CN112184773A - Helmet wearing detection method and system based on deep learning - Google Patents
Helmet wearing detection method and system based on deep learning Download PDFInfo
- Publication number
- CN112184773A CN112184773A CN202011064308.4A CN202011064308A CN112184773A CN 112184773 A CN112184773 A CN 112184773A CN 202011064308 A CN202011064308 A CN 202011064308A CN 112184773 A CN112184773 A CN 112184773A
- Authority
- CN
- China
- Prior art keywords
- safety helmet
- detection
- headgear
- wearing
- moving
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20228—Disparity calculation for image-based rendering
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a safety helmet wearing detection method and system based on deep learning, belonging to the field of security monitoring and comprising the following steps: s1, at the server side, moving object detection is carried out on all the camera videos on site, and all moving objects in the camera videos are extracted to obtain moving object images; s2, inputting the moving target image into a pre-trained safety helmet detection model, and detecting the wearing condition of the safety helmet; the safety helmet detection model is a deep learning model; the data set used to train the headgear detection model includes an image that is annotated as to whether a worker is wearing a headgear. The method fully utilizes the human body information of field workers and video data of a large number of cameras, filters invalid information in the camera videos by extracting the moving targets in the camera videos, and then detects moving target images based on the deep learning model to obtain the wearing condition of the safety helmet, and is high in accuracy and speed.
Description
Technical Field
The invention belongs to the field of security monitoring, and particularly relates to a safety helmet wearing detection method and system based on deep learning.
Background
In the building construction operation process, there are many potential safety hazards for the incidence of accident is high. In the practical process, the detection on the behavior ability of construction workers and the wearing of safety facilities before the construction operation is carried out is found, so that the probability of accidents can be effectively reduced. Therefore, in daily construction work, it is important to supervise whether safety facilities such as safety helmets of workers are worn or not. However, most of the existing sites adopt manual monitoring, depend on field-experienced managers, require the managers to observe and inspect in real time, are time-consuming and labor-consuming, have the conditions of low automation level, large workload, limited inspection projects, and are easy to cause potential safety hazards due to the fact that the inspection is missed and the like.
In order to solve the above problems, in the prior art, one method is to detect the wearing condition of the safety helmet by using a Viola-Jones detector, and express the safety helmet by using two rectangular features in the edge features, and when the background is complicated or a shielded scene exists, the detection result is not very stable, and the accuracy is low. The other method is that the safety helmet is connected with a sensing chip, a positioning identification chip and a voice chip into a whole, a Zigbee network is erected in a construction site, signals are transmitted between the safety helmet and a central control room through the Zigbee wireless network, the sensing chip in the safety helmet can detect and monitor the environmental condition of a production site in real time, and detected information is converted into signals which are transmitted to the central control room through the Zigbee wireless network and are processed and analyzed. The method depends on a hardware chip, has higher cost and has higher requirements on scenes. Moreover, the RGB color space representation is used in the deployment, the RGB color space representation is extremely sensitive to the change of illumination, the color of the safety helmet cannot be reflected through a relatively stable threshold value, and the accuracy is reduced.
Disclosure of Invention
In view of the above defects or improvement requirements of the prior art, the present invention provides a method and a system for detecting wearing of a safety helmet based on deep learning, and aims to solve the technical problem of low accuracy in the prior art.
In order to achieve the above object, in a first aspect, the present invention provides a method for detecting wearing of a safety helmet based on deep learning, including the following steps:
s1, at the server side, moving object detection is carried out on all the camera videos on site, and all moving objects in the camera videos are extracted to obtain moving object images;
s2, inputting the moving target image into a pre-trained safety helmet detection model, and detecting the wearing condition of the safety helmet;
the safety helmet detection model is a deep learning model; the data set used to train the crash helmet detection model includes an image that indicates whether a worker is wearing a crash helmet.
Further preferably, the above S1 includes the steps of:
s11, acquiring a foreground target image of a current frame in the camera video by adopting a background difference method;
s12, respectively carrying out edge detection on the current frame and the previous frame and the next frame of the current frame; carrying out difference operation on the edge detection images of the current frame and the previous frame to obtain a difference image; carrying out XOR operation on the edge detection images of the current frame and the next frame to obtain an XOR image;
s13, after the obtained exclusive-or image and the foreground target image of the current frame are subjected to OR operation, and the obtained result and the obtained difference image are subjected to AND operation;
s14, performing morphological processing and continuity analysis on the obtained and operation result to obtain a binary template of the moving target, and extracting each moving target in the current frame based on the obtained binary template to obtain a moving target image;
and S15, repeating the steps S11-S14 to iterate until the moving object extraction of all images in the camera video is completed.
Further preferably, the helmet detection model is the Yolov5 model.
Further preferably, if it is detected that the crash helmet is not worn, an alarm is issued.
In a second aspect, the present invention provides a deep learning based helmet wearing detection system, including: the safety helmet wearing condition detection module comprises a moving target image extraction module and a safety helmet wearing condition detection module;
the moving target image extraction module is used for detecting moving objects of all camera videos on site, extracting all moving targets in the camera videos to obtain moving target images, and inputting the moving target images into the safety helmet wearing condition detection module;
the safety helmet wearing condition detection module is used for inputting images of moving objects into a safety helmet detection model which is trained in advance and detecting the wearing condition of the safety helmet;
the safety helmet detection model is a deep learning model; the data set used to train the crash helmet detection model includes an image that indicates whether a worker is wearing a crash helmet.
Further preferably, the system for detecting wearing of safety helmet based on deep learning provided by the second aspect of the present invention further comprises: and the warning module is used for sending out a warning when the safety helmet is detected not worn.
Further preferably, the helmet wearing detection system based on deep learning provided by the second aspect of the present invention is deployed on a server.
In a third aspect, the present invention further provides a computer-readable storage medium, where the computer-readable storage medium includes a stored computer program, where the computer program, when executed by a processor, controls an apparatus in which the storage medium is located to perform the method for detecting wearing of a safety helmet based on deep learning according to the first aspect of the present invention.
In general, compared with the prior art, the above technical solution contemplated by the present invention can achieve the following beneficial effects:
1. the invention provides a safety helmet wearing detection method based on deep learning, which fully utilizes human body information of field workers and video data of a large number of cameras, filters invalid information in the camera videos by extracting moving targets in the camera videos, and then detects moving target images based on a deep learning model to obtain the wearing condition of the safety helmet, and is high in accuracy and speed.
2. The safety helmet wearing detection method based on deep learning provided by the invention is characterized in that a safety helmet detection model is a Yolov5 model which has a dense connection network, so that fusion of multilayer characteristics can be realized, and characteristics of each layer are fully utilized; in addition, a mobile network strategy is introduced into the model, and the standard convolution is segmented into a depth-separable convolution and a 1 x 1 convolution, so that the calculation amount and the size of the model can be obviously reduced; in addition, in the training process of the model, the anchor frame is calculated in a self-adaptive mode, and the accuracy is high.
3. According to the method for detecting the wearing of the safety helmet based on the deep learning, the video data of all the cameras on the construction site are analyzed, so that whether a worker wears the safety helmet or not can be detected when the worker enters the construction site, the result is automatically obtained and fed back to a construction site manager, meanwhile, a reminder can be given to a constructor who does not wear the safety helmet, convenience is provided for construction site management, the workload of security personnel is greatly reduced, the possibility of hidden danger caused by manual neglect is reduced, and a safer construction environment is provided for the construction site; in addition, the invention fully utilizes the video data of the camera on the construction site, only one server needs to be deployed on the hardware, a large amount of hardware facilities do not need to be additionally equipped, and the cost is lower.
Drawings
Fig. 1 is a flowchart of a method for detecting wearing of a safety helmet based on deep learning according to embodiment 1 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Examples 1,
A method for detecting wearing of safety helmets based on deep learning is disclosed, as shown in figure 1, and comprises the following steps:
s1, at the server side, moving object detection is carried out on all the camera videos on site, and all moving objects in the camera videos are extracted to obtain moving object images;
in this embodiment, the video data of the cameras at the entrance and exit of the site are acquired, and after being transmitted to the server, the moving object detection is performed on all the camera videos at the site at the server side, specifically, the method includes the following steps:
s11, acquiring a foreground target image of a current frame in the camera video by adopting a background difference method;
s12, respectively carrying out edge detection on the current frame and the previous frame and the next frame of the current frame; carrying out difference operation on the edge detection images of the current frame and the previous frame to obtain a difference image; carrying out XOR operation on the edge detection images of the current frame and the next frame to obtain an XOR image; specifically, a Canny algorithm is adopted to carry out edge detection on a current frame and a previous frame and a next frame of the current frame respectively;
s13, after the obtained exclusive-or image and the foreground target image of the current frame are subjected to OR operation, and the obtained result and the obtained difference image are subjected to AND operation;
and S14, performing morphological processing and continuity analysis on the obtained and operation result to obtain a binary template of the moving target, and extracting each moving target in the current frame based on the obtained binary template to obtain a moving target image. It should be noted that, in the binary template of the moving object, the position of the moving object takes a value of 1, and the rest positions take values of 0;
and S15, repeating the steps S11-S14 to iterate until the moving object extraction of all images in the camera video is completed.
The method is simple to implement, is not influenced by light and dynamic scene change, avoids the phenomena of false edges, cavities and the like, has good robustness, is quicker to update compared with other methods, has strong adaptability and meets the requirement of real-time property.
S2, inputting the moving target image into a pre-trained safety helmet detection model, detecting the wearing condition of the safety helmet, and sending out a warning if the safety helmet is detected not to be worn; the safety helmet detection model is a deep learning model;
in this embodiment, the helmet detection model is Yolov5 model. The Yolov5 model is a dense connection network, and has the biggest characteristic that the input of each layer is connected with the outputs of all the previous layers, so that the fusion of multilayer characteristics can be realized, and the characteristics of each layer are fully utilized. And by introducing a mobile network strategy, the standard convolution is segmented into a depth-separable convolution and a 1 x 1 convolution, so that the calculation amount and the model size can be obviously reduced. Specifically, the convolutional neural network of the Yolov5 model divides the input picture into grids with the size of S × S, and each cell is responsible for detecting the targets whose central points fall on the cell. Each cell predicts B bounding boxes and the confidence of the bounding boxes. The confidence level actually includes two aspects, namely the probability size of the bounding box containing the target and the accuracy of the bounding box. The former is denoted as pr (object), and when the bounding box is background (i.e. contains no object), pr (object) is 0; and Pr (object) is 1 when the bounding box contains the target. The accuracy of the bounding box can be characterized by the intersection over Intersection (IOU) of the predicted box and the actual box, and is noted as
And with a confidence of
The size and position of the bounding box can be characterized by 4 values: (x, y, w, h), where (x, y) is the center coordinates of the bounding box, and w and h are the width and height of the bounding box. It is also noted that the predicted value of the center coordinate (x, y) is the offset value relative to the upper left coordinate point of each cell, and the unit is relative to the cell size, and the predicted values of w and h for the bounding box are the ratio of width to height relative to the entire picture, so that theoretically the size of the 4 elements should be [0,1]And (3) a range. So each edgeThe predicted value of the bounding box actually contains 5 elements: (x, y, w, h, c), where the first 4 characterize the size and position of the bounding box and the last value is the confidence.
Further, the data set used to train the crash helmet detection model includes an image that is annotated as to whether the worker is wearing a crash helmet, wherein the image that the worker is wearing the crash helmet is a positive sample and the other images are negative samples. In this embodiment, the Yolov5 model was trained using a safety helmet wearing detection Dataset (SHWD) containing 7581 images in total, including 9044 safety helmet-worn bounding boxes (positive class) and 111514 non-worn bounding boxes (negative class), all of which are labeled with their target regions and categories. It should be noted that, in the training process of the Yolov5 model, a self-adaptive anchor frame calculation process is introduced, and the selection of positive and negative samples is automatically performed according to the relevant statistical characteristics of the target. For each actual frame, firstly finding k candidate anchor frames (non-prediction results) with the nearest central points in each feature layer, calculating the intersection ratio between the candidate anchor frames and the actual frame, and calculating the mean value m of the intersection ratiogAnd standard deviation vgTo obtain the cross-over ratio threshold tg＝mg+vgFinally, the threshold is selected to be greater than tgAs the final output. If the candidate anchor box corresponds to multiple actual boxes, the actual box with the largest intersection is selected.
By the method, whether the worker wears the safety helmet or not is accurately detected when the worker enters the construction site, the result is automatically obtained and fed back to construction site management personnel, convenience is provided for construction site management, and the workload of security personnel is greatly reduced.
Examples 2,
A deep learning based headgear wear detection system deployed on a server, comprising: the safety helmet wearing condition detection module comprises a moving target image extraction module and a safety helmet wearing condition detection module;
the moving object image extraction module is used for detecting moving objects of all camera videos on site, extracting all moving objects in the camera videos to obtain moving object images, and inputting the moving object images into the safety helmet wearing condition detection module;
the safety helmet wearing condition detection module is used for inputting images of moving objects into a safety helmet detection model which is trained in advance and detecting the wearing condition of the safety helmet;
the safety helmet detection model is a deep learning model; the data set used to train the crash helmet detection model includes an image that indicates whether a worker is wearing a crash helmet.
The related technical solution is the same as embodiment 1, and is not described herein again.
Examples 3,
A computer-readable storage medium, which includes a stored computer program, wherein when the computer program is executed by a processor, the apparatus where the storage medium is located is controlled to execute a method for detecting wearing of a safety helmet based on deep learning according to a first aspect of the present invention. The related technical solution is the same as embodiment 1, and is not described herein again.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (8)
1. A safety helmet wearing detection method based on deep learning is characterized by comprising the following steps:
s1, at the server side, moving object detection is carried out on all the camera videos on site, and all moving objects in the camera videos are extracted to obtain moving object images;
s2, inputting the moving target image into a pre-trained safety helmet detection model, and detecting the wearing condition of the safety helmet;
the safety helmet detection model is a deep learning model; the data set used to train the headgear detection model includes an image that is annotated as to whether a worker is wearing a headgear.
2. The headgear wearing detection method according to claim 1, wherein the S1 includes the steps of:
s11, acquiring a foreground target image of a current frame in the camera video by adopting a background difference method;
s12, respectively carrying out edge detection on the current frame and the previous frame and the next frame of the current frame; carrying out difference operation on the edge detection images of the current frame and the previous frame to obtain a difference image; carrying out XOR operation on the edge detection images of the current frame and the next frame to obtain an XOR image;
s13, after the exclusive-OR image and the foreground target image of the current frame are subjected to OR operation, and the obtained result and the difference image are subjected to AND operation;
s14, performing morphological processing and continuity analysis on the AND operation result of S13 to obtain a binary template of the moving target, and extracting each moving target in the current frame based on the binary template to obtain a moving target image;
and S15, repeating the steps S11-S14 to iterate until the extraction of the moving objects in all the images of the camera video is completed.
3. The headgear wear detection method of claim 1, wherein the headgear detection model is a Yolov5 model.
4. The headgear wearing detection method according to any one of claims 1 to 3, wherein a warning is issued if it is detected that the headgear is not worn.
5. A helmet wearing detection system based on deep learning is characterized by comprising: the safety helmet wearing condition detection module comprises a moving target image extraction module and a safety helmet wearing condition detection module;
the moving object image extraction module is used for detecting moving objects of all camera videos on site, extracting all moving objects in the camera videos to obtain moving object images, and inputting the moving object images into the safety helmet wearing condition detection module;
the safety helmet wearing condition detection module is used for inputting images of moving objects into a safety helmet detection model which is trained in advance and detecting the wearing condition of the safety helmet;
the safety helmet detection model is a deep learning model; the data set used to train the headgear detection model includes an image that is annotated as to whether a worker is wearing a headgear.
6. The headgear wear detection system of claim 5, further comprising a warning module for issuing a warning when an unworn headgear is detected.
7. The headgear wear detection system of claim 5 or 6, deployed on a server.
8. A computer-readable storage medium, comprising a stored computer program, wherein the computer program, when executed by a processor, controls an apparatus in which the storage medium is located to perform the method of headgear wear detection of any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011064308.4A CN112184773A (en) | 2020-09-30 | 2020-09-30 | Helmet wearing detection method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011064308.4A CN112184773A (en) | 2020-09-30 | 2020-09-30 | Helmet wearing detection method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112184773A true CN112184773A (en) | 2021-01-05 |
Family
ID=73947703
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011064308.4A Pending CN112184773A (en) | 2020-09-30 | 2020-09-30 | Helmet wearing detection method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112184773A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112784821A (en) * | 2021-03-06 | 2021-05-11 | 深圳市安比智慧科技有限公司 | Building site behavior safety detection and identification method and system based on YOLOv5 |
| CN112966618A (en) * | 2021-03-11 | 2021-06-15 | 京东数科海益信息科技有限公司 | Dressing identification method, device, equipment and computer readable medium |
| CN113408365A (en) * | 2021-05-26 | 2021-09-17 | 广东能源集团科学技术研究院有限公司 | Method and device for identifying safety helmet in complex scene |
| CN113838094A (en) * | 2021-09-28 | 2021-12-24 | 贵州乌江水电开发有限责任公司 | Safety early warning method based on intelligent video identification |
| CN116959028A (en) * | 2023-07-20 | 2023-10-27 | 山东产研鲲云人工智能研究院有限公司 | Method for supervising safety of high-altitude operation, inspection equipment and computing equipment |
| CN117095273A (en) * | 2023-10-13 | 2023-11-21 | 中山大学深圳研究院 | Online collaborative industrial vision detection method and terminal equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102024146A (en) * | 2010-12-08 | 2011-04-20 | 江苏大学 | Method for extracting foreground in piggery monitoring video |
| US20150334398A1 (en) * | 2014-05-15 | 2015-11-19 | Daniel Socek | Content adaptive background foreground segmentation for video coding |
| JP2018025914A (en) * | 2016-08-09 | 2018-02-15 | 大日本印刷株式会社 | Detection device for individual in shot image |
| CN108319926A (en) * | 2018-02-12 | 2018-07-24 | 安徽金禾软件股份有限公司 | A kind of the safety cap wearing detecting system and detection method of building-site |
| CN108921878A (en) * | 2018-04-30 | 2018-11-30 | 武汉工程大学 | Hazardous gas spillage infrared video detection method under moving-target low contrast |
| CN109377515A (en) * | 2018-08-03 | 2019-02-22 | 佛山市顺德区中山大学研究院 | A kind of moving target detecting method and system based on improvement ViBe algorithm |
| CN110188724A (en) * | 2019-06-05 | 2019-08-30 | 中冶赛迪重庆信息技术有限公司 | The method and system of safety cap positioning and color identification based on deep learning |
-
2020
- 2020-09-30 CN CN202011064308.4A patent/CN112184773A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102024146A (en) * | 2010-12-08 | 2011-04-20 | 江苏大学 | Method for extracting foreground in piggery monitoring video |
| US20150334398A1 (en) * | 2014-05-15 | 2015-11-19 | Daniel Socek | Content adaptive background foreground segmentation for video coding |
| JP2018025914A (en) * | 2016-08-09 | 2018-02-15 | 大日本印刷株式会社 | Detection device for individual in shot image |
| CN108319926A (en) * | 2018-02-12 | 2018-07-24 | 安徽金禾软件股份有限公司 | A kind of the safety cap wearing detecting system and detection method of building-site |
| CN108921878A (en) * | 2018-04-30 | 2018-11-30 | 武汉工程大学 | Hazardous gas spillage infrared video detection method under moving-target low contrast |
| CN109377515A (en) * | 2018-08-03 | 2019-02-22 | 佛山市顺德区中山大学研究院 | A kind of moving target detecting method and system based on improvement ViBe algorithm |
| CN110188724A (en) * | 2019-06-05 | 2019-08-30 | 中冶赛迪重庆信息技术有限公司 | The method and system of safety cap positioning and color identification based on deep learning |
Non-Patent Citations (3)
| Title |
|---|
| LUO YANLI ET AL.: "Adaptive moving vehicle detection algorithm based on hybrid background subtraction and frame difference", 《TRANS TECH PUBLICATIONS》 * |
| WILLIAM: "一文读懂yolov5与yolov4", 《HTTP：//ZHUANLAN.ZHIHU.COM/P/161083602》 * |
| 郑茂凯: "运动目标识别与跟踪方法的研究", 《中国优秀博硕士学位论文全文数据库（硕士） 信息科技辑》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112784821A (en) * | 2021-03-06 | 2021-05-11 | 深圳市安比智慧科技有限公司 | Building site behavior safety detection and identification method and system based on YOLOv5 |
| CN112966618A (en) * | 2021-03-11 | 2021-06-15 | 京东数科海益信息科技有限公司 | Dressing identification method, device, equipment and computer readable medium |
| CN112966618B (en) * | 2021-03-11 | 2024-02-09 | 京东科技信息技术有限公司 | Dressing recognition method, apparatus, device and computer readable medium |
| CN113408365A (en) * | 2021-05-26 | 2021-09-17 | 广东能源集团科学技术研究院有限公司 | Method and device for identifying safety helmet in complex scene |
| CN113408365B (en) * | 2021-05-26 | 2023-09-08 | 广东能源集团科学技术研究院有限公司 | Safety helmet identification method and device under complex scene |
| CN113838094A (en) * | 2021-09-28 | 2021-12-24 | 贵州乌江水电开发有限责任公司 | Safety early warning method based on intelligent video identification |
| CN113838094B (en) * | 2021-09-28 | 2024-03-05 | 贵州乌江水电开发有限责任公司 | Safety early warning method based on intelligent video identification |
| CN116959028A (en) * | 2023-07-20 | 2023-10-27 | 山东产研鲲云人工智能研究院有限公司 | Method for supervising safety of high-altitude operation, inspection equipment and computing equipment |
| CN116959028B (en) * | 2023-07-20 | 2024-03-01 | 山东产研鲲云人工智能研究院有限公司 | Method for supervising safety of high-altitude operation, inspection equipment and computing equipment |
| CN117095273A (en) * | 2023-10-13 | 2023-11-21 | 中山大学深圳研究院 | Online collaborative industrial vision detection method and terminal equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112184773A (en) | Helmet wearing detection method and system based on deep learning | |
| CN111898514B (en) | Multi-target visual supervision method based on target detection and action recognition | |
| CN111881730A (en) | Wearing detection method for on-site safety helmet of thermal power plant | |
| US9412025B2 (en) | Systems and methods to classify moving airplanes in airports | |
| CN111241959A (en) | Method for detecting person without wearing safety helmet through construction site video stream | |
| CN102509075A (en) | Remnant object detection method and device | |
| CN104966304A (en) | Kalman filtering and nonparametric background model-based multi-target detection tracking method | |
| CN114241298A (en) | Tower crane environment target detection method and system based on laser radar and image fusion | |
| KR20190046351A (en) | Method and Apparatus for Detecting Intruder | |
| CN104680145B (en) | The on off state change detecting method and device of a kind of | |
| CN111062303A (en) | Image processing method, system and computer storage medium | |
| CN112784821A (en) | Building site behavior safety detection and identification method and system based on YOLOv5 | |
| CN112183472A (en) | Method for detecting whether test field personnel wear work clothes or not based on improved RetinaNet | |
| CN112819068A (en) | Deep learning-based real-time detection method for ship operation violation behaviors | |
| CN112287823A (en) | Facial mask identification method based on video monitoring | |
| CN113723300A (en) | Artificial intelligence-based fire monitoring method and device and storage medium | |
| CN112532927A (en) | Intelligent safety management and control system for construction site | |
| CN115995097A (en) | Deep learning-based safety helmet wearing standard judging method | |
| CN114169404A (en) | Method for intelligently acquiring quantitative information of slope diseases based on images | |
| Tao | Statistical calculation of dense crowd flow antiobscuring method considering video continuity | |
| CN114979567B (en) | Object and region interaction method and system applied to video intelligent monitoring | |
| CN114783000B (en) | Method and device for detecting dressing standard of worker in bright kitchen range scene | |
| CN115578827B (en) | Method and system for tracking dangerous chemical article receiving container based on AI video | |
| CN114022685B (en) | Method and system for detecting position of security officer | |
| CN118015330A (en) | Method and equipment for identifying personnel intrusion during operation of coal conveyor and computing device |
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 |
Application publication date: 20210105 |
|
| RJ01 | Rejection of invention patent application after publication |