trained CNN neural network model, and specifically comprises the following steps:
and a maximum pooling layer is arranged between the first convolution layer and the second convolution layer of the CNN convolution neural network, and is used for reducing the output dimension of the convolution layers and the sensitivity to input data. Adopting maximum pooling operation, selecting the maximum value in the local area as output, and carrying out batch normalization to improve the network performance;
the third convolution layer to the fifth convolution layer of the CNN convolution neural network adopt a downsampling residual error module and a lightweight residual error module, and different parameters are selected according to the task of defect detection to optimize the network performance;
outputting the data trained by the multi-layer convolution layer to an average pooling layer, and selecting an average value in a local area as output;
Performing verification and parameter adjustment, testing and deployment, repeating the processes of image acquisition and image preprocessing, and training a CNN convolutional neural network;
and obtaining the CNN neural network model after training.
Each layer in the CNN convolutional neural network adopts a channel attention mechanism, and a channel attention mechanism is adopted between the average pooling layer and the full-connection layer. The channel attention mechanism is an attention mechanism for enhancing CNN performance, and aims to adaptively learn the importance of each channel so as to better focus on useful features in the feature extraction process, improve the performance of a network in a computer vision task, inhibit irrelevant channels and improve the feature characterization capability, thereby improving the generalization capability and robustness of a model.
In step S3 of the present embodiment, the identifying and positioning are performed on the light intensity and the position information of the lighting area of the intelligent indoor lamp through the CNN neural network model, and whether the light intensity and the position information meet the specification requirements is determined, which specifically includes the following steps:
acquiring the light intensity and the position information of a lighting area of the intelligent indoor lamp, analyzing image data through a CNN neural network model, inputting the preprocessed intelligent indoor lamp surface image to a first layer convolution layer of the CNN neural network model, and extracting initial defect edge characteristics, wherein the defect characteristics of the intelligent indoor lamp surface image have defects of local darkness, uneven brightness, light intensity and the like;
Then stacking residual blocks, extracting further defect edge characteristics, and further refining and screening initial defect edge characteristics;
then global average pooling is carried out, and the integral characteristics of the defect characteristic diagram are reserved to obtain identification, positioning and classification information of the defects;
And processing and predicting through the CNN neural network model, and finally carrying out defect identification and classification on the intelligent indoor lamp surface image.
The technical problems, technical solutions and advantageous effects solved by the present invention have been further described in detail in the above-described embodiments, and it should be understood that the above-described embodiments are only illustrative of the present invention and are not intended to limit the present invention, and any modifications, equivalent substitutions, improvements, etc. within the spirit and principle of the present invention should be included in the scope of protection of the present invention.
Claims (9)
1. An intelligent indoor lamp detection device based on CNN convolutional neural network, which is characterized in that the device comprises:
The robot is used for performing multi-point position touch control on the intelligent indoor lamp to be detected;
the machine vision equipment is used for acquiring an image of the surface of the intelligent indoor lamp to be detected;
The image processing module is used for receiving the image data transmitted by the machine vision equipment and preprocessing the image; extracting features through a CNN neural network model to obtain feature information in an image; acquiring the light intensity of the intelligent indoor lamp and the position information of the lighting area according to the characteristic information; judging whether the intelligent indoor lamp meets the standard requirements or not according to the acquired intelligent indoor lamp light intensity and the position information of the lighting area, and transmitting the detection result to a display storage module; the method for judging whether the intelligent indoor lamp meets the standard requirement or not according to the acquired intelligent indoor lamp light intensity and the lighting area position information specifically comprises the following steps: acquiring the light intensity and the position information of a lighting area of the intelligent indoor lamp, analyzing image data through a CNN neural network model, inputting the preprocessed intelligent indoor lamp surface image to a first layer convolution layer of the CNN neural network model, and extracting initial defect edge characteristics; then stacking residual blocks, extracting further defect edge characteristics, and further refining and screening initial defect edge characteristics; then global average pooling is carried out, and the integral characteristics of the defect characteristic diagram are reserved to obtain identification, positioning and classification information of the defects; performing processing and prediction through a CNN neural network model, and finally performing defect identification and classification on the intelligent indoor lamp surface image;
And the display storage module is used for storing and displaying the detection result.
2. The CNN convolutional neural network-based intelligent indoor light detection device of claim 1, wherein: the machine vision device is an industrial camera.
3. The CNN convolutional neural network-based intelligent indoor light detection device of claim 1, wherein: the display storage module comprises a data storage device and a display, wherein the data storage device is used for storing a detection report, and the display is used for displaying a detection result.
4. A method for detecting an intelligent indoor lamp detection device based on CNN convolutional neural network according to any one of claims 1 to 3, comprising:
S1, collecting an intelligent indoor lamp surface image by a machine vision device;
S2, preprocessing the acquired image data;
step S3, training is carried out based on the preprocessed image data, a CNN neural network model is constructed, the intelligent indoor lamp light intensity and the lighting area position information are identified and positioned through the CNN neural network model, and whether the standard requirement is met or not is judged;
And S4, sending the judging result to a display storage module.
5. The method for detecting the intelligent indoor lamp detection device based on the CNN convolutional neural network according to claim 4, wherein the step S2 specifically comprises the following steps:
Step S21, gray scale is adopted The weighted average method of (2) carries out graying treatment on the image data, and the calculation formula of the graying treatment is as follows:
;
Wherein 0.299, 0.587 and 0.114 are based on the human eye to red Green/>Blue/>Weights obtained by calculating the sensitivity of the three colors;
step S22, a canny edge detector is adopted to extract the ROI area in the gray level image, then a gradient amplitude sharpening image is used, f (x, y) is made to represent an input image, G (x, y) is made to represent a Gaussian function, and a calculation formula of the Gaussian function G (x, y) is as follows:
;
Wherein, Representing standard deviation, x representing horizontal coordinates of the image, and y representing vertical coordinates of the image;
Convolving G and f to form a smoothed image :
;
Calculating image gradient amplitudeAnd direction/>:
;
Wherein,Representing the gradient in the x-direction; /(I)Representing the gradient in the y-direction.
6. The method for detecting the intelligent indoor lamp detection device based on the CNN convolutional neural network according to claim 4, wherein in the step S3, training is performed based on the preprocessed image data, and a CNN neural network model is constructed, specifically comprising the following steps:
The CNN neural network model comprises a plurality of convolution layers, an average pooling layer and a full connection layer;
Defining a training data set, a test data set and a verification data set, dividing the preprocessed image data into various label sets, and completing the manufacture of the training data set, the test data set and the verification data set;
And training and optimizing the CNN convolutional neural network by adopting the CNN convolutional neural network to obtain a CNN neural network model after training.
7. The method for detecting the intelligent indoor lamp detection device based on the CNN convolutional neural network according to claim 6, wherein the training optimization is performed on the CNN convolutional neural network by adopting the CNN convolutional neural network to obtain a trained CNN neural network model, and the method specifically comprises the following steps:
setting a maximum pooling layer between a first convolution layer and a second convolution layer of a CNN convolution neural network, adopting maximum pooling operation, and selecting the maximum value in a local area as output;
the third to fifth convolution layers of the CNN convolution neural network adopt a downsampling residual error module and a lightweight residual error module;
outputting the data trained by the multi-layer convolution layer to an average pooling layer, and selecting an average value in a local area as output;
Performing verification and parameter adjustment, testing and deployment, repeating the processes of image acquisition and image preprocessing, and training a CNN convolutional neural network;
and obtaining the CNN neural network model after training.
8. The detection method of the intelligent indoor lamp detection device based on the CNN convolutional neural network, as set forth in claim 7, is characterized in that: each layer in the CNN convolutional neural network adopts a channel attention mechanism, and a channel attention mechanism is adopted between the average pooling layer and the full-connection layer.
9. The method for detecting the intelligent indoor lamp detection device based on the CNN convolutional neural network according to claim 4, wherein in the step S3, the intelligent indoor lamp intensity and the position information of the lighting area are identified and positioned through the CNN neural network model, and whether the standard requirement is met is judged, which specifically comprises the following steps:
Acquiring the light intensity and the position information of a lighting area of the intelligent indoor lamp, analyzing image data through a CNN neural network model, inputting the preprocessed intelligent indoor lamp surface image to a first layer convolution layer of the CNN neural network model, and extracting initial defect edge characteristics;
Then stacking residual blocks, extracting further defect edge characteristics, and further refining and screening initial defect edge characteristics;
then global average pooling is carried out, and the integral characteristics of the defect characteristic diagram are reserved to obtain identification, positioning and classification information of the defects;
And processing and predicting through the CNN neural network model, and finally carrying out defect identification and classification on the intelligent indoor lamp surface image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410353973.7A CN117952981B (en) | 2024-03-27 | 2024-03-27 | Intelligent indoor lamp detection device and method based on CNN convolutional neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410353973.7A CN117952981B (en) | 2024-03-27 | 2024-03-27 | Intelligent indoor lamp detection device and method based on CNN convolutional neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN117952981A CN117952981A (en) | 2024-04-30 |
CN117952981B true CN117952981B (en) | 2024-06-21 |
Family
ID=90798223
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410353973.7A Active CN117952981B (en) | 2024-03-27 | 2024-03-27 | Intelligent indoor lamp detection device and method based on CNN convolutional neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117952981B (en) |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110135343A (en) * | 2019-05-15 | 2019-08-16 | 南通大学 | A kind of street lamp intelligent measurement and night status judgment method |
CN114022820A (en) * | 2021-11-09 | 2022-02-08 | 交通运输部北海航海保障中心烟台航标处 | Intelligent beacon light quality detection method based on machine vision |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10936916B1 (en) * | 2019-10-31 | 2021-03-02 | Booz Allen Hamilton Inc. | System and method for classifying image data |
CN114364099B (en) * | 2022-01-13 | 2023-07-18 | 达闼机器人股份有限公司 | Method for adjusting intelligent light equipment, robot and electronic equipment |
CN116433568A (en) * | 2022-04-15 | 2023-07-14 | 电子科技大学中山学院 | LED lamp strip defect detection method and system based on depth support vector description |
CN115457444A (en) * | 2022-09-21 | 2022-12-09 | 上海施成照明工程有限公司 | Method and system for automatically identifying lamp fault, electronic equipment and storage medium |
-
2024
- 2024-03-27 CN CN202410353973.7A patent/CN117952981B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110135343A (en) * | 2019-05-15 | 2019-08-16 | 南通大学 | A kind of street lamp intelligent measurement and night status judgment method |
CN114022820A (en) * | 2021-11-09 | 2022-02-08 | 交通运输部北海航海保障中心烟台航标处 | Intelligent beacon light quality detection method based on machine vision |
Also Published As
Publication number | Publication date |
---|---|
CN117952981A (en) | 2024-04-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106875381B (en) | Mobile phone shell defect detection method based on deep learning | |
CN106875373B (en) | Mobile phone screen MURA defect detection method based on convolutional neural network pruning algorithm | |
CN110389127B (en) | System and method for identifying metal ceramic parts and detecting surface defects | |
CN108960245B (en) | Tire mold character detection and recognition method, device, equipment and storage medium | |
CN111833306B (en) | Defect detection method and model training method for defect detection | |
CN113592845A (en) | Defect detection method and device for battery coating and storage medium | |
CN106599854B (en) | Automatic facial expression recognition method based on multi-feature fusion | |
CN109509187B (en) | Efficient inspection algorithm for small defects in large-resolution cloth images | |
CN109636772A (en) | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning | |
CN111815564B (en) | Method and device for detecting silk ingots and silk ingot sorting system | |
CN111932501A (en) | Seal ring surface defect detection method based on semantic segmentation | |
CN111612737B (en) | Artificial board surface flaw detection device and detection method | |
CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
CN111667455A (en) | AI detection method for various defects of brush | |
CN112907519A (en) | Metal curved surface defect analysis system and method based on deep learning | |
CN112949704B (en) | Tobacco leaf maturity state identification method and device based on image analysis | |
CN110866915A (en) | Circular inkstone quality detection method based on metric learning | |
CN115775236A (en) | Surface tiny defect visual detection method and system based on multi-scale feature fusion | |
CN115294033A (en) | Tire belt layer difference level and misalignment defect detection method based on semantic segmentation network | |
CN116678826A (en) | Appearance defect detection system and method based on rapid three-dimensional reconstruction | |
CN116559111A (en) | Sorghum variety identification method based on hyperspectral imaging technology | |
CN117011260A (en) | Automatic chip appearance defect detection method, electronic equipment and storage medium | |
CN115829942A (en) | Electronic circuit defect detection method based on non-negative constraint sparse self-encoder | |
CN117557565B (en) | Detection method and device for lithium battery pole piece | |
CN112561885B (en) | YOLOv 4-tiny-based gate valve opening detection method |
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 | ||
GR01 | Patent grant |