CN112200092A - Intelligent smoking detection method based on variable-focus movement of dome camera - Google Patents
Intelligent smoking detection method based on variable-focus movement of dome camera Download PDFInfo
- Publication number
- CN112200092A CN112200092A CN202011089599.2A CN202011089599A CN112200092A CN 112200092 A CN112200092 A CN 112200092A CN 202011089599 A CN202011089599 A CN 202011089599A CN 112200092 A CN112200092 A CN 112200092A
- Authority
- CN
- China
- Prior art keywords
- face
- image
- camera
- calculating
- smoking
- 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.)
- Granted
Links
Images
Classifications
-
- 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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- 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
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- 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
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/242—Aligning, centring, orientation detection or correction of the image by image rotation, e.g. by 90 degrees
-
- 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
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/166—Detection; Localisation; Normalisation using acquisition arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/61—Control of cameras or camera modules based on recognised objects
- H04N23/611—Control of cameras or camera modules based on recognised objects where the recognised objects include parts of the human body
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/67—Focus control based on electronic image sensor signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/695—Control of camera direction for changing a field of view, e.g. pan, tilt or based on tracking of objects
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Studio Devices (AREA)
- Image Processing (AREA)
Abstract
An intelligent smoking detection method based on variable focus movement of a dome camera comprises the following steps: the method comprises the steps of firstly, initializing, secondly, inspecting by a dome camera, thirdly, detecting human faces, fourthly, calculating human face positions, fifthly, calculating human face ratios, sixthly, adjusting camera directions, seventhly, adjusting camera focal lengths, eighthly, acquiring a human face area diagram, nineteenth, smoking detection, tenthly, recognizing human faces, and eleventh, and recording warnings; the invention improves the traditional mode that the spherical camera needs to manually adjust the focal length and the direction, changes the traditional mode into the mode that the focal length and the direction are automatically adjusted by matching with an intelligent algorithm, improves the efficiency of the spherical camera, expands the application field of the spherical camera, can be used for training and detecting human faces and improves the accuracy rate of smoking detection.
Description
Technical Field
The invention relates to the technical field of image recognition, in particular to an intelligent smoking detection method based on variable-focus movement of a dome camera.
Background
The ball machine is a ball type camera and is a monitoring camera integrating a camera system, a zoom lens and an electronic pan-tilt. The ball machine can realize 360-degree rotation through the electronic holder system, and can realize full coverage of a monitoring area. The zoom lens of the dome camera can realize clear monitoring of far and near targets by changing the focal length. The spherical monitoring is mainly applied to household monitoring, public traffic safety monitoring and factory production safety monitoring.
Face detection and face recognition technologies are widely applied to various industries, face recognition is a means for performing biological identity verification by using face information, and face detection is a key step in face recognition and is effective detection on face images through the characteristic attributes of faces. The invention uses face detection to position the face of a person, detects whether cigarettes exist in a face image, and uses a face recognition technology to recognize and record the facial features and identities of smokers so as to realize the detection and the record.
According to statistics, a great part of fire disasters in the current park are caused by smoking, and laws and regulations for forbidding smoking in public places are set in each big city, but smoking behaviors are difficult to supervise, and difficult penalties are always a problem which plagues managers. The invention can effectively reduce the management difficulty of smoking detection in specific occasions by utilizing a visual detection technology, so that smoking behavior can be detected in real time and recorded on a case. The utilization scene of the invention is mainly concentrated in other indoor and outdoor smoke-forbidden areas such as gas stations, bus stations, railway stations, parks and the like.
Disclosure of Invention
The invention aims to provide an intelligent smoking detection method based on variable-focus movement of a dome camera, which can realize real-time movement detection of smoking behaviors.
The intelligent smoking detection method based on the variable focus movement of the dome camera adopts a learning training mode of a deep neural network, and designs an algorithm capable of detecting smoking behaviors from different angle distances according to the characteristic of the variable focus movement of the dome camera. The smoking behavior of specific occasions can be detected and identified in an all-round, multi-angle and long-distance mode. And face recognition is carried out on the smokers in specific occasions, and a blacklist library is established.
The method can utilize the deep neural network to train and identify the smoking photos, and simultaneously, the characteristic of movable rotation and zooming of the dome camera is utilized to design a smoking detection system capable of carrying out omnibearing and multi-angle detection, thereby greatly expanding the detectable range and precision. And finally, the face recognition technology is used for recognizing and recording the smokers, so that the functions of the system are improved, and the deterrence of the system is improved.
The initial state of the dome camera is long focus, the dome camera is inspected at a certain rotating speed, the human face is continuously identified in the inspection process, and the human face is identified by utilizing a neural network algorithm. After the face is recognized, the angle of the face is adjusted to enable the face picture to be centered, then the focal length of the camera is adjusted to amplify the face until the face proportion reaches one half of the picture. The mode of amplifying the human face is to calculate the distance to the human face by using the current focal distance human face ratio. The size of the face.
The cigarette has small target and single characteristic, and a good identification result is difficult to achieve by using the traditional method. In the past, the optimization of the algorithm is emphasized, and an identification method combining hardware and the algorithm is ignored. For the identification distance limited target which is difficult to overcome by the algorithm, the use efficiency of resources is increased, and the identification range is improved. By adopting the method of cloud platform inspection, the cost can be effectively saved, and the efficiency is improved.
The technical scheme provided by the invention is as follows: an intelligent smoking detection method based on variable focus movement of a dome camera comprises the following steps:
step one, initialization: initially, initializing a ball machine, setting the inspection mode of the ball machine to be from left to right, setting the focal length of the ball machine to be minimum, and starting to detect a scene;
step two, ball machine inspection; the face detection is carried out in the routing inspection process, and as the minimum focal length is set, the detection range is the widest, more faces can be detected;
step three, face detection: detecting a human face by using a human face detection technology;
step four, calculating the position of the face: calculating the position of the center point of the face in the image, and converting the position of the face on the image into the position under a camera coordinate system;
step five, calculating the face proportion: calculating the proportion of the human face in the image to judge the size of the human face;
step six, adjusting the direction of the camera: and adjusting the vertical and horizontal positions of the camera according to the position of the center point of the face in the image, so that the face is positioned in the center of the image.
Step seven, adjusting the focal length of the camera: and adjusting the focal length of the camera of the ball machine according to the position of the face in the image and the proportion of the face, so that the image of the face is centered and occupies half of the size of the image.
Step eight, obtaining a face region image: after the ball machine is adjusted in place, the face image is subjected to matting, and 20 pixels are respectively extended to four directions of the face image during matting to prepare for subsequent detection.
Step nine, smoking detection: after the face image is extracted, whether smoking behaviors exist in the image is detected by using a trained smoking detection algorithm.
Step ten, face recognition: when smoking behavior is detected, the human face is subjected to face recognition and recorded on a case.
Eleventh, warning recording: and (4) giving an alarm or penalty to the identified smoking personnel.
The intelligent smoking detection method based on the zoom movement of the dome camera has the following beneficial effects:
1. the invention adopts an advanced intelligent means to detect whether people smoke in a specific occasion, and changes the traditional mode of detecting smoking by using smoke sensation by using a video detection mode. The smoking behavior is detected in a video mode, the detection range is expanded, the detection efficiency is improved, and the detection cost is reduced.
2. The invention adopts a variable-focus movable dome camera to match with an intelligent detection means, improves the traditional mode that the dome camera needs to be manually adjusted in the past, changes the mode into the mode that the focus and the direction are automatically adjusted by matching with an intelligent algorithm, improves the efficiency of the dome camera, and expands the application field of the dome camera.
3. The invention designs a real-time light detection network which can be used for training and detecting human faces and detecting smoking, has simple structure and higher speed, can achieve the real-time detection effect on the edge side such as android equipment, and has higher accuracy, fewer network layers and higher speed than the traditional network.
Drawings
FIG. 1 is a block flow diagram of the present invention;
FIG. 2 is a schematic diagram of the present invention.
In the figure, 1 is a spherical monitoring camera; and 2, a cloud server platform.
Detailed Description
The invention is further illustrated by the following specific examples in conjunction with the accompanying drawings. Fig. 1 is a flow block diagram of the invention, and as shown in the drawing, in the intelligent smoking detection method based on zoom movement of a dome camera in the embodiment of the invention, a Hai Congwei-view 2DC6420IW-A dome camera is selected as an adaptive dome camera in the invention, the adaptive dome camera is provided with a cloud platform, and the position and direction of the dome camera can be adjusted by adjusting the steering torque of a motor of the cloud platform through commands. The method comprises the following steps:
step one, initializing S1: the method comprises the following steps that firstly, an algorithm deployed at the cloud end is used for carrying out initialization setting on a ball machine through an API (application program interface) of the ball machine, the inspection mode of the ball machine is set to be from left to right, the focal length of the ball machine is set to be minimum, after the setting is completed, the algorithm sets the ball machine to start to inspect a scene, and at the moment, the ball machine starts to rotate from left to right;
the implementation process of the invention adopts a method of deployment on a cloud-end platform, firstly, an algorithm is deployed on the cloud platform, the cloud platform and a ball machine are connected together through a network, the cloud platform can acquire images shot by the ball machine in real time, and meanwhile, the cloud platform detects the current state of the ball machine in real time, and the method comprises the following steps: the current angle of the ball machine and the focal distance of the ball machine. The image and the state of the ball machine are input into the algorithm, the new position of the ball machine is obtained through algorithm calculation, the image information after the ball machine is adjusted is obtained, the face of the image information obtained after adjustment is clearer, the information of the cigarette with the important smoking characteristic is amplified, and the smoking behavior can be easily detected by the smoking detection network. The deployment diagram is shown in fig. 2.
Step two, ball machine inspection S2; in the process of inspection, the dome camera continuously takes a snapshot, wherein the snapshot frequency is every second, ten photos are taken, and the snapshot photos are transmitted to the cloud end;
step three, face detection S3: the method comprises the steps that a face is detected by using a face detection algorithm of a cloud, face detection is carried out on each picture at the cloud, more faces can be detected due to the fact that the set minimum focal length is set and the detection range is the widest, and a detected face image is a minimum rectangular frame containing all the faces;
the invention provides a deep neural network-light weight detection network for training face detection and smoking detection, wherein the light weight detection neural network consists of a convolution network with six layers, and further comprises a classification layer and a positioning layer. The input of the picture is set to be 300X300, and the human face and smoking images can be effectively detected by using the input of the size. In order to effectively identify targets with different sizes by the neural network, the method effectively clusters the sizes of the human faces and the smoking photos, obtains effective sampling sizes, and improves the accuracy of the model.
Step four, calculating the face position S4: calculating the position of the center point of each detected face in the image, namely calculating the position (X, Y) of the center position of the rectangular frame in the whole image, calculating the actual position (X, Y, Z) of the face according to the position of the face on the image, and transmitting the actual position of the face to the step six;
the specific calculation process is that firstly, the vector distance from the center point of the face to the center point of the image, namely the distance containing the position relation of the angle, is calculated. And then, through the conversion of a coordinate system, the vector of the coordinate system is projected to the three-dimensional coordinate system space of the camera, wherein the space relation of the third dimension can be calculated through the proportion of the human face in the image under different focal lengths, and the distance between the camera and the target. The distance and the angle required for moving the visual center of the camera to the center of the alignment target can be obtained through the conversion of the coordinate system, so that the position of the camera can be adjusted. We rely on the pixel locations in the image to compute the image of the true location of the target as follows:
D＝(J×N)÷T (1)
wherein J is the focal length of the camera obtained by reading camera parameters, T is the pixel distance occupied in the image, N is an empirical coefficient obtained through experiments, different cameras are different coefficients, and D obtained through calculation is the actual distance of the target. Wherein (X, Y, Z) can be obtained by the formula (1);
step five, calculating the face proportion S5: calculating the proportion of the face in the image through the face, namely calculating the proportion of pixels occupied by each rectangular frame containing the face to pixels of the whole image to judge the proportion N of the face, and transmitting the data of the face proportion to the sixth step;
step six, adjusting the camera direction S6: and according to the face position (X, Y, Z) and the face proportion image size N acquired in the fourth step and the fifth step, calculating the current position of the camera to obtain positions D1 and D2 of the camera which need to move up, down, left and right, and enabling the face to be located at the center of the image.
Step seven, adjusting the focal length of the camera S7: when the camera adjusting position, let the people's face behind the central point, according to the size D that accounts for of people's face, the high in the clouds control ball machine camera adjusts the focus for the image of people's face is back placed in the middle, accounts for image size half at least, and N > is 0.5 promptly.
Step eight, acquiring a face region image S8: after the ball machine is adjusted in place, the face image is subjected to matting, namely a rectangular frame is scratched out of the image, and in the matting process, the rectangular frame containing the face needs to extend 20 pixels in four directions respectively to prepare for subsequent detection.
Step nine, smoking detection S9: after picking out the face image, sending the picked-out image block into an algorithm for smoking detection, detecting smoking behaviors in the image by using a trained smoking detection algorithm, and dividing a detection result into two types: smoking behavior and non-smoking behavior.
Step ten, face recognition S10: when smoking behavior is detected, the identity of the person who detects smoking behavior is authenticated, and at the moment, the person can recognize the smoking behavior or recognize the smoking behavior by using a mature face recognition method and record the identity.
Eleventh, warning record S11: alerting or penalising identified smokers
Fig. 2 is a schematic diagram of the implementation of the present invention, and as shown in fig. 2, the dome surveillance camera 1 is mainly used for acquiring images and feeding back the current state in real time: the method comprises the following steps: the cloud server 2 is mainly used for deploying algorithms, controlling the dome camera by using the algorithms after receiving the information from the dome camera, and detecting images by using the algorithms.
In order to verify the technical effect of the invention, through a simulation actual condition test, the inventor tests the accuracy rate calculated by 1000 times through a human simulation actual condition test, namely, through a smoking behavior of a person at a distance of 5 to 10m from a dome camera and a common monitoring camera, and simultaneously detects by respectively utilizing the zoom moving method of the dome camera and the monitoring camera which does not use the method of the invention, so as to obtain the following data. Specific values are detailed in table 1: the obtained beneficial effects are shown in the table.
Table 1: table of the beneficial effects obtained
| Smoking identification network identification accuracy | |
| Detection without ball machine cooperation | 85％ |
| Detection after ball machine matching | 98％ |
According to the intelligent smoking detection method based on variable-focus movement of the dome camera, the target can be automatically zoomed and amplified, a far-distance clearer target image can be obtained, even though the same detection algorithm is used in the background, the result shows that under the same condition, the smoking identification network recognition accuracy of detection matched with the dome camera is 98%, the smoking identification network recognition accuracy of detection not matched with the dome camera is 85%, the detection accuracy is improved by 13% compared with that of the detection method without the dome camera, and the practical effect of the method is obvious.
Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art will understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (6)
1. A functional smoking detection method based on zoom movement of a dome camera comprises the following steps:
step one, initialization: initially, initializing a ball machine, setting the inspection mode of the ball machine to be from left to right, setting the focal length of the ball machine to be minimum, and starting to detect a scene;
step two, ball machine inspection; the face detection is carried out in the routing inspection process, and as the minimum focal length is set, the detection range is the widest, more faces can be detected;
step three, face detection: detecting a human face by using a human face detection technology;
step four, calculating the position of the face: calculating the position of the center point of the face in the image, and converting the position of the face on the image into the position under a camera coordinate system;
step five, calculating the face proportion: calculating the proportion of the human face in the image to judge the size of the human face;
step six, adjusting the direction of the camera: and adjusting the vertical and horizontal positions of the camera according to the position of the center point of the face in the image, so that the face is positioned in the center of the image.
Step seven, adjusting the focal length of the camera: and adjusting the focal length of the camera of the ball machine according to the position of the face in the image and the proportion of the face, so that the image of the face is centered and occupies half of the size of the image.
Step eight, obtaining a face region image: after the ball machine is adjusted in place, the face image is subjected to matting, and 20 pixels are respectively extended to four directions of the face image during matting to prepare for subsequent detection.
Step nine, smoking detection: after the face image is extracted, whether smoking behaviors exist in the image is detected by using a trained smoking detection algorithm.
Step ten, face recognition: when smoking behavior is detected, the human face is subjected to face recognition and recorded on a case.
Eleventh, warning recording: and (4) giving an alarm or penalty to the identified smoking personnel.
2. The functional smoking detection method based on the zoom movement of the dome camera according to claim 1, wherein: and step four, calculating the face position, specifically, calculating the position of the center point of each detected face in the image, calculating the position (X, Y) of the center position of the rectangular frame in the whole image, calculating the actual position (X, Y, Z) of the face according to the position of the face on the image, and transmitting the actual position of the face to the step six.
3. The functional smoking detection method based on the zoom movement of the dome camera according to claim 2, wherein: step four, calculating the position of the face, which is characterized in that: the calculation process of calculating the actual position (X, Y, Z) of the face from the position of the face on the image is to first calculate the vector distance from the center point of the face to the center point of the image, i.e., the distance including the positional relationship of the angles. And then, through the conversion of a coordinate system, the vector of the coordinate system is projected to the three-dimensional coordinate system space of the camera, wherein the space relation of the third dimension can be calculated through the proportion of the human face in the image under different focal lengths, and the distance between the camera and the target. The distance and the angle required for moving the visual center of the camera to the center of the alignment target can be obtained through the conversion of the coordinate system, so that the position of the camera can be adjusted. We rely on the pixel locations in the image to compute the image of the true location of the target as follows:
D＝(J×N)÷T (1)
wherein J is the focal length of the camera obtained by reading camera parameters, T is the pixel distance occupied in the image, N is an empirical coefficient obtained through experiments, different cameras are different coefficients, and D obtained through calculation is the actual distance of the target. Wherein (X, Y, Z) can be obtained by the formula (1).
4. The functional smoking detection method based on the zoom movement of the dome camera according to claim 1, wherein: and step five, calculating the face proportion, namely calculating the proportion of the face in the image through the face, namely calculating the proportion of pixels occupied by each rectangular frame containing the face to pixels of the whole image to judge the proportion N of the face, and transmitting the face proportion data to the step six.
5. The functional smoking detection method based on the zoom movement of the dome camera according to claim 1, wherein: and step six, adjusting the direction of the camera, specifically, calculating positions D1 and D2 of the camera which need to move up and down and left and right according to the position (X, Y, Z) of the face and the size N of the face image acquired in the step four and the step five, and enabling the face to be located at the center of the image.
6. The functional smoking detection method based on the zoom movement of the dome camera according to claim 1, wherein: and seventhly, adjusting the focal length of the camera, specifically, when the camera is adjusted to be in a position, the camera of the cloud control dome camera adjusts the focal length according to the proportion D of the face, so that the image of the face is centered and at least occupies half of the image size, namely N > is 0.5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011089599.2A CN112200092B (en) | 2020-10-13 | 2020-10-13 | Intelligent smoking detection method based on zoom movement of dome camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011089599.2A CN112200092B (en) | 2020-10-13 | 2020-10-13 | Intelligent smoking detection method based on zoom movement of dome camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112200092A true CN112200092A (en) | 2021-01-08 |
| CN112200092B CN112200092B (en) | 2023-06-23 |
Family
ID=74010003
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011089599.2A Active CN112200092B (en) | 2020-10-13 | 2020-10-13 | Intelligent smoking detection method based on zoom movement of dome camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112200092B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113343859A (en) * | 2021-06-10 | 2021-09-03 | 浙江大华技术股份有限公司 | Smoking behavior detection method and device, storage medium and electronic device |
| CN113392710A (en) * | 2021-05-19 | 2021-09-14 | 上海可深信息科技有限公司 | Intelligent recognition method and system for smoking behavior |
| CN113392706A (en) * | 2021-05-13 | 2021-09-14 | 上海湃道智能科技有限公司 | Device and method for detecting smoking and using mobile phone behaviors |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101236599A (en) * | 2007-12-29 | 2008-08-06 | 浙江工业大学 | Human face recognition detection device based on multi- video camera information integration |
| CN107590461A (en) * | 2017-09-12 | 2018-01-16 | 广东欧珀移动通信有限公司 | Face identification method and Related product |
| CN108419014A (en) * | 2018-03-20 | 2018-08-17 | 北京天睿空间科技股份有限公司 | The method for capturing face using panoramic camera and the linkage of Duo Tai candid cameras |
| CN109711307A (en) * | 2018-12-19 | 2019-05-03 | 中科天网（广东）科技有限公司 | A kind of smoking evidence collecting method based on recognition of face |
| CN110569771A (en) * | 2019-08-30 | 2019-12-13 | 重庆博拉智略科技有限公司 | smoking identification method and device, storage medium and electronic equipment |
| CN110852275A (en) * | 2019-11-12 | 2020-02-28 | 成都新舟锐视科技有限公司 | Face detection re-amplification tracking method in dome camera tracking process |
| 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 |
-
2020
- 2020-10-13 CN CN202011089599.2A patent/CN112200092B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101236599A (en) * | 2007-12-29 | 2008-08-06 | 浙江工业大学 | Human face recognition detection device based on multi- video camera information integration |
| CN107590461A (en) * | 2017-09-12 | 2018-01-16 | 广东欧珀移动通信有限公司 | Face identification method and Related product |
| CN108419014A (en) * | 2018-03-20 | 2018-08-17 | 北京天睿空间科技股份有限公司 | The method for capturing face using panoramic camera and the linkage of Duo Tai candid cameras |
| CN109711307A (en) * | 2018-12-19 | 2019-05-03 | 中科天网（广东）科技有限公司 | A kind of smoking evidence collecting method based on recognition of face |
| CN110569771A (en) * | 2019-08-30 | 2019-12-13 | 重庆博拉智略科技有限公司 | smoking identification method and device, storage medium and electronic equipment |
| CN110852275A (en) * | 2019-11-12 | 2020-02-28 | 成都新舟锐视科技有限公司 | Face detection re-amplification tracking method in dome camera tracking process |
| 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 |
Non-Patent Citations (1)
| Title |
|---|
| 韩贵金 等: "基于Faster R-CNN的吸烟快速检测算法", 《西安邮电大学学报》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113392706A (en) * | 2021-05-13 | 2021-09-14 | 上海湃道智能科技有限公司 | Device and method for detecting smoking and using mobile phone behaviors |
| CN113392710A (en) * | 2021-05-19 | 2021-09-14 | 上海可深信息科技有限公司 | Intelligent recognition method and system for smoking behavior |
| CN113343859A (en) * | 2021-06-10 | 2021-09-03 | 浙江大华技术股份有限公司 | Smoking behavior detection method and device, storage medium and electronic device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112200092B (en) | 2023-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112200092A (en) | Intelligent smoking detection method based on variable-focus movement of dome camera | |
| CN101119482B (en) | Overall view monitoring method and apparatus | |
| CN108615321B (en) | Security pre-warning system and method based on radar detecting and video image behavioural analysis | |
| CN108062349B (en) | Video monitoring method and system based on video structured data and deep learning | |
| CN110675592B (en) | High-altitude parabolic early warning protection system based on target identification and control method | |
| CN101123722B (en) | Panorama video intelligent monitoring method and system | |
| CN109272530A (en) | Method for tracking target and device towards space base monitoring scene | |
| CN104794468A (en) | Human face detection and tracking method based on unmanned aerial vehicle mobile platform | |
| CN111062234A (en) | Monitoring method, intelligent terminal and computer readable storage medium | |
| CN109218667B (en) | Public place safety early warning system and method | |
| KR20180133745A (en) | Flying object identification system using lidar sensors and pan/tilt zoom cameras and method for controlling the same | |
| CN111679695A (en) | Unmanned aerial vehicle cruising and tracking system and method based on deep learning technology | |
| US20110157360A1 (en) | Surveillance system and method | |
| CN110991297A (en) | Target positioning method and system based on scene monitoring | |
| CN114998737A (en) | Remote smoke detection method, system, electronic equipment and medium | |
| CN106845410A (en) | A kind of flame identification method based on deep learning model | |
| CN104506827A (en) | Interactive visual shooting device having human body tracking function | |
| CN114120171A (en) | Fire smoke detection method, device and equipment based on video frame and storage medium | |
| CN115761627A (en) | Fire smoke flame image identification method | |
| CN115035470A (en) | Low, small and slow target identification and positioning method and system based on mixed vision | |
| CN111103891B (en) | Unmanned aerial vehicle rapid posture control system and method based on skeleton point detection | |
| CN111860457A (en) | Fighting behavior recognition early warning method and recognition early warning system thereof | |
| CN111466892A (en) | Remote temperature measurement system and method based on thermal imaging and face recognition | |
| CN108876824B (en) | Target tracking method, device and system and dome camera | |
| US11509831B2 (en) | Synchronous head movement (SHMOV) detection systems and methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |