CN110728223A - Helmet wearing identification method based on deep learning - Google Patents
Helmet wearing identification method based on deep learning Download PDFInfo
- Publication number
- CN110728223A CN110728223A CN201910949981.7A CN201910949981A CN110728223A CN 110728223 A CN110728223 A CN 110728223A CN 201910949981 A CN201910949981 A CN 201910949981A CN 110728223 A CN110728223 A CN 110728223A
- Authority
- CN
- China
- Prior art keywords
- training
- deep learning
- neural network
- model
- network
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
Abstract
The invention discloses a safety helmet wearing identification method based on deep learning, which comprises the following steps: a) acquiring a video image; b) manually marking the images, wherein the marking content indicates whether a safety helmet is worn, so that a database consisting of training samples and test samples is obtained; c) forming a neural network target detection model for safety helmet detection based on the target detection SSD network after deep learning; d) detecting the accuracy of the neural network, wherein the model accuracy also slowly increases with the increase of the training times, and finally converges to 99.8 percent, which exceeds the average human recognition rate; e) real-time detection of headgear wear. According to the safety helmet wearing identification method based on deep learning, the image characteristics are processed through the convolutional neural network to obtain the neural network model, the model is trained continuously, the network model is obtained finally with high accuracy, and reliable basis is provided for safety helmet wearing identification; the loss value is reduced and the network is optimized.
Description
Technical Field
The invention relates to a safety helmet wearing identification method, in particular to a safety helmet wearing identification method based on deep learning.
Background
The safety helmet has the advantages that the safety helmet has a protection effect on the head of a human body from being damaged by external force, for places such as building construction sites and ship construction with high falling object risks, safety production specifications require field operation personnel to wear the safety helmet, but some workers and even managers still have the functions of disregarding the protection function of the safety helmet and the method for correctly using the safety helmet, the safety helmet enters a construction area with the high falling object risks and is not required to be matched and put on, and once an accident occurs, the personnel who are required to wear the safety helmet are greatly damaged. At present, the wearing of the safety helmet can only depend on manual inspection, the supervision difficulty is high, so that the execution of a plurality of construction sites is not in place, and important potential safety hazards exist. Therefore, it is necessary to develop a method for automatically recognizing the wearing of the safety helmet.
Disclosure of Invention
In order to overcome the defects of the technical problems, the invention provides a safety helmet wearing identification method based on deep learning.
The invention discloses a safety helmet wearing identification method based on deep learning, which is characterized by comprising the following steps of:
a) video image acquisition, namely performing video acquisition on the field operation condition to be monitored by using a camera and framing the acquired video;
b) establishing a database, detecting images with human bodies appearing in the frame images obtained in the step a) by adopting an SSD algorithm, manually marking the images, and marking whether a safety helmet is worn or not, thereby obtaining the database consisting of training samples and test samples;
c) building a training model, taking the SSD network as a training neural network based on the target detection after deep learning, and adjusting neural network parameters in multiple training so that the loss function of the network tends to converge in the iterative process, and finally forming a neural network target detection model for helmet detection;
d) detecting the accuracy of the neural network, wherein in the process of training the neural network model in the step c), 50000 times of training are carried out in total, the accuracy of the model is tested every 20 times of training, and along with the increase of the training times, the accuracy of the model also slowly increases and finally converges to 99.8 percent, which exceeds the average human recognition rate; test sample data does not participate in training in the training process, but is used for evaluating the network accuracy;
e) and detecting whether the safety helmet is worn in real time, acquiring a figure picture in a shooting range, identifying the person who does not wear the safety helmet by using the trained neural network target detection model, and performing real-time voice alarm.
The helmet wearing identification method based on deep learning comprises the following steps of b), in the process of collecting image data of a person, due to the fact that data quantity is insufficient and data general contrast is low, the performance of an SSD (solid State disk) can be improved by adopting data amplification, the method adopting data amplification comprises image horizontal turning, random collection of a graphic block domain and image enhancement based on log transformation, wherein the image enhancement method based on log transformation is shown as a formula (1):
I(x,y)＝C·log(V(x,y)+1) (1)
wherein, V is the original pixel value, I is the pixel after correction, and C is the adjustment coefficient.
In the method for identifying wearing of the safety helmet based on deep learning, in the step c), the target detection network SSD adopts the VGG16 as a basic model, and then a convolution layer is added on the basis of the VGG16 to obtain more characteristic maps for detection.
The invention discloses a safety helmet wearing identification method based on deep learning, wherein parameters defined by a neural network model in the step c) comprise: the method comprises the steps of (1) training by utilizing images in a test sample and outputting a training log, wherein the total number of categories, the size of a rectangular frame, the learning rate and the weight attenuation rate are obtained; and calculating the change of the accuracy by using the test sample so as to adjust the network parameters according to the change of the accuracy and obtain the network model meeting the requirement.
In the step c), in the process of training a neural network target detection model, different characteristic pictures are extracted by using a convolutional neural network, each characteristic picture generates a rectangular frame, and the image is subjected to convolution operation through a formula (2):
wherein f represents the convolved image, h represents a convolution kernel, and g represents the result of the image after convolution; i. j denotes the width and height of the image.
In the method for identifying the wearing of the safety helmet based on the deep learning, in the step c) in the process of training a neural network model, the SSD network has the following loss functions that the loss function cannot be converged in the later training period and the loss value is high because the number of positive and negative samples is unbalanced and different samples are not distinguished, so that the SSD network loss function is modified, the network is optimized, and the loss function based on the deep learning is obtained through a formula (3):
wherein, alpha is a weighting coefficient, and N is the number of samples; l isconfThe calculation is performed by equation (4):
wherein r is a parameter for reducing the loss of the easily classified sample, and the value of r is 2; alpha is alpha1The parameters are hyper-parameters, are used for balancing the uneven proportion of the positive and negative samples, and are adjusted according to the training result; x represents each training sample, y represents the true label of the sample, and y' represents the predicted label of the sample;
wherein L islocThe calculation is performed by equation (5):
wherein:
lmrepresentation prediction boxCoordinates of the center point and width and height, gmRepresenting the coordinates of the center point of the prediction box and the width and height.
In the step d), in the process of training the neural network model, the accuracy is calculated by the formula (7) to obtain:
where tp is an abbreviation for true posives meaning the number of instances that are actually positive and divided into positive by the classifier, and fp is an abbreviation for false posives meaning the number of instances that are actually negative but divided into positive by the classifier.
In the method for identifying the wearing of the safety helmet based on the deep learning, in the step d), when the result of each picture is calculated, a non-maximum value inhibition algorithm is utilized, but the traditional non-maximum value inhibition is calculated based on the cross-over ratio of two images, the cross-over ratio formula is subjected to the deep learning based on the deep learning, so that the accuracy is improved, and the cross-over ratio formula based on the deep learning is as follows:
wherein SA，SBRespectively representing the area sizes of the A and B targets.
The invention has the beneficial effects that: according to the safety helmet wearing identification method based on deep learning, the image characteristics are processed through the convolutional neural network to obtain the neural network model, the model is trained continuously, the network model is obtained finally with high accuracy, and reliable basis is provided for safety helmet wearing identification; according to the invention, the SSD network loss function is modified, so that the loss value is reduced, and the network is optimized; meanwhile, the invention utilizes a non-maximum suppression algorithm and improves the identification accuracy.
Drawings
FIG. 1 is a schematic illustration of the wearing and non-wearing of the helmet identified in the present invention;
fig. 2 is a flowchart of the method for identifying wearing of a safety helmet based on deep learning according to the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
As shown in fig. 1, a schematic diagram of the wearing and non-wearing of the safety helmet identified by the present invention is given, and fig. 2 is a flowchart of a deep learning-based safety helmet wearing identification method of the present invention, which is implemented by the following steps:
a) video image acquisition, namely performing video acquisition on the field operation condition to be monitored by using a camera and framing the acquired video;
b) establishing a database, detecting images with human bodies appearing in the frame images obtained in the step a) by adopting an SSD algorithm, manually marking the images, and marking whether a safety helmet is worn or not, thereby obtaining the database consisting of training samples and test samples;
in the step, in the process of collecting the image data of the person, due to insufficient data volume and low general contrast of the data, the performance of the SSD can be improved by adopting data amplification, and the image enhancement method based on log transformation comprises the steps of image horizontal inversion, random acquisition of a graphic block domain and log-transformation-based image enhancement, wherein the log-transformation-based image enhancement method is shown as a formula (1):
I(x,y)＝C·log(V(x,y)+1) (1)
wherein, V is the original pixel value, I is the pixel after correction, and C is the adjustment coefficient.
c) Building a training model, taking the SSD network as a training neural network based on the target detection after deep learning, and adjusting neural network parameters in multiple training so that the loss function of the network tends to converge in the iterative process, and finally forming a neural network target detection model for helmet detection;
in this step, the target detection network SSD adopts VGG16 as a basic model, and then adds convolution layer on the basis of VGG16 to obtain more feature maps for detection.
In this step, the parameters defined by the neural network model include: the method comprises the steps of (1) training by utilizing images in a test sample and outputting a training log, wherein the total number of categories, the size of a rectangular frame, the learning rate and the weight attenuation rate are obtained; and calculating the change of the accuracy by using the test sample so as to adjust the network parameters according to the change of the accuracy and obtain the network model meeting the requirement.
In the step, in the process of training a neural network target detection model, different feature pictures are extracted by using a convolutional neural network, each feature picture generates a rectangular frame, and the image is subjected to convolution operation through a formula (2):
wherein f represents the convolved image, h represents a convolution kernel, and g represents the result of the image after convolution; i. j denotes the width and height of the image. For the prior rectangular frame selection, 9 rectangles with different sizes are selected as the prior rectangular frames according to the labeled data set clustering by adopting a clustering algorithm proposed by YOLOV 3.
In the step, in the process of training the neural network model, the SSD network modifies the SSD network loss function to optimize the network because the number of positive and negative samples is unbalanced and different samples are not distinguished, so that the loss function cannot be converged at the later stage of training, and the loss value is high, and the loss function after deep learning is obtained by formula (3):
wherein, alpha is a weighting coefficient, and N is the number of samples; l isconfThe calculation is performed by equation (4):
wherein r is a parameter for reducing the loss of the easily classified sample, and the value of r is 2; alpha is alpha1Is a hyper-parameter, is used to balance positive and negative samplesThe province is not uniform in proportion and is adjusted according to the training result; x represents each training sample, y represents the true label of the sample, and y' represents the predicted label of the sample;
wherein L islocThe calculation is performed by equation (5):
wherein:
lmcoordinates of center point and width and height, g, representing the prediction boxmRepresenting the coordinates of the center point of the prediction box and the width and height.
d) Detecting the accuracy of the neural network, wherein in the process of training the neural network model in the step c), 50000 times of training are carried out in total, the accuracy of the model is tested every 20 times of training, and along with the increase of the training times, the accuracy of the model also slowly increases and finally converges to 99.8 percent, which exceeds the average human recognition rate; test sample data does not participate in training in the training process, but is used for evaluating the network accuracy;
in the step, in the process of training the neural network model, the accuracy is calculated by the formula (7):
where tp is an abbreviation for true posives meaning the number of instances that are actually positive and divided into positive by the classifier, and fp is an abbreviation for false posives meaning the number of instances that are actually negative but divided into positive by the classifier.
In the step, when calculating the result of each picture, a non-maximum suppression algorithm is used, but the traditional non-maximum suppression is calculated based on the cross-over ratio of two images, and the cross-over ratio formula is subjected to deep learning based on the deep learning, so that the accuracy is improved, and the cross-over ratio formula based on the deep learning is as follows:
wherein SA，SBRespectively representing the area sizes of the A and B targets.
e) And detecting whether the safety helmet is worn in real time, acquiring a figure picture in a shooting range, identifying the person who does not wear the safety helmet by using the trained neural network target detection model, and performing real-time voice alarm.
Claims (8)
1. A safety helmet wearing identification method based on deep learning is characterized by being realized through the following steps:
a) video image acquisition, namely performing video acquisition on the field operation condition to be monitored by using a camera and framing the acquired video;
b) establishing a database, detecting images with human bodies appearing in the frame images obtained in the step a) by adopting an SSD algorithm, manually marking the images, and marking whether a safety helmet is worn or not, thereby obtaining the database consisting of training samples and test samples;
c) building a training model, taking the SSD network as a training neural network based on the target detection after deep learning, and adjusting neural network parameters in multiple training so that the loss function of the network tends to converge in the iterative process, and finally forming a neural network target detection model for helmet detection;
d) detecting the accuracy of the neural network, wherein in the process of training the neural network model in the step c), 50000 times of training are carried out in total, the accuracy of the model is tested every 20 times of training, and along with the increase of the training times, the accuracy of the model also slowly increases and finally converges to 99.8 percent, which exceeds the average human recognition rate; test sample data does not participate in training in the training process, but is used for evaluating the network accuracy;
e) and detecting whether the safety helmet is worn in real time, acquiring a figure picture in a shooting range, identifying the person who does not wear the safety helmet by using the trained neural network target detection model, and performing real-time voice alarm.
2. The deep learning-based helmet wearing identification method according to claim 1, characterized in that: step b), in the process of collecting image data of a person, due to insufficient data volume and low general contrast of the data, the performance of the SSD can be improved by adopting data amplification, and the image enhancement method based on log transformation comprises the steps of image horizontal inversion, random acquisition of a graphic block domain and log-transformation-based image enhancement, wherein the log-transformation-based image enhancement method is shown as a formula (1):
I(x,y)＝C·log(V(x,y)+1) (1)
wherein, V is the original pixel value, I is the pixel after correction, and C is the adjustment coefficient.
3. The deep learning-based helmet wearing identification method according to claim 1 or 2, characterized in that: in the step c), the target detection network SSD adopts the VGG16 as a basic model, and then a convolution layer is added on the basis of the VGG16 to obtain more feature maps for detection.
4. The deep learning-based helmet wearing identification method according to claim 1 or 2, characterized in that: the parameters defined by the neural network model in the step c) comprise: the method comprises the steps of (1) training by utilizing images in a test sample and outputting a training log, wherein the total number of categories, the size of a rectangular frame, the learning rate and the weight attenuation rate are obtained; and calculating the change of the accuracy by using the test sample so as to adjust the network parameters according to the change of the accuracy and obtain the network model meeting the requirement.
5. The deep learning-based helmet wearing identification method according to claim 1 or 2, characterized in that: in the step c), in the process of training a neural network target detection model, different feature pictures are extracted by using a convolutional neural network, each feature picture generates a rectangular frame, and the image is subjected to convolution operation through a formula (2):
wherein f represents the convolved image, h represents a convolution kernel, and g represents the result of the image after convolution; i. j denotes the width and height of the image.
6. The deep learning-based helmet wearing identification method according to claim 1 or 2, characterized in that: step c) in the process of training the neural network model, the SSD network modifies the SSD network loss function so as to optimize the network because the quantities of positive and negative samples are unbalanced and different samples are not distinguished, and the loss function cannot be converged at the later training stage, and the loss value is high, and the loss function after deep learning is solved through a formula (3):
wherein, alpha is a weighting coefficient, and N is the number of samples; l isconfThe calculation is performed by equation (4):
wherein r is a parameter for reducing the loss of the easily classified sample, and the value of r is 2; alpha is alpha1The parameters are hyper-parameters, are used for balancing the uneven proportion of the positive and negative samples, and are adjusted according to the training result; x represents each training sample, y represents the true label of the sample, and y' represents the predicted label of the sample;
wherein L islocThe calculation is performed by equation (5):
wherein:
lmcoordinates of center point and width and height, g, representing the prediction boxmRepresenting the coordinates of the center point of the prediction box and the width and height.
7. The method for identifying wearing of safety helmet based on deep learning according to claim 1 or 2, wherein in the step d), the accuracy is calculated by formula (7) in the process of training the neural network model:
where tp is an abbreviation for true posives meaning the number of instances that are actually positive and divided into positive by the classifier, and fp is an abbreviation for false posives meaning the number of instances that are actually negative but divided into positive by the classifier.
8. The method for identifying wearing of safety helmet based on deep learning of claim 1 or 2, wherein in the step d), a non-maximum suppression algorithm is used in calculating the result of each picture, but the conventional non-maximum suppression is calculated based on the cross-over ratio of two images, wherein the cross-over ratio formula is based on deep learning, so as to improve the accuracy, and the cross-over ratio formula after deep learning is as follows:
wherein SA，SBRespectively representing the area sizes of the A and B targets.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910949981.7A CN110728223A (en) | 2019-10-08 | 2019-10-08 | Helmet wearing identification method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910949981.7A CN110728223A (en) | 2019-10-08 | 2019-10-08 | Helmet wearing identification method based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110728223A true CN110728223A (en) | 2020-01-24 |
Family
ID=69220752
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910949981.7A Pending CN110728223A (en) | 2019-10-08 | 2019-10-08 | Helmet wearing identification method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110728223A (en) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111401310A (en) * | 2020-04-08 | 2020-07-10 | 天津中科智能识别产业技术研究院有限公司 | Kitchen health safety supervision and management method based on artificial intelligence |
| CN111476083A (en) * | 2020-02-07 | 2020-07-31 | 山东理工大学 | Automatic identification method for wearing of safety helmet of electric power staff |
| CN111488804A (en) * | 2020-03-19 | 2020-08-04 | 山西大学 | Labor insurance product wearing condition detection and identity identification method based on deep learning |
| CN111582068A (en) * | 2020-04-22 | 2020-08-25 | 北京交通大学 | Method for detecting wearing state of personal mask |
| CN111598040A (en) * | 2020-05-25 | 2020-08-28 | 中建三局第二建设工程有限责任公司 | Construction worker identity identification and safety helmet wearing detection method and system |
| CN111639546A (en) * | 2020-05-07 | 2020-09-08 | 金钱猫科技股份有限公司 | Small-scale target cloud computing identification method and device based on neural network |
| CN111652046A (en) * | 2020-04-17 | 2020-09-11 | 济南浪潮高新科技投资发展有限公司 | Safe wearing detection method, equipment and system based on deep learning |
| CN111753805A (en) * | 2020-07-08 | 2020-10-09 | 深延科技(北京)有限公司 | Method and device for detecting wearing of safety helmet |
| CN111783564A (en) * | 2020-06-15 | 2020-10-16 | 青岛众瑞智能仪器有限公司 | Method for rapidly detecting wearing safety of respiratory tract protection equipment |
| CN111914743A (en) * | 2020-07-31 | 2020-11-10 | 广东电网有限责任公司清远供电局 | Method and device for detecting safety helmet of transformer substation worker |
| CN111950457A (en) * | 2020-08-12 | 2020-11-17 | 海默潘多拉数据科技(深圳)有限公司 | Oil field safety production image identification method and system |
| CN112070043A (en) * | 2020-09-15 | 2020-12-11 | 常熟理工学院 | Safety helmet wearing convolutional network based on feature fusion, training and detecting method |
| CN112183532A (en) * | 2020-09-28 | 2021-01-05 | 福建师范大学福清分校 | Safety helmet identification method based on weak supervision collaborative learning algorithm and storage medium |
| CN112215809A (en) * | 2020-09-25 | 2021-01-12 | 北京林业大学 | Prawn processing shelling rate detection method and device, electronic equipment and storage medium |
| CN112464822A (en) * | 2020-11-30 | 2021-03-09 | 深圳市捷顺科技实业股份有限公司 | Helmet wearing detection method and device based on feature enhancement |
| CN112598055A (en) * | 2020-12-21 | 2021-04-02 | 电子科技大学 | Helmet wearing detection method, computer-readable storage medium and electronic device |
| CN113435369A (en) * | 2021-07-01 | 2021-09-24 | 华风数据(深圳)有限公司 | Photovoltaic fault intelligent identification method based on SSD algorithm model and far infrared video |
| CN113808197A (en) * | 2021-09-17 | 2021-12-17 | 山西大学 | Automatic workpiece grabbing system and method based on machine learning |
| CN114299429A (en) * | 2021-12-24 | 2022-04-08 | 宁夏广天夏电子科技有限公司 | Human body recognition method, system and device based on deep learning |
| CN114595759A (en) * | 2022-03-07 | 2022-06-07 | 卡奥斯工业智能研究院(青岛)有限公司 | Protective tool identification method and device, electronic equipment and storage medium |
| CN114627425A (en) * | 2021-06-11 | 2022-06-14 | 珠海路讯科技有限公司 | Method for detecting whether worker wears safety helmet or not based on deep learning |
| CN114782986A (en) * | 2022-03-28 | 2022-07-22 | 佳源科技股份有限公司 | Helmet wearing detection method, device, equipment and medium based on deep learning |
| CN114782986B (en) * | 2022-03-28 | 2024-04-26 | 佳源科技股份有限公司 | Deep learning-based safety helmet wearing detection method, device, equipment and medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256464A (en) * | 2018-01-12 | 2018-07-06 | 适普远景遥感信息技术（北京）有限公司 | High-resolution remote sensing image urban road extracting method based on deep learning |
| CN108304787A (en) * | 2018-01-17 | 2018-07-20 | 河南工业大学 | Road target detection method based on convolutional neural networks |
| US20180330166A1 (en) * | 2017-05-09 | 2018-11-15 | Blue River Technology Inc. | Automated plant detection using image data |
| CN110287804A (en) * | 2019-05-30 | 2019-09-27 | 广东电网有限责任公司 | A kind of electric operating personnel's dressing recognition methods based on mobile video monitor |
-
2019
- 2019-10-08 CN CN201910949981.7A patent/CN110728223A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180330166A1 (en) * | 2017-05-09 | 2018-11-15 | Blue River Technology Inc. | Automated plant detection using image data |
| CN108256464A (en) * | 2018-01-12 | 2018-07-06 | 适普远景遥感信息技术（北京）有限公司 | High-resolution remote sensing image urban road extracting method based on deep learning |
| CN108304787A (en) * | 2018-01-17 | 2018-07-20 | 河南工业大学 | Road target detection method based on convolutional neural networks |
| CN110287804A (en) * | 2019-05-30 | 2019-09-27 | 广东电网有限责任公司 | A kind of electric operating personnel's dressing recognition methods based on mobile video monitor |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111476083A (en) * | 2020-02-07 | 2020-07-31 | 山东理工大学 | Automatic identification method for wearing of safety helmet of electric power staff |
| CN111488804A (en) * | 2020-03-19 | 2020-08-04 | 山西大学 | Labor insurance product wearing condition detection and identity identification method based on deep learning |
| CN111401310B (en) * | 2020-04-08 | 2023-08-29 | 天津中科智能识别产业技术研究院有限公司 | Kitchen sanitation safety supervision and management method based on artificial intelligence |
| CN111401310A (en) * | 2020-04-08 | 2020-07-10 | 天津中科智能识别产业技术研究院有限公司 | Kitchen health safety supervision and management method based on artificial intelligence |
| CN111652046A (en) * | 2020-04-17 | 2020-09-11 | 济南浪潮高新科技投资发展有限公司 | Safe wearing detection method, equipment and system based on deep learning |
| CN111582068A (en) * | 2020-04-22 | 2020-08-25 | 北京交通大学 | Method for detecting wearing state of personal mask |
| CN111639546A (en) * | 2020-05-07 | 2020-09-08 | 金钱猫科技股份有限公司 | Small-scale target cloud computing identification method and device based on neural network |
| CN111598040A (en) * | 2020-05-25 | 2020-08-28 | 中建三局第二建设工程有限责任公司 | Construction worker identity identification and safety helmet wearing detection method and system |
| CN111783564A (en) * | 2020-06-15 | 2020-10-16 | 青岛众瑞智能仪器有限公司 | Method for rapidly detecting wearing safety of respiratory tract protection equipment |
| CN111753805A (en) * | 2020-07-08 | 2020-10-09 | 深延科技(北京)有限公司 | Method and device for detecting wearing of safety helmet |
| CN111914743A (en) * | 2020-07-31 | 2020-11-10 | 广东电网有限责任公司清远供电局 | Method and device for detecting safety helmet of transformer substation worker |
| CN111950457A (en) * | 2020-08-12 | 2020-11-17 | 海默潘多拉数据科技(深圳)有限公司 | Oil field safety production image identification method and system |
| CN112070043A (en) * | 2020-09-15 | 2020-12-11 | 常熟理工学院 | Safety helmet wearing convolutional network based on feature fusion, training and detecting method |
| CN112070043B (en) * | 2020-09-15 | 2023-11-10 | 常熟理工学院 | Feature fusion-based safety helmet wearing convolution network, training and detection method |
| CN112215809A (en) * | 2020-09-25 | 2021-01-12 | 北京林业大学 | Prawn processing shelling rate detection method and device, electronic equipment and storage medium |
| CN112183532A (en) * | 2020-09-28 | 2021-01-05 | 福建师范大学福清分校 | Safety helmet identification method based on weak supervision collaborative learning algorithm and storage medium |
| CN112464822A (en) * | 2020-11-30 | 2021-03-09 | 深圳市捷顺科技实业股份有限公司 | Helmet wearing detection method and device based on feature enhancement |
| CN112598055B (en) * | 2020-12-21 | 2022-06-17 | 电子科技大学 | Helmet wearing detection method, computer-readable storage medium and electronic device |
| CN112598055A (en) * | 2020-12-21 | 2021-04-02 | 电子科技大学 | Helmet wearing detection method, computer-readable storage medium and electronic device |
| CN114627425A (en) * | 2021-06-11 | 2022-06-14 | 珠海路讯科技有限公司 | Method for detecting whether worker wears safety helmet or not based on deep learning |
| CN113435369A (en) * | 2021-07-01 | 2021-09-24 | 华风数据(深圳)有限公司 | Photovoltaic fault intelligent identification method based on SSD algorithm model and far infrared video |
| CN113808197A (en) * | 2021-09-17 | 2021-12-17 | 山西大学 | Automatic workpiece grabbing system and method based on machine learning |
| CN114299429A (en) * | 2021-12-24 | 2022-04-08 | 宁夏广天夏电子科技有限公司 | Human body recognition method, system and device based on deep learning |
| CN114595759A (en) * | 2022-03-07 | 2022-06-07 | 卡奥斯工业智能研究院(青岛)有限公司 | Protective tool identification method and device, electronic equipment and storage medium |
| CN114782986A (en) * | 2022-03-28 | 2022-07-22 | 佳源科技股份有限公司 | Helmet wearing detection method, device, equipment and medium based on deep learning |
| CN114782986B (en) * | 2022-03-28 | 2024-04-26 | 佳源科技股份有限公司 | Deep learning-based safety helmet wearing detection method, device, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110728223A (en) | Helmet wearing identification method based on deep learning | |
| CN106600888B (en) | Automatic forest fire detection method and system | |
| CN112435221B (en) | Image anomaly detection method based on generated countermeasure network model | |
| CN104123544B (en) | Anomaly detection method and system based on video analysis | |
| CN104063722B (en) | A kind of detection of fusion HOG human body targets and the safety cap recognition methods of SVM classifier | |
| CN107451999A (en) | foreign matter detecting method and device based on image recognition | |
| CN110378232B (en) | Improved test room examinee position rapid detection method of SSD dual-network | |
| CN108615226A (en) | A kind of image defogging method fighting network based on production | |
| CN111611907B (en) | Image-enhanced infrared target detection method | |
| CN106056079A (en) | Image acquisition device and facial feature occlusion detection method | |
| CN111126293A (en) | Flame and smoke abnormal condition detection method and system | |
| CN104463253B (en) | Passageway for fire apparatus safety detection method based on adaptive background study | |
| CN107798279A (en) | Face living body detection method and device | |
| CN111091110B (en) | Reflection vest wearing recognition method based on artificial intelligence | |
| CN111626169B (en) | Image-based railway dangerous falling rock size judgment method | |
| CN111507227B (en) | Multi-student individual segmentation and state autonomous identification method based on deep learning | |
| CN111401310B (en) | Kitchen sanitation safety supervision and management method based on artificial intelligence | |
| CN111046827A (en) | Video smoke detection method based on convolutional neural network | |
| CN112183472A (en) | Method for detecting whether test field personnel wear work clothes or not based on improved RetinaNet | |
| CN109034038B (en) | Fire identification device based on multi-feature fusion | |
| CN109063589A (en) | Instrument and equipment on-line monitoring method neural network based and system | |
| CN109325472B (en) | Face living body detection method based on depth information | |
| CN106372566A (en) | Digital signage-based emergency evacuation system and method | |
| CN107704818A (en) | A kind of fire detection system based on video image | |
| CN114332739A (en) | Smoke detection method based on moving target detection and deep learning technology |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200124 |