CN110046557A - Safety cap, Safe belt detection method based on deep neural network differentiation - Google Patents
Safety cap, Safe belt detection method based on deep neural network differentiation Download PDFInfo
- Publication number
- CN110046557A CN110046557A CN201910237877.5A CN201910237877A CN110046557A CN 110046557 A CN110046557 A CN 110046557A CN 201910237877 A CN201910237877 A CN 201910237877A CN 110046557 A CN110046557 A CN 110046557A
- Authority
- CN
- China
- Prior art keywords
- safety
- neural network
- cap
- deep neural
- belt
- 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
Classifications
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Abstract
The invention discloses a kind of safety caps differentiated based on deep neural network, Safe belt detection method, include: the human body image data training deep neural network with safety cap, safety belt image data and non-safe wearing cap, safety belt, obtains the deep neural network model that can detect safety cap, belt position in the picture；Safety cap, seatbelt wearing scene are monitored using video capture device, acquire video image to be detected；Video image to be detected is detected using obtained deep neural network model, determine in video human body whether safe wearing cap, safety belt.The present invention passes through deep neural network model and the spatial coherence model of safety cap, safety belt and human body, worker's image in video is detected, and differentiate its safe wearing cap, safety belt situation, influence of the factors such as construction site cement to picture quality effectively is avoided, improves the precision of safety cap, safety belt detection.
Description
Technical field
The present invention relates to field more particularly to a kind of safety caps differentiated based on deep neural network, safety belt detection side
Method.
Background technique
As country increasingly payes attention to safety in production, every profession and trade also all adopt various measures ensure the safety in production of employee from
And the interests of enterprise are ensured.In building trade, worker on duty safe wearing cap and does not do related security dangerous operation,
As non-safe wearing cap, safety belt and caused by injures and deaths happen occasionally.Tradition can only lean on artificial inspection safety cap, safety belt
Wear condition, great waste of manpower.And the mode based on image procossing, since in building trade, the spots such as cement are to safety
Cap, the Imaging of safety belt are very big, the detection image for being difficult to obtain high quality are caused, so that verification and measurement ratio is not high.The present invention
It is directed to a kind of safety cap, Safe belt detection method for learning differentiation based on deep neural network, to improve Detection accuracy.
Summary of the invention
The purpose of the present invention is a kind of safety caps differentiated based on deep neural network study, Safe belt detection method, should
Method is by deep neural network model and the spatial coherence model of safety cap, safety belt and human body, to the work in video
People's image detects, and differentiates its safe wearing cap, safety belt situation, to improve Detection accuracy.
For achieving the above object, the technical scheme is that a kind of safety differentiated based on deep neural network
Cap, Safe belt detection method, comprising the following steps:
Step 1) is deep with the human body image data training of safety cap, safety belt image data and non-safe wearing cap, safety belt
Neural network is spent, the deep neural network model that can detect safety cap, belt position in the picture is obtained；
Step 2 is monitored safety cap, seatbelt wearing scene using video capture device, acquires video image to be detected；
Step 3) detects video image to be detected using deep neural network model obtained in step 1), determines video
Middle human body whether safe wearing cap, safety belt.
Further, the step 3) specifically:
Step 3.1) uses background modeling method, the moving region in video image to be detected is extracted, as area to be tested；
Step 3.2) uses human testing algorithm, identifies the human body image in area to be tested, obtains human detection result；
Step 3.3) is using head and trunk of the deep neural network model to the human body in area to be tested described in step 1)
Position is detected respectively, obtains safety cap or safety belt testing result；
Step 3.4) carries out integrated location space to the result that step 3.2), step 3.3) obtain using spatial coherence model and sentences
It is fixed, if calculated result is greater than threshold value, determines human body safe wearing cap or safety belt, otherwise do not wear.
Further, the human testing algorithm in the step 3.2) be SSD algorithm, to human body detected the result is that
One rectangle frame comprising pedestrian contour.
The beneficial effects of the present invention are: the present invention passes through deep neural network model and the space of safety cap, safety belt
Correlation models detect worker's image in video, and differentiate its safe wearing cap, safety belt situation, effectively keep away
Exempt from influence of the factors such as construction site cement to picture quality, improves the precision of safety cap, safety belt detection.
Detailed description of the invention
Fig. 1 is the calibration maps of safety cap, safety belt on human body.
Specific embodiment
Below in conjunction with attached drawing, technical scheme in the embodiment of the invention is clearly and completely described.
It is a kind of based on deep neural network differentiate safety cap, Safe belt detection method, comprising the following steps:
Step 1) is deep with the human body image data training of safety cap, safety belt image data and non-safe wearing cap, safety belt
Neural network is spent, the deep neural network model that can detect safety cap, belt position in the picture is obtained；
Step 2 is monitored safety cap, seatbelt wearing scene using video capture device, acquires video image to be detected；
Step 3) detects video image to be detected using deep neural network model obtained in step 1), determines video
Middle human body whether safe wearing cap, safety belt.Specifically:
Step 3.1) uses background modeling method, the moving region in video image to be detected is extracted, as area to be tested；
In order to improve detection speed, the moving region in video is detected using frame difference method, and using gaussian filtering and
Region growing algorithm removes the noise region of frame difference result；
Step 3.2) uses human testing algorithm, identifies the human body image in area to be tested, obtains human detection result；
Human testing algorithm is the SSD algorithm based on deep neural network, using the algorithm to the pedestrian contour of area to be tested
Mark rectangle frame to get to a human detection result p；
Head and metastomium of the step 3.3) using the SSD algorithm based on deep neural network to the human body in area to be tested
It is detected respectively, obtains safety cap or safety belt testing result, as shown in Figure 1；
Step 3.4) carries out integrated location space to the result that step 3.2), step 3.3) obtain using spatial coherence model and sentences
It is fixed, if calculated result is greater than threshold value, determines human body safe wearing cap or safety belt, otherwise do not wear.Integrated location is empty
Between decision process are as follows:
The result that step 3.2), step 3.3) are obtained substitutes into spatial coherence model
M=d (pos (h), pos (p)) * DET (h)+DET (p), M > N
d(x,y) = (x - y)*(x-y)
Wherein, M is detection score；D () is the spatial coherence model of safety cap, safe bandgap calibration position；Pos () is nerve
The coordinate that network exports；DET () is the confidence level that neural network exports；N is threshold value, as M > N, is determined as wearing
It wears, otherwise, is determined as not wearing.The model is intended to indicate that the matching of safety cap, safety belt and the pedestrian detected, improves
The precision of detection.
The present invention is by deep neural network model and the spatial coherence model of safety cap, safety belt, in video
Worker's image detect, and differentiate its safe wearing cap, safety belt situation, effectively avoid construction site cement etc. because
Influence of the element to picture quality improves the precision of safety cap, safety belt detection
Described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Based in the present invention
Embodiment, every other embodiment obtained by those of ordinary skill in the art without making creative efforts, all
Belong to the scope of the present invention.
Claims (3)
1. a kind of safety cap differentiated based on deep neural network, Safe belt detection method, which is characterized in that including following step
It is rapid:
Step 1) is deep with the human body image data training of safety cap, safety belt image data and non-safe wearing cap, safety belt
Neural network is spent, the deep neural network model that can detect safety cap, belt position in the picture is obtained；
Step 2 is monitored safety cap, seatbelt wearing scene using video capture device, acquires video image to be detected；
Step 3) detects video image to be detected using deep neural network model obtained in step 1), determines video
Middle human body whether safe wearing cap, safety belt.
2. a kind of safety cap differentiated based on deep neural network as described in claim 1, Safe belt detection method, feature
It is, the step 3) specifically:
Step 3.1) uses background modeling method, the moving region in video image to be detected is extracted, as area to be tested；
Step 3.2) uses human testing algorithm, identifies the human body image in area to be tested, obtains human detection result；
Step 3.3) is using head and trunk of the deep neural network model to the human body in area to be tested described in step 1)
Position is detected respectively, obtains safety cap or safety belt testing result；
Step 3.4) carries out integrated location space to the result that step 3.2), step 3.3) obtain using spatial coherence model and sentences
It is fixed, if calculated result is greater than threshold value, determines human body safe wearing cap or safety belt, otherwise do not wear.
3. a kind of safety cap differentiated based on deep neural network as claimed in claim 2, Safe belt detection method, feature
Be, the human testing algorithm in the step 3.2) be SSD algorithm, to human body detected the result is that one include pedestrian
The rectangle frame of profile.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910237877.5A CN110046557A (en) | 2019-03-27 | 2019-03-27 | Safety cap, Safe belt detection method based on deep neural network differentiation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910237877.5A CN110046557A (en) | 2019-03-27 | 2019-03-27 | Safety cap, Safe belt detection method based on deep neural network differentiation |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110046557A true CN110046557A (en) | 2019-07-23 |
Family
ID=67275311
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910237877.5A Pending CN110046557A (en) | 2019-03-27 | 2019-03-27 | Safety cap, Safe belt detection method based on deep neural network differentiation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110046557A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222478A (en) * | 2020-01-11 | 2020-06-02 | 深圳奥腾光通系统有限公司 | Construction site safety protection detection method and system |
| CN111310592A (en) * | 2020-01-20 | 2020-06-19 | 杭州视在科技有限公司 | Detection method based on scene analysis and deep learning |
| 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 |
| CN111669548A (en) * | 2020-06-04 | 2020-09-15 | 赛特斯信息科技股份有限公司 | Method for realizing safety supervision and treatment aiming at pole climbing operation of power distribution network |
| CN111931642A (en) * | 2020-08-07 | 2020-11-13 | 上海商汤临港智能科技有限公司 | Safety belt wearing detection method and device, electronic equipment and storage medium |
| CN112101260A (en) * | 2020-09-22 | 2020-12-18 | 广东电科院能源技术有限责任公司 | Method, device, equipment and storage medium for identifying safety belt of operator |
| 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 |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106682600A (en) * | 2016-12-15 | 2017-05-17 | 深圳市华尊科技股份有限公司 | Method and terminal for detecting targets |
| US20170169313A1 (en) * | 2015-12-14 | 2017-06-15 | Samsung Electronics Co., Ltd. | Image processing apparatus and method based on deep learning and neural network learning |
| CN107133564A (en) * | 2017-03-26 | 2017-09-05 | 天津普达软件技术有限公司 | A kind of frock work hat detection method |
| CN107666594A (en) * | 2017-09-18 | 2018-02-06 | 广东电网有限责任公司东莞供电局 | A kind of video monitoring monitors the method operated against regulations in real time |
| US20180150701A1 (en) * | 2016-11-29 | 2018-05-31 | Samsung Electronics Co., Ltd. | Method and apparatus for determining abnormal object |
| CN109117827A (en) * | 2018-09-05 | 2019-01-01 | 武汉市蓝领英才科技有限公司 | Work clothes work hat wearing state automatic identifying method and alarm system based on video |
-
2019
- 2019-03-27 CN CN201910237877.5A patent/CN110046557A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170169313A1 (en) * | 2015-12-14 | 2017-06-15 | Samsung Electronics Co., Ltd. | Image processing apparatus and method based on deep learning and neural network learning |
| US20180150701A1 (en) * | 2016-11-29 | 2018-05-31 | Samsung Electronics Co., Ltd. | Method and apparatus for determining abnormal object |
| CN106682600A (en) * | 2016-12-15 | 2017-05-17 | 深圳市华尊科技股份有限公司 | Method and terminal for detecting targets |
| CN107133564A (en) * | 2017-03-26 | 2017-09-05 | 天津普达软件技术有限公司 | A kind of frock work hat detection method |
| CN107666594A (en) * | 2017-09-18 | 2018-02-06 | 广东电网有限责任公司东莞供电局 | A kind of video monitoring monitors the method operated against regulations in real time |
| CN109117827A (en) * | 2018-09-05 | 2019-01-01 | 武汉市蓝领英才科技有限公司 | Work clothes work hat wearing state automatic identifying method and alarm system based on video |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111222478A (en) * | 2020-01-11 | 2020-06-02 | 深圳奥腾光通系统有限公司 | Construction site safety protection detection method and system |
| CN111310592A (en) * | 2020-01-20 | 2020-06-19 | 杭州视在科技有限公司 | Detection method based on scene analysis and deep learning |
| 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 |
| 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 |
| CN111669548A (en) * | 2020-06-04 | 2020-09-15 | 赛特斯信息科技股份有限公司 | Method for realizing safety supervision and treatment aiming at pole climbing operation of power distribution network |
| CN111669548B (en) * | 2020-06-04 | 2021-11-26 | 赛特斯信息科技股份有限公司 | Method for realizing safety supervision and treatment aiming at pole climbing operation of power distribution network |
| CN111931642A (en) * | 2020-08-07 | 2020-11-13 | 上海商汤临港智能科技有限公司 | Safety belt wearing detection method and device, electronic equipment and storage medium |
| CN112101260A (en) * | 2020-09-22 | 2020-12-18 | 广东电科院能源技术有限责任公司 | Method, device, equipment and storage medium for identifying safety belt of operator |
| CN112101260B (en) * | 2020-09-22 | 2023-09-26 | 南方电网电力科技股份有限公司 | Method, device, equipment and storage medium for identifying safety belt of operator |
| 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 |
| 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 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110046557A (en) | Safety cap, Safe belt detection method based on deep neural network differentiation | |
| CN109919132B (en) | Pedestrian falling identification method based on skeleton detection | |
| CN109522793B (en) | Method for detecting and identifying abnormal behaviors of multiple persons based on machine vision | |
| CN104554341B (en) | The system and method for detecting rail smooth degree | |
| CN109670396A (en) | A kind of interior Falls Among Old People detection method | |
| CN104063722B (en) | A kind of detection of fusion HOG human body targets and the safety cap recognition methods of SVM classifier | |
| CN106846734B (en) | A kind of fatigue driving detection device and method | |
| CN108734125A (en) | A kind of cigarette smoking recognition methods of open space | |
| CN108319934A (en) | Safety cap wear condition detection method based on video stream data | |
| CN102483851B (en) | Method and apparatus for prominent facial features recognition | |
| CN101893580A (en) | Digital image based detection method of surface flaw of steel rail | |
| US8965068B2 (en) | Apparatus and method for discriminating disguised face | |
| CN109002801A (en) | A kind of face occlusion detection method and system based on video monitoring | |
| CN114155601A (en) | Vision-based method and system for detecting dangerous behaviors of operating personnel | |
| CN103630093B (en) | Image analysis method for concrete surface roughness measurement | |
| CN110751051A (en) | Abnormal driving behavior detection method based on machine vision | |
| Bandini et al. | Automatic Detection of Orofacial Impairment in Stroke. | |
| CN103310444A (en) | Method of monitoring pedestrians and counting based on overhead camera | |
| CN104777522B (en) | Foreign matter detection system and detection method between a kind of Train door and platform screen door based on compound detection | |
| CN110796049A (en) | Production worker safety helmet wearing detection method and system based on image processing | |
| CN110136172A (en) | The detection method that safeguard is worn before a kind of miner goes into the well | |
| CN110096980A (en) | Character machining identifying system | |
| CN111401310B (en) | Kitchen sanitation safety supervision and management method based on artificial intelligence | |
| CN107463873A (en) | A kind of real-time gesture analysis and evaluation methods and system based on RGBD depth transducers | |
| CN115223249A (en) | Quick analysis and identification method for unsafe behaviors of underground personnel based on machine vision |
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: 20190723 |