[go: up one dir, main page]

CN108111808B - Drowning detection method based on real-time video image analysis - Google Patents

Drowning detection method based on real-time video image analysis Download PDF

Info

Publication number
CN108111808B
CN108111808B CN201711231670.4A CN201711231670A CN108111808B CN 108111808 B CN108111808 B CN 108111808B CN 201711231670 A CN201711231670 A CN 201711231670A CN 108111808 B CN108111808 B CN 108111808B
Authority
CN
China
Prior art keywords
drowning
detection
video
drowning detection
alarm
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
Application number
CN201711231670.4A
Other languages
Chinese (zh)
Other versions
CN108111808A (en
Inventor
贠周会
王旭
叶超
吴斌
应艳丽
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangxi Hongdu Aviation Industry Group Co Ltd
Original Assignee
Jiangxi Hongdu Aviation Industry Group Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangxi Hongdu Aviation Industry Group Co Ltd filed Critical Jiangxi Hongdu Aviation Industry Group Co Ltd
Priority to CN201711231670.4A priority Critical patent/CN108111808B/en
Publication of CN108111808A publication Critical patent/CN108111808A/en
Application granted granted Critical
Publication of CN108111808B publication Critical patent/CN108111808B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N7/00Television systems
    • H04N7/18Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/11Region-based segmentation
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00Scenes; Scene-specific elements
    • G06V20/50Context or environment of the image
    • G06V20/52Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • GPHYSICS
    • G08SIGNALLING
    • G08BSIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
    • G08B21/00Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
    • G08B21/02Alarms for ensuring the safety of persons
    • G08B21/08Alarms for ensuring the safety of persons responsive to the presence of persons in a body of water, e.g. a swimming pool; responsive to an abnormal condition of a body of water
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00Details of television systems
    • H04N5/76Television signal recording
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10016Video; Image sequence
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30232Surveillance

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Signal Processing (AREA)
  • Business, Economics & Management (AREA)
  • Emergency Management (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Image Analysis (AREA)
  • Emergency Alarm Devices (AREA)

Abstract

A drowning detection method based on real-time video image analysis is characterized in that a drowning detection device based on real-time video image analysis is adopted to carry out drowning detection, the drowning detection device based on real-time video image analysis comprises an IP camera, an alarm, a switch and a drowning detection server, the IP camera and the alarm are installed in a detection area, the IP camera is connected with the switch, the switch is connected with the drowning detection server, the drowning detection server is connected with the alarm, a drowning detection module and a manual one-key confirmation module are arranged in the drowning detection server, when the drowning detection module judges that a drowning event occurs, an operator confirms the drowning event through the manual one-key confirmation module, corresponding measures are taken according to confirmation results, a large amount of manpower and material resources are effectively saved, the alarm accuracy is improved, and the false alarm rate.

Description

Drowning detection method based on real-time video image analysis
Technical Field
The invention relates to the technical field of video analysis, in particular to a drowning detection method based on real-time video image analysis.
Background
With the rapid development of national economy and basic science, people begin to have higher standards and requirements for public safety guarantee in daily life. People hope to ensure the personal safety and property safety of the people by enhancing the safety infrastructure construction, and even hope that the early warning or the alarm can be given by high-tech means when the danger is about to occur or just occurs, so as to reduce the personal injury and the property loss. If more lakes exist in a city, the duration of hot weather is long in summer, and more people are gathered around the lakes to enjoy the cool in the summer, and people still fall into the water to cause life danger although warning boards prohibiting swimming are erected around the lakes. The number of deaths from drowning is over tens of thousands of people each year, with most deaths being children under 12 years of age, with little success despite measures such as real-time video monitoring, 24-hour emergency rescue, etc.
The prior patent disclosed in the aspect of falling water detection is mainly directed to falling water detection of objects (such as containers and the like), and most of the methods are that an induction trigger is arranged on the objects, and a strategy is triggered once falling water occurs; the number of patents aiming at the drowning detection of human bodies and adopting video analysis is few. In addition, through research and analysis, in the fields of digital image processing technology and computer vision technology, the application cases successful in this aspect are few and few, and the main reasons are: the false alarm rate cannot be reduced, the video analysis method is adopted to carry out drowning detection, and the intrusion principle in video analysis is actually applied, namely whether pixel lumps (namely intrusion lumps) with changed gray levels exist in a marking area is detected, if so, the pixel lumps are understood as intrusion, after the marking area is finished on a natural water area, ripples generated by water and sunlight irradiation are caused by blowing of wind, so that the pixel lumps with changed gray levels are easily generated in the marking area by the glistening shark, and an alarm is generated after the number of the pixels occupied by the lumps is greater than a set threshold value.
Disclosure of Invention
The present invention provides a method for detecting drowning based on real-time video image analysis, so as to solve the above-mentioned drawbacks in the background art.
The technical problem solved by the invention is realized by adopting the following technical scheme:
the utility model provides a detection method falls into water based on real-time video image analysis, adopt the detection device that falls into water based on real-time video image analysis to implement and fall into water and detect, the detection device that falls into water based on real-time video image analysis includes IP camera, alarm, switch and the detection server that falls into water, IP camera and alarm are installed in the detection zone, and IP camera is connected with the switch, the switch is connected with the detection server that falls into water, the detection server that falls into water is connected with the alarm, and be provided with in the detection server that falls into water and fall into water detection module and artifical one-touch confirmation module, the detection step that specifically falls into water is as follows:
1) the drowning detection server reads and decodes the video stream collected by the IP camera in real time through the switch, and the decoded video data is temporarily stored in a cache;
2) the drowning detection module adopts a VIBE background modeling algorithm to perform background generation and background updating, performs VIBE algorithm calculation on the video data stored in the cache, and stores the calculation result into the cache, wherein the result is a first background and an updated background;
3) performing polygonal closed scribing on an actual region to be detected in a video by utilizing an OpenGL algorithm, wherein the scribed region is a detection region, and a scribing closed frame of the detection region is always displayed on the video; meanwhile, carrying out polygonal closed scribing in the area above the horizontal plane in the detection area to determine the area above the horizontal plane;
4) when the lumps appear in the detection area defined in the step 3), the water falling detection module calculates the number of the lumps and the sizes of the lumps by utilizing an interframe difference method and a morphological algorithm;
5) when the calculation of the lumps in the step 4) meets the following condition reservation: firstly, the size of the lump is larger than a set threshold value; secondly, the geometric centroid of the block mass enters a detection area, meanwhile, in a marking area above the horizontal plane of the detection area, the block mass meeting the two conditions is subjected to target tracking, the block mass with a vertical downward displacement component is reserved, when the block mass enters the horizontal plane of the marking area, transverse ripples are generated immediately, one-time falling water detection is marked, and at the moment, a falling water detection server immediately gives an early warning; when one-time drowning detection occurs, the drowning detection module extracts the video frame 30 seconds before the early warning moment from the cache region, repeatedly and quickly plays the video frame on the display interface of the drowning detection server and immediately stores the video frame in a designated folder in the drowning detection server;
6) an operator checks short videos 30 seconds before the early warning moment played on a display interface of the drowning detection server to manually determine for the second time, if the drowning event is determined, the manual one-key type confirmation module confirms that a confirmation key is pressed, the drowning detection server keeps the short videos and continuously stores the short videos in a designated folder, triggers an alarm signal to prompt an alarm installed in a detection area to give an alarm, and simultaneously informs a field ready rescue team to rapidly go out of the alarm for rescue; if the video is not the drowning event, a negative key is pressed, the drowning detection server deletes the video and cancels early warning.
In the present invention, in step 1), the drop-in-water detection server decodes the video stream using an FFmpeg video decoder.
In the invention, in the step 5), when the drowning detection server gives an early warning, the marking-off closed frame turns red, and an early warning prompt tone is sent out.
In the invention, in step 5), the drowning detection module uses an FFmpeg video encoder to perform video encoding on a video frame 30 seconds before the early warning moment.
Has the advantages that:
1) the invention effectively reduces the false alarm rate, improves the alarm accuracy and further improves the working efficiency;
2) according to the invention, a large amount of manpower is not needed for on-site patrol, a worker is not needed to check the monitoring video within 24 hours, only one video monitoring worker needs to check the short video in time after hearing the early warning, and one-click operation is carried out after judgment is made, so that a large amount of manpower and material resources are effectively saved.
Drawings
FIG. 1 is a flow chart illustrating a preferred embodiment of the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further explained below by combining the specific drawings.
Referring to fig. 1, a method for detecting falling into water based on real-time video image analysis, which implements the detection of falling into water by using a device for detecting falling into water based on real-time video image analysis, the device for detecting falling into water based on real-time video image analysis includes an IP camera, an alarm, a switch and a server for detecting falling into water, the IP camera and the alarm are installed in a detection area, the IP camera is connected to the switch, the switch is connected to the server for detecting falling into water, the server for detecting falling into water is connected to the alarm, and a module for detecting falling into water and a manual one-key confirmation module are installed in the server for detecting falling into water, when the module for detecting falling into water determines that an event occurs, an operator confirms the event by the manual one-key confirmation module:
1) the drowning detection server reads the video stream collected by the IP camera in real time through the switch, decodes the video stream by using an FFmpeg video decoder, and temporarily stores the decoded video data in a cache;
2) the water-falling detection module adopts a VIBE background modeling algorithm to generate and update a background, performs VIBE algorithm calculation on the video data stored in the cache, and stores a calculation result into the cache, wherein the result is a first background and an updated background;
3) performing polygonal closed scribing on an actual region to be detected in a video by utilizing an OpenGL algorithm, wherein the scribed region is a detection region, and a scribing closed frame of the detection region is always displayed on the video; meanwhile, polygonal closed drawing is carried out in the area above the horizontal plane in the detection area to determine the area above the horizontal plane, and the closed line of the area is not required to be displayed on a video;
4) when the lumps appear in the detection area defined in the step 3), the water falling detection module calculates the number of the lumps and the sizes of the lumps by utilizing an interframe difference method and a morphological algorithm;
5) when the calculation of the lumps in the step 4) meets the following condition reservation: firstly, the size of the lump is larger than a set threshold value; secondly, the geometric centroid of the block mass enters a detection area, meanwhile, in a marking area above the horizontal plane of the detection area, the block mass meeting the two conditions is subjected to target tracking, the block mass with a vertical downward displacement component is reserved, when the block mass enters the horizontal plane of the marking area, transverse ripples (gray level difference displayed in a gray level image is transversely and continuously enlarged) are immediately generated, one-time falling water detection is marked, at the moment, a falling water detection server immediately gives an early warning, a marking closed frame turns red, and an early warning prompt tone is sent out; when one-time drowning detection occurs, the drowning detection module extracts a video frame 30 seconds before the early warning moment from the buffer area, uses an FFmpeg video encoder to perform video encoding, repeatedly and quickly plays the video frame on a display interface of a drowning detection server and immediately stores the video frame in a designated folder in the drowning detection server so as to empty the buffer area to reserve a space for subsequent short video storage;
6) an operator checks short videos 30 seconds before the early warning moment played on a display interface of the drowning detection server to manually determine for the second time, if the drowning event is determined, the manual one-key type confirmation module confirms that a confirmation key is pressed, the drowning detection server keeps the short videos and continuously stores the short videos in a designated folder, triggers an alarm signal to prompt an alarm installed in a detection area to give an alarm, and simultaneously informs a field ready rescue team to rapidly go out of the alarm for rescue; if the video is not the drowning event, a negative key is pressed, the drowning detection server deletes the video and cancels early warning.
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (4)

1. The drowning detection method based on real-time video image analysis is characterized in that the drowning detection is implemented by adopting a drowning detection device based on real-time video image analysis, the drowning detection device based on real-time video image analysis comprises an IP camera, an alarm, a switch and a drowning detection server, the IP camera and the alarm are installed in a detection area, the IP camera is connected with the switch, the switch is connected with the drowning detection server, the drowning detection server is connected with the alarm, a drowning detection module and a manual one-key confirmation module are arranged in the drowning detection server, and the drowning detection steps are as follows:
1) the drowning detection server reads and decodes the video stream collected by the IP camera in real time through the switch, and the decoded video data is temporarily stored in a cache;
2) the drowning detection module adopts a VIBE background modeling algorithm to perform background generation and background updating, performs VIBE algorithm calculation on the video data stored in the cache, and stores the calculation result into the cache, wherein the result is a first background and an updated background;
3) performing polygonal closed scribing on an actual region to be detected in a video by utilizing an OpenGL algorithm, wherein the scribed region is a detection region, and a scribing closed frame of the detection region is always displayed on the video; meanwhile, carrying out polygonal closed scribing in the area above the horizontal plane in the detection area to determine the area above the horizontal plane;
4) when the lumps appear in the detection area defined in the step 3), the water falling detection module calculates the number of the lumps and the sizes of the lumps by utilizing an interframe difference method and a morphological algorithm;
5) when the calculation of the lumps in the step 4) meets the following condition reservation: firstly, the size of the lump is larger than a set threshold value; secondly, the geometric centroid of the block mass enters a detection area, meanwhile, in a marking area above the horizontal plane of the detection area, the block mass meeting the two conditions is subjected to target tracking, the block mass with a vertical downward displacement component is reserved, when the block mass enters the horizontal plane of the marking area, transverse ripples are generated immediately, one-time falling water detection is marked, and at the moment, a falling water detection server immediately gives an early warning; when one-time drowning detection occurs, the drowning detection module extracts the video frame 30 seconds before the early warning moment from the cache region, repeatedly and quickly plays the video frame on the display interface of the drowning detection server and immediately stores the video frame in a designated folder in the drowning detection server;
6) an operator checks short videos 30 seconds before the early warning moment played on a display interface of the drowning detection server to manually determine for the second time, if the drowning event is determined, the manual one-key type confirmation module confirms that a confirmation key is pressed, the drowning detection server keeps the short videos and continuously stores the short videos in a designated folder, triggers an alarm signal to prompt an alarm installed in a detection area to give an alarm, and simultaneously informs a field ready rescue team to rapidly go out of the alarm for rescue; if the video is not the drowning event, a negative key is pressed, the drowning detection server deletes the video and cancels early warning.
2. The method of claim 1, wherein the video stream is decoded by the drowning detection server using a FFmpeg video decoder in step 1).
3. The drowning detection method based on real-time video image analysis according to claim 1, characterized in that in step 5), when the drowning detection server gives an early warning, the closed frame is marked to turn red, and an early warning prompt tone is given.
4. The method of claim 1, wherein in step 5), the video coding module uses a FFmpeg video encoder to perform video coding on the video frame 30 seconds before the early warning time.
CN201711231670.4A 2017-11-30 2017-11-30 Drowning detection method based on real-time video image analysis Active CN108111808B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711231670.4A CN108111808B (en) 2017-11-30 2017-11-30 Drowning detection method based on real-time video image analysis

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711231670.4A CN108111808B (en) 2017-11-30 2017-11-30 Drowning detection method based on real-time video image analysis

Publications (2)

Publication Number Publication Date
CN108111808A CN108111808A (en) 2018-06-01
CN108111808B true CN108111808B (en) 2020-04-28

Family

ID=62208800

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711231670.4A Active CN108111808B (en) 2017-11-30 2017-11-30 Drowning detection method based on real-time video image analysis

Country Status (1)

Country Link
CN (1) CN108111808B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102018215125A1 (en) * 2018-09-06 2020-03-12 Robert Bosch Gmbh Monitoring device and method for man overboard monitoring
CN110148283A (en) * 2019-05-16 2019-08-20 安徽天帆智能科技有限责任公司 It is a kind of to fall water monitoring system in real time based on convolutional neural networks
CN111028480A (en) * 2019-12-06 2020-04-17 江西洪都航空工业集团有限责任公司 Drowning detection and alarm system
CN111310556A (en) * 2019-12-20 2020-06-19 山东汇佳软件科技股份有限公司 Drowning prevention safety supervision system based on primary and middle school student area and monitoring method thereof

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102693606A (en) * 2011-03-24 2012-09-26 中国人民解放军国防科学技术大学 Shooting network-based method and system for warning of drowning prevention in swimming pool
WO2012145800A1 (en) * 2011-04-29 2012-11-01 Preservation Solutions Pty Ltd Monitoring the water safety of at least one person in a body of water
CN103578226A (en) * 2012-07-31 2014-02-12 青岛联合创新技术服务平台有限公司 Video capturing alarm device used for swimming pool and warning method thereof
EP2722636A1 (en) * 2012-10-22 2014-04-23 The Boeing Company Water area management system
CN104301662A (en) * 2013-07-20 2015-01-21 郑州学生宝电子科技有限公司 Monitoring system for preventing students from drowning
CN104469317A (en) * 2014-12-18 2015-03-25 天津市亚安科技股份有限公司 Video monitoring system for pipeline safety early warning
CN104537892A (en) * 2015-01-14 2015-04-22 山东中创软件工程股份有限公司 Information processing method and system
CN106951838A (en) * 2017-03-07 2017-07-14 四川省建筑设计研究院 A kind of overboard warning system based on image water wave
WO2017130187A1 (en) * 2016-01-26 2017-08-03 Coral Detection Systems Ltd. Methods and systems for drowning detection

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102693606A (en) * 2011-03-24 2012-09-26 中国人民解放军国防科学技术大学 Shooting network-based method and system for warning of drowning prevention in swimming pool
WO2012145800A1 (en) * 2011-04-29 2012-11-01 Preservation Solutions Pty Ltd Monitoring the water safety of at least one person in a body of water
CN103578226A (en) * 2012-07-31 2014-02-12 青岛联合创新技术服务平台有限公司 Video capturing alarm device used for swimming pool and warning method thereof
EP2722636A1 (en) * 2012-10-22 2014-04-23 The Boeing Company Water area management system
CN104301662A (en) * 2013-07-20 2015-01-21 郑州学生宝电子科技有限公司 Monitoring system for preventing students from drowning
CN104469317A (en) * 2014-12-18 2015-03-25 天津市亚安科技股份有限公司 Video monitoring system for pipeline safety early warning
CN104537892A (en) * 2015-01-14 2015-04-22 山东中创软件工程股份有限公司 Information processing method and system
WO2017130187A1 (en) * 2016-01-26 2017-08-03 Coral Detection Systems Ltd. Methods and systems for drowning detection
CN106951838A (en) * 2017-03-07 2017-07-14 四川省建筑设计研究院 A kind of overboard warning system based on image water wave

Also Published As

Publication number Publication date
CN108111808A (en) 2018-06-01

Similar Documents

Publication Publication Date Title
CN108111808B (en) Drowning detection method based on real-time video image analysis
CN102693606B (en) A kind of swimming pool anti-drowning method for early warning based on shooting network and system
CN102013147B (en) High voltage power transmission tower intelligent anti-theft method for supervising and device
CN103475882B (en) Surveillance video encoding and recognizing method and surveillance video encoding and recognizing system
CN103714325B (en) Left object and lost object real-time detection method based on embedded system
CN111967393A (en) Helmet wearing detection method based on improved YOLOv4
CN106023076B (en) Panorama stitching method and method for detecting defect state of guardrail of high-speed railway
CN110070033A (en) Safety cap wearing state detection method in a kind of power domain dangerous work region
CN117218801A (en) Urban flood disaster monitoring and early warning method and device
CN103093192B (en) The recognition methods that high voltage transmission line is waved
CN103442209A (en) Video monitoring method of electric transmission line
CN103150856A (en) Fire flame video monitoring and early warning system and fire flame detection method
CN113936248B (en) A hazard early warning method for beach personnel based on image recognition
CN104165668B (en) A kind of water conservancy monitoring method and its system of automatic early-warning water conservancy dangerous situation
CN205281786U (en) Alarm system
CN107426533A (en) A kind of video monitoring image recognition system based on video-encryption compression and image identification
CN111028480A (en) Drowning detection and alarm system
CN116311727B (en) Intrusion response methods, apparatus, devices and readable storage media
CN112906533B (en) Safety helmet wearing detection method based on self-adaptive detection area
CN104853154B (en) A kind of extraction of crag deformation information and alarm method based on moving image
CN103338336A (en) Video detection method for firework early warning of overhead power transmission line passage
CN103152558A (en) Intrusion detection method based on scene recognition
CN114155470A (en) River channel area intrusion detection method, system and storage medium
CN116527850A (en) A self-adaptive detection method for drowning prevention based on video surveillance
CN117612109A (en) Step-by-step detection method for non-specific foreign matter invasion limit in railway running track range

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: Nanchang high tech Industrial Development Zone, Jiangxi Province

Patentee after: JIANGXI HONGDU AVIATION INDUSTRY GROUP Co.,Ltd.

Address before: 330000 Jiangxi city in Nanchang Province, the new bridge box 460 box 5001

Patentee before: JIANGXI HONGDU AVIATION INDUSTRY GROUP Co.,Ltd.