CN108280952B - Passenger trailing monitoring method based on foreground object segmentation - Google Patents
Passenger trailing monitoring method based on foreground object segmentation Download PDFInfo
- Publication number
- CN108280952B CN108280952B CN201810074904.7A CN201810074904A CN108280952B CN 108280952 B CN108280952 B CN 108280952B CN 201810074904 A CN201810074904 A CN 201810074904A CN 108280952 B CN108280952 B CN 108280952B
- Authority
- CN
- China
- Prior art keywords
- monitoring
- monitoring system
- image
- control computer
- trailing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 155
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000011218 segmentation Effects 0.000 title claims abstract description 22
- 238000003062 neural network model Methods 0.000 claims abstract description 8
- 230000008859 change Effects 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 230000007547 defect Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000010191 image analysis Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000037361 pathway Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B13/00—Burglar, theft or intruder alarms
- G08B13/18—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength
- G08B13/189—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems
- G08B13/194—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems
- G08B13/196—Actuation by interference with heat, light, or radiation of shorter wavelength; Actuation by intruding sources of heat, light, or radiation of shorter wavelength using passive radiation detection systems using image scanning and comparing systems using television cameras
- G08B13/19602—Image analysis to detect motion of the intruder, e.g. by frame subtraction
-
- 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
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Burglar Alarm Systems (AREA)
Abstract
The invention discloses a passenger trailing monitoring method based on foreground object segmentation, which comprises the following steps: step S1, the monitoring system collects the image; step S2, mode selection; step S3, trailing monitoring: the monitoring system utilizes a deep neural network model to carry out pixel-level segmentation on the pedestrian image aiming at the collected image, then multi-frame comprehensive judgment is carried out on whether more than one person exists in the segmented image, and if yes, a trailing alarm signal is sent to a master control computer; step S4, intrusion monitoring: the monitoring system compares the current image with the background image, divides the monitoring area into a plurality of sub-areas, and sends an 'intrusion alarm' signal to the master control computer if an object appears in the sub-areas; in step S5, the monitoring is ended. The invention greatly improves the monitoring accuracy, reduces the working intensity of workers, and has low cost and easy realization.
Description
Technical Field
The invention relates to a pedestrian monitoring method, in particular to a passenger trailing monitoring method based on foreground object segmentation.
Background
Pedestrian monitoring system is more common in crowd intensive places such as station, and current self-service passageway is equipped with two doors and keeps apart the passenger, requires can only one passenger to pass through the passageway at every turn, and after this passenger accomplished to inspect and left the passageway, perhaps next passenger gets into the passageway, also does not allow adult to take child to get into the passageway, but in the in-service use has someone to get into the passageway with normal passenger with multiple mode (including upright walking, squat, crawling or bow to move ahead etc.), also has adult to take child to get into the condition in passageway. In response to these problems, the existing pedestrian monitoring methods include the following:
the first is that the light curtain monitors the object, one side of the light curtain is equipped with a plurality of infrared transmitting tubes at equal intervals, the other side is equipped with the same number of infrared receiving tubes which are arranged in the same way, each infrared transmitting tube is equipped with a corresponding infrared receiving tube correspondingly, and is installed on the same straight line. When no barrier exists between the infrared transmitting tube and the infrared receiving tube on the same straight line, the modulation signal (optical signal) sent by the infrared transmitting tube can smoothly reach the infrared receiving tube. After the infrared receiving tube receives the modulation signal, the corresponding internal circuit outputs low level, and under the condition of an obstacle, the modulation signal (optical signal) sent by the infrared transmitting tube can not smoothly reach the infrared receiving tube, at this moment, the infrared receiving tube can not receive the modulation signal, and the output of the corresponding internal circuit is high level. When no object passes through the light curtain, the modulated signals (light signals) sent by all the infrared transmitting tubes can smoothly reach the corresponding infrared receiving tube on the other side, so that all the internal circuits output low level. Thus, the presence or absence of an object can be monitored by analyzing the state of the internal circuit. The light curtain sensor technology has the defects that whether a person enters a channel or not can be accurately monitored, two persons enter the channel side by side or two thin and small persons follow the channel, one person or two persons cannot be distinguished, missing report or false report can be caused, and false alarm can be caused if dust obstructs emission and acceptance of infrared rays. It is also possible for a large piece of luggage carried by a passenger to be misjudged as a person.
The second is depth image analysis, which divides the depth image into layers, analyzes the connected domain of each layer and judges whether two-person characteristics appear; the lower floor has two independent connected domains, whether there is the motion trail through the difference monitoring of interframe, then can tentatively judge for two people, and then through the multilayer joint analysis, for example, if there is three layers to satisfy above condition, then send "follow alarm" signal to the master control computer. And if no monitoring signal exists, entering intrusion monitoring. The depth image analysis technology has the defects that a front person carries a mountain bag and two persons enter the depth image in a way of being embraced from front to back, the upper layer of a depth image is similar to a circle and is difficult to distinguish, and false alarm or missing alarm can be caused; the later people squat to crawl and the foremen drag large luggage to be difficult to distinguish, and false alarm or missing report can be caused.
The third is an object monitoring algorithm, the pedestrians in the channel are monitored by using a conventional object monitoring algorithm, whether trailing conditions occur or not is determined according to the number of the monitored pedestrians, a single person is monitored, the pedestrians normally pass through the channel, and two or more than two persons are monitored to give an alarm. The target monitoring algorithm has the defects that double-person imaging is easy to overlap due to the problem of the field angle, and the rectangular frame result output by the target monitoring algorithm has great deviation and is difficult to avoid false detection and missing detection.
Disclosure of Invention
The invention aims to solve the technical problem that the method for monitoring the trailing of the passenger based on the foreground object segmentation is provided aiming at the defects of the prior art, so that the method is used for sending out an accurate alarm when a plurality of people enter a channel, and has the effects of short monitoring time, wide monitoring range, accurate monitoring of the trailing person, timely alarm and the like.
In order to solve the technical problems, the invention adopts the following technical scheme.
A passenger trailing monitoring method based on foreground object segmentation is realized based on a monitoring system and a master control computer, and comprises the following steps: step S1, the monitoring system collects images; step S2, mode selection: the monitoring system selects a mode according to a control instruction sent by the master control computer, if the monitoring system enters a trailing monitoring mode, the step S3 is executed, and if the monitoring system enters an intrusion monitoring mode, the step S4 is executed; step S3, trailing monitoring: the monitoring system utilizes a deep neural network model to carry out pixel-level segmentation on the pedestrian image aiming at the collected image, then multi-frame comprehensive judgment is carried out on whether more than one person exists in the segmented image, and if yes, a trailing alarm signal is sent to the master control computer; step S4, intrusion monitoring: the monitoring system compares the current image with the background image, divides the monitoring area into a plurality of sub-areas, and sends an 'intrusion alarm' signal to the master control computer if an object appears in the sub-areas; step S5, end monitoring: and if the monitoring system receives the 'monitoring ending' instruction sent by the master control computer, the monitoring is stopped, and if the 'monitoring ending' instruction is not received, the monitoring is automatically ended after the preset time is continued.
Preferably, in step S1, the image captured by the monitoring system is a color video image.
Preferably, the step S1 is followed by a waiting monitoring step: and the monitoring system determines whether an object enters according to the change of the background image and waits for the master control computer to send a control instruction.
Preferably, in step S2, when someone swipes a card to enter a monitoring area, the main control computer sends a control instruction to the monitoring system.
Preferably, in the step S3, the deep neural network model is obtained by the monitoring system through offline training in advance.
Preferably, in step S4, the monitoring system divides the monitoring area into 5 sub-areas.
Preferably, in step S5, if the "end monitoring" command is not received, the monitoring system automatically ends monitoring after 5 seconds.
Preferably, in step S5, if the monitoring system does not generate the "tail alarm" signal and the "intrusion alarm" signal, the "normal pass" is displayed.
Preferably, the method further comprises the step S6 of resetting: and when the pedestrian leaves the monitoring area, the outlet is closed, the monitoring system receives a 'clear alarm' instruction sent by the master control computer, and updates the background image.
Preferably, the monitoring area is a pedestrian pathway.
Compared with the prior art, the passenger trailing monitoring method based on foreground object segmentation has the advantages that the monitoring system can accurately and effectively find the condition that multiple persons enter the channel and give an alarm through trailing monitoring and intrusion monitoring, simultaneously stores the trailing-prevention video, effectively reduces false alarm and missing alarm, greatly improves the monitoring accuracy, reduces the working intensity of workers, and is low in cost and easy to implement.
Drawings
FIG. 1 is a flow chart of a passenger trailing monitoring method based on foreground object segmentation according to the present invention.
Detailed Description
The invention is described in more detail below with reference to the figures and examples.
The invention discloses a passenger trailing monitoring method based on foreground object segmentation, which is realized based on a monitoring system and a master control computer and comprises the following steps:
step S1, the monitoring system collects images;
step S2, mode selection: the monitoring system selects a mode according to a control instruction sent by the master control computer, if the monitoring system enters a trailing monitoring mode, the step S3 is executed, and if the monitoring system enters an intrusion monitoring mode, the step S4 is executed;
step S3, trailing monitoring: the monitoring system utilizes a deep neural network model to carry out pixel-level segmentation on the pedestrian image aiming at the collected image, then multi-frame comprehensive judgment is carried out on whether more than one person exists in the segmented image, and if yes, a trailing alarm signal is sent to the master control computer;
step S4, intrusion monitoring: the monitoring system compares the current image with the background image, divides the monitoring area into a plurality of sub-areas, and sends an 'intrusion alarm' signal to the master control computer if an object appears in the sub-areas;
step S5, end monitoring: and if the monitoring system receives the 'monitoring ending' instruction sent by the master control computer, the monitoring is stopped, and if the 'monitoring ending' instruction is not received, the monitoring is automatically ended after the preset time is continued.
According to the method, the monitoring system can accurately and effectively find the condition that multiple people enter the channel and give an alarm through trailing monitoring and intrusion monitoring, and meanwhile, the anti-trailing video is stored.
Regarding the captured image, in step S1, the image captured by the monitoring system is a color video image.
In this embodiment, after the step S1, the method further includes a waiting monitoring step: and the monitoring system determines whether an object enters according to the change of the background image and waits for the master control computer to send a control instruction.
Further, in step S2, when someone swipes a card to enter the monitoring area, the main control computer sends a control instruction to the monitoring system.
Preferably, in step S3, the deep neural network model is obtained by the monitoring system through offline training in advance.
In step S4 of the method of the present invention, the monitoring system divides the monitoring area into 5 sub-areas.
In a preferred embodiment, in step S5, if the "end monitoring" command is not received, the monitoring system automatically ends monitoring after 5 seconds.
In the case of normal traffic, if the monitoring system does not generate the "follow alarm" signal and the "intrusion alarm" signal in step S5, "normal pass" is displayed.
The present embodiment further includes step S6, resetting: and when the pedestrian leaves the monitoring area, the outlet is closed, the monitoring system receives a 'clear alarm' instruction sent by the master control computer, and updates the background image.
In this embodiment, the monitoring area is a pedestrian passage.
The invention discloses a passenger trailing monitoring method based on foreground object segmentation, as shown in fig. 1, the execution process in the practical application process can refer to the following embodiments:
step 1, collecting color video images.
Step 2, waiting for monitoring: and determining whether an object enters or not through the background difference, and waiting for various command signals sent by the channel master control computer.
And step 3, starting a monitoring mode: when someone swipes the card to enter the channel, the master control computer sends a 'start monitoring' instruction to enter the trailing mode monitoring, and if the 'start monitoring' instruction is not received, the master control computer enters the break-in mode monitoring.
Step 4, follow-up monitoring: the input image is subjected to pixel-level segmentation of pedestrians by a deep neural network model which is trained in advance off-line, then whether more than one person exists in the segmented image is comprehensively judged by multiple frames, and if the existence of multiple persons is found, a trailing alarm signal is sent to a master control computer.
Step 5, intrusion monitoring: comparing the current image with the background, dividing the passage area into 5 small areas, and if an object is found at the exit, sending an 'intrusion alarm' signal to a master control computer.
And 6, finishing monitoring: if receiving the instruction of finishing monitoring from the main control computer, stopping monitoring, if not, finishing monitoring automatically after 5 seconds. If no alarm is found, a 'normal pass' is displayed.
And 7, resetting: after people go out, the exit door is closed, and the 'clear alarm' instruction sent by the master control computer is received, so that the background is updated.
The invention discloses a passenger trailing monitoring method based on foreground object segmentation, which is characterized in that a Mask RCNN (context-based neural network) based example segmentation algorithm for deep learning can realize accurate pixel-level class prediction of foreground objects, the Mask RCNN is divided into three parts, the first part is a backbone network for feature extraction, the second part is a head structure for boundary frame identification (classification and regression), and the third part is Mask prediction for pixel-level segmentation of each foreground object. Considering that only people are processed and knapsack luggage is a background class in the channel anti-tailing scene, pixel-level screening can be performed on the complex scene of the anti-tailing channel by adopting an example segmentation algorithm, and if two people are tightly held together, the people can normally detect whether the front people stand normally and the back people squat down and bend down; particularly for infants and children, the pixel-level segmentation can be used for accurately distinguishing the infants and the infants from luggage according to the detail difference, so that the false detection is reduced. Aiming at the deployment of embedded equipment, a Mask RCNN backbone network is replaced by a Mobilenet network structure specially designed for a mobile terminal, so that the real-time anti-following monitoring of an anti-following scene is realized. Compared with the prior art, the invention can monitor the condition that a person (one person or a plurality of persons) enters the passage along with the passenger in various ways, including the condition that an adult enters the passage along with children, can identify the luggage articles entering the passage along with the passenger, and can not judge the luggage articles as persons by mistake. The invention prompts accurate alarm under the condition that a plurality of people enter the channel, has short monitoring time, covers the whole channel in the monitoring range, can be found by the following people when entering the channel, and can alarm in time.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents or improvements made within the technical scope of the present invention should be included in the scope of the present invention.
Claims (2)
1. A passenger trailing monitoring method based on foreground object segmentation is characterized in that the method is realized based on a monitoring system and a master control computer, and the method comprises the following steps:
step S1, the monitoring system collects images;
step S2, mode selection: the monitoring system selects a mode according to a control instruction sent by the master control computer, if the monitoring system enters a trailing monitoring mode, the step S3 is executed, and if the monitoring system enters an intrusion monitoring mode, the step S4 is executed;
step S3, trailing monitoring: the monitoring system utilizes a deep neural network model to carry out pixel-level segmentation on the pedestrian image aiming at the collected image, then multi-frame comprehensive judgment is carried out on whether more than one person exists in the segmented image, and if yes, a trailing alarm signal is sent to the master control computer;
step S4, intrusion monitoring: the monitoring system compares the current image with the background image, divides the monitoring area into a plurality of sub-areas, and sends an 'intrusion alarm' signal to the master control computer if an object appears in the sub-areas;
step S5, end monitoring: if the monitoring system receives a 'monitoring ending' instruction sent by the master control computer, stopping monitoring, and if the 'monitoring ending' instruction is not received, automatically ending monitoring after the preset time is continued;
in step S3, the deep neural network model is obtained by the monitoring system through offline training in advance;
in step S4, the monitoring system divides the monitoring area into 5 sub-areas;
in step S5, if the command of "end monitoring" is not received, the monitoring system automatically ends monitoring after 5 seconds;
in step S5, if the monitoring system does not generate the "follow alarm" signal and the "intrusion alarm" signal, it displays "normal pass";
further comprising a step S6 of resetting: when the pedestrian leaves the monitoring area, the outlet is closed, the monitoring system receives a 'clear alarm' instruction sent by a master control computer, and updates a background image;
the monitoring area is a pedestrian passage;
in step S1, the image collected by the monitoring system is a color video image;
the step S1 is followed by a waiting monitoring step: and the monitoring system determines whether an object enters according to the change of the background image and waits for the master control computer to send a control instruction.
2. The passenger tailgating monitoring method based on foreground object segmentation as claimed in claim 1, wherein in step S2, when someone swipes a card to enter the monitoring area, the main control computer sends a control command to the monitoring system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810074904.7A CN108280952B (en) | 2018-01-25 | 2018-01-25 | Passenger trailing monitoring method based on foreground object segmentation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810074904.7A CN108280952B (en) | 2018-01-25 | 2018-01-25 | Passenger trailing monitoring method based on foreground object segmentation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108280952A CN108280952A (en) | 2018-07-13 |
| CN108280952B true CN108280952B (en) | 2020-03-27 |
Family
ID=62805158
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810074904.7A Active CN108280952B (en) | 2018-01-25 | 2018-01-25 | Passenger trailing monitoring method based on foreground object segmentation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108280952B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109118519A (en) * | 2018-07-26 | 2019-01-01 | 北京纵目安驰智能科技有限公司 | Target Re-ID method, system, terminal and the storage medium of Case-based Reasoning segmentation |
| CN110930427B (en) * | 2018-09-20 | 2022-05-24 | 银河水滴科技（北京）有限公司 | Image segmentation method, device and storage medium based on semantic contour information |
| CN109635740B (en) * | 2018-12-13 | 2020-07-03 | 深圳美图创新科技有限公司 | Video target detection method and device and image processing equipment |
| CN109657608B (en) * | 2018-12-17 | 2023-10-13 | 中通服公众信息产业股份有限公司 | Trailing analysis method based on face recognition technology |
| CN109977796A (en) * | 2019-03-06 | 2019-07-05 | 新华三技术有限公司 | Trail current detection method and device |
| CN110930568A (en) * | 2019-12-05 | 2020-03-27 | 江苏中云智慧数据科技有限公司 | Video anti-trailing system and method |
| CN111144231B (en) * | 2019-12-09 | 2022-04-15 | 深圳市鸿逸达科技有限公司 | Self-service channel anti-trailing detection method and system based on depth image |
| CN112037403A (en) * | 2020-07-22 | 2020-12-04 | 四川科达乐气象科技有限公司 | Personnel safety monitoring system and method based on embedded gateway |
| CN112597928B (en) * | 2020-12-28 | 2024-05-14 | 深圳市捷顺科技实业股份有限公司 | Event detection method and related device |
| CN118038340B (en) * | 2024-04-15 | 2024-10-11 | 盛视科技股份有限公司 | Anti-trailing detection system based on video image |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012012555A1 (en) * | 2010-07-20 | 2012-01-26 | SET Corporation | Methods and systems for audience digital monitoring |
| DE102011011929A1 (en) * | 2011-02-18 | 2012-08-23 | Hella Kgaa Hueck & Co. | Method for detecting target objects in a surveillance area |
| CN105141885B (en) * | 2014-05-26 | 2018-04-20 | 杭州海康威视数字技术股份有限公司 | Carry out the method and device of video monitoring |
| CN105139425B (en) * | 2015-08-28 | 2018-12-07 | 浙江宇视科技有限公司 | A kind of demographic method and device |
| CN106961576A (en) * | 2017-03-10 | 2017-07-18 | 盛视科技股份有限公司 | video anti-tailing method, device and system |
| CN107610148B (en) * | 2017-09-19 | 2020-07-28 | 电子科技大学 | A foreground segmentation method based on binocular stereo vision system |
| CN107563349A (en) * | 2017-09-21 | 2018-01-09 | 电子科技大学 | A kind of Population size estimation method based on VGGNet |
-
2018
- 2018-01-25 CN CN201810074904.7A patent/CN108280952B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108280952A (en) | 2018-07-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108280952B (en) | Passenger trailing monitoring method based on foreground object segmentation | |
| CN109686109B (en) | Parking lot safety monitoring management system and method based on artificial intelligence | |
| CN111144247B (en) | A method for retrograde detection of escalator passengers based on deep learning | |
| CN110002302B (en) | Elevator door opening and closing detection system and method based on deep learning | |
| CN105744232B (en) | A kind of method of the transmission line of electricity video external force damage prevention of Behavior-based control analytical technology | |
| CN101577812B (en) | Method and system for post monitoring | |
| Dick et al. | Issues in automated visual surveillance | |
| CN101334924B (en) | Fire detection system and fire detection method thereof | |
| US8301577B2 (en) | Intelligent monitoring system for establishing reliable background information in a complex image environment | |
| CN104847211B (en) | Auxiliary system for platform safety door and car door section safety and its realization method | |
| KR20200071799A (en) | object recognition and counting method using deep learning artificial intelligence technology | |
| CN112669497A (en) | Pedestrian passageway perception system and method based on stereoscopic vision technology | |
| KR20130097868A (en) | Intelligent parking management method and system based on camera | |
| JPH08123935A (en) | Method and device for counting moving object by direction | |
| US20220351515A1 (en) | Method for perceiving event tagging-based situation and system for same | |
| KR101656642B1 (en) | Group action analysis method by image | |
| KR102113489B1 (en) | Action Detection System and Method Using Deep Learning with Camera Input Image | |
| KR100967456B1 (en) | Railroad crossing obstacle image detection device and control method | |
| CN110211159A (en) | A kind of aircraft position detection system and method based on image/video processing technique | |
| CN108109304A (en) | A kind of subway platform invader identifying system and recognition methods | |
| KR101736431B1 (en) | System and Method for Smart Traffic Monitoring Using Multiple Image | |
| KR20100051775A (en) | Real-time image analysis using the multi-object tracking and analyzing vehicle information and data management system and how to handle | |
| JP7597533B2 (en) | Station platform monitoring system | |
| KR100699596B1 (en) | Automatic detection image processing alarm device of railway crossing | |
| CN114802351A (en) | Vehicle control method, device and system and computer equipment |
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 | ||
| CP02 | Change in the address of a patent holder | ||
| CP02 | Change in the address of a patent holder |
Address after: 1601 and 1605, East Block 2, Tian'an innovation Plaza, chegongmiao, Futian District, Shenzhen City, Guangdong Province Patentee after: Maxvision Technology Corp. Address before: 518000 Guangdong city of Shenzhen province Futian District Shatou street Che Kung Temple Tairan six road Tairan Cangsong building 511, South Tower 512, 513, 515, 516, 517, 518 (only office) Patentee before: Maxvision Technology Corp. |