CN106060467A - Safety monitoring system - Google Patents

Safety monitoring system Download PDF

Info

Publication number
CN106060467A
CN106060467A CN201610457785.4A CN201610457785A CN106060467A CN 106060467 A CN106060467 A CN 106060467A CN 201610457785 A CN201610457785 A CN 201610457785A CN 106060467 A CN106060467 A CN 106060467A
Authority
CN
China
Prior art keywords
monitoring
module
video
video monitoring
audio
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.)
Withdrawn
Application number
CN201610457785.4A
Other languages
Chinese (zh)
Inventor
易飞
缪达瑞
钱志
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.)
Chengdu Jiaze Generale Technology Co Ltd Ste
Original Assignee
Chengdu Jiaze Generale Technology Co Ltd Ste
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 Chengdu Jiaze Generale Technology Co Ltd Ste filed Critical Chengdu Jiaze Generale Technology Co Ltd Ste
Priority to CN201610457785.4A priority Critical patent/CN106060467A/en
Publication of CN106060467A publication Critical patent/CN106060467A/en
Withdrawn legal-status Critical Current

Links

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
    • H04N7/181Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00Details of television systems
    • H04N5/76Television signal recording
    • H04N5/91Television signal processing therefor
    • H04N5/92Transformation of the television signal for recording, e.g. modulation, frequency changing; Inverse transformation for playback
    • H04N5/9201Transformation of the television signal for recording, e.g. modulation, frequency changing; Inverse transformation for playback involving the multiplexing of an additional signal and the video signal
    • H04N5/9202Transformation of the television signal for recording, e.g. modulation, frequency changing; Inverse transformation for playback involving the multiplexing of an additional signal and the video signal the additional signal being a sound signal

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Closed-Circuit Television Systems (AREA)
  • Alarm Systems (AREA)

Abstract

The invention provides a safety monitoring system. The safety monitoring system comprises an audio-video monitoring synchronization module, an audio monitoring module and a video monitoring module, wherein the audio-video monitoring synchronization module can send same clock information into the audio monitoring module and the video monitoring module; the audio monitoring module performs continuous audio monitoring; and the video monitoring module controls video monitoring according to sounds monitored by the audio monitoring module.

Description

A kind of safety monitoring system
Technical field
The present invention relates to security monitoring field, specifically a kind of audio frequency and video Cooperative Security monitoring system.
Background technology
Along with social development, the progress of science and technology, monitoring becomes the part that society is indispensable already.Monitor and pacify in society Entirely, traffic administration, each face of each side such as intelligent city plays the most important role.
Even to this day, video monitoring is still the absolute leading role in monitoring field, the probe transport that such as highway is seen everywhere, Closed monitor system of community etc..
In video monitoring system, the video that photographic head is shot with video-corder, need to carry out storing in case later retrieval is inquired about, by Huge in video source code flow data volume, this is accomplished by source code flow is carried out Video coding process.
Common video encoding standard is for the high definition video encoding and decoding of consumer entertainment purpose, in this case, Owing to raw video image googol is according to amount (a two field picture initial data often needs several million even tens), in order to reach real Time coding, generally require special coding chip and process, the most such, the compressed bit stream produced after coding will occupy phase When big memory space.Dedicated video coding chip function complexity, construction cycle length, later stage cannot be carried out Function Extension, operation Power consumption is high, and needs the data/address bus that huge memory space and height handle up.And monitoring system real-time is the highest, and need Will not stopped monitoring record for a long time.So common video high definition coding is not suitable for monitoring system.So in reality In monitoring system, its coding standard also has different from common video encoding standard, and pickup quality is the most relatively low.But low pixel Quality monitoring sometimes the retrieval and inquisition in later stage can be produced certain harmful effect, the most many times highway monitoring probe Monitor video does not sees the number-plate number at all.
Meanwhile, video monitoring also has the shortcoming that it is intrinsic, as in night or fog weather effect can big heavy discount, its There is the biggest blind area in monitoring range.
Increasing phenomenon shows, simple video monitoring the most increasingly cannot meet the monitoring demand of sustainable growth ?.
Summary of the invention
In order to overcome above-mentioned technical problem, the invention provides a kind of safety monitoring system, it includes that audio-video monitoring is same Step module, Voice Surveillance module and video monitoring module, it is characterised in that:
Audio frequency can be sent identical clock information depending on monitoring synchronization module and enter Voice Surveillance module and video monitoring module;
Voice Surveillance module carries out successional Voice Surveillance;
The sound control video monitoring that video monitoring module arrives according to Voice Surveillance module monitors.
Wherein video monitoring module according to Voice Surveillance module monitors to sound control video monitoring specifically include:
Calculate the audio energy values of current audio frame;
When described audio energy values exceedes described first threshold with background sound energy value variable quantity, it is determined that different for occurring Ordinary affair part, and start the video monitoring to described anomalous event.
Accompanying drawing explanation
Fig. 1 shows audio frequency and video cooperative monitoring method flow diagram based on audio detection;
Fig. 2 shows the video frequency monitoring method flow chart under audio frequency control;
Fig. 3 shows the audio frequency and video cooperative monitoring device surveyed based on sound inspection;
Detailed description of the invention
The development of Voice Surveillance lags behind video monitoring due to various reasons always, but Voice Surveillance but has The advantage that many video monitorings are incomparable, so Voice Surveillance the most increasingly comes into one's own.
Compared with video image, Voice Surveillance needs data volume to be processed much smaller, so audio coding processing speed Comparatively fast, it might even be possible to directly using the embedded type CPUs such as ARM to carry out coded treatment, the construction cycle is relatively short.Owing to being software, Later stage function expansibility is strong.Further, there is not blind area in Voice Surveillance, though night or serious fog weather normal work in the same old way Make.
The shortcoming that visible Voice Surveillance compensate for video monitoring completely.These features just because of Voice Surveillance so that Voice Surveillance seems more and more important in monitoring field.
Audio frequency can be blocked one side alone under many circumstances, but more in the case of, independent Voice Surveillance or video prison Control all cannot well play monitoring effect, is at this time accomplished by Voice & Video and monitors collaborative carrying out.
For monitoring, every kind of monitoring has its fixing monitoring objective event, the traffic road such as certain section is easily caused trouble The monitoring objective event of section is vehicle accident.But for any monitoring, we cannot be known a priori by all possible abnormal thing Part type and set monitoring objective event according to anomalous event type, namely we cannot ensure to occur in monitoring range All anomalous events be all monitoring objective event, although for example vehicle accident is monitoring objective event, but also can supervise Control is to other anomalous events many, and for example vehicle passes through, pedestrian ran and some criminal activity.And these anomalous events Although not being monitoring objective event, but the most all can play a big part when such as the police solve a case, say, that at these Closing, in addition to monitoring objective event, other a lot of anomalous events are required for carrying out record, have arrived in the dead of night, and highway does not has car No one, is the most obviously made without monitoring record.
Just because of this feature of monitoring, this patent proposes a kind of audio frequency and video cooperative monitoring controlled by Voice Surveillance Method, uses the change of Voice Surveillance detection surrounding enviroment, if periphery situation is calm, supervises only with simple audio frequency Control, closes video monitoring, runs power consumption and data space to save;When Voice Surveillance detects the notable change of peripheral sound Change, calculate the general direction in abnormal sound source, adjust video monitoring probe and obtain optimal video monitoring direction, open and regard Frequently watch-dog record the video monitoring information of current time, the image before the most also occurring according to anomalous event and abnormal thing Image in part generation, the image change after anomalous event generation find anomalous event, and the characteristic analysis according to monitor event is different Whether the type of ordinary affair part is monitoring objective event, if it is, perform set operation, as report to the police, otherwise, record one different Regular signal.
Audio-video monitoring process under audio frequency control is as shown in Figure 1.
Wherein Voice Surveillance carry out all the time with ensure monitoring persistence (in some application scenario, it is possible to only detect different Voice Surveillance coding record is just carried out during Chang Shengyin), double track is (in some application scenario, it is also possible to increase more sound channel and carry High monitoring performance, but monitoring cost can be improved) audio sample PCM Zhi Hou mono-road delivers to audio coder and encodes, another road Carry out sound variation detection.Video monitoring is closed under normal circumstances, if it occur that anomalous event, and must be along with The notable change of sound, when detecting this sound variation, records the time of anomalous event generation and writes Voice Surveillance Code stream.Analyze the orientation of sound simultaneously, thus adjust video monitoring photographic head and most preferably monitor angle to obtain, open video prison Control, record anomalous event send out sum time so that and Voice Surveillance code stream synchronize, carry out video coding and storage;Meanwhile, contrast Front and back the difference of image determines the image block corresponding to anomalous event and attempts to be analyzed, if it is possible to successful analysis goes out different Often event type is monitoring objective event, then process according to the processing mode set, as reported to the police;Failing to detect as monitoring Object event, then recording exceptional event is in case monitoring personnel's later retrieval is inquired about.
Flow chart is described in more detail as shown in Figure 2 for the video monitoring part under audio frequency control:
In upper map flow chart, Voice Surveillance is stereo double channel, in actual application, can increase choosing according to practical situation Select more multichannel.
First several noun is defined:
1. audio frame: the processing unit of audio coding, comprises some sound channels some PCM value (such as AAC audio standard alliteration Road encodes, and each of which audio frame comprises two sound channels, each 1024 the PCM values of each sound channel.Channel number is permissible according to different situations Difference, but for same audio standard, in every frame, the PCM value number of every sound channel is fixed value, and AAC standard is fixed as 1024 Point).Here, audio frame can be further divided into normal sound frame and abnormal sound frame.
2. normal sound frame: refer to be not detected by the audio frame of abnormal sound.This audio frame represents does not has anomalous event to send out Background sound time raw.
3. abnormal sound frame: refer to the audio frame that audio frequency changes be detected, it is meant that (anomalous event is divided to occur in that anomalous event For monitoring objective event and non-supervised object event, every kind of monitoring application has monitoring purpose, such as, unattended equipment room is prevented Stealing monitoring system, properly functioning unmanned entrance of machine room is normal event, and someone trespasses as monitoring objective event, slight earthquake Article are caused to drop as non-supervised object event), in this case need open video monitoring and carry out anomalous event analysis.
4.E_curr_l: present frame L channel PCM energy value.
5.E_curr_r: present frame R channel PCM energy value.
6.E_curr: present frame PCM energy value, i.e. left and right acoustic channels PCM average energy, E_curr=(E_curr_l+E_ curr_r)/2。
7.E_back: background sound average energy, the i.e. average energy of normal sound frame.Front M can be used just Often the meansigma methods of voiced frame is used for calculating background sound average energy E_back, and M value can be according to practical situation monitoring objective thing The speed of part sets to reach optimum efficiency, such as 100.
8.Threshold: abnormal sound frame threshold value, if present frame E_curr contrast E_back exceedes threshold value, then Mean to detect obvious sound variation, namely mean the generation of anomalous event.Due to different use occasion background sounds Average energy is different, monitoring objective event is different, and its value can be varied from, depending on needing according to practical situation.
There is the moment in 9.T: anomalous event.Voice Surveillance and video monitoring share a clock, abnormal thing once detected Part occurs, and needs to record generation time information, and recorded Voice Surveillance code stream and video monitoring code stream to ensure sound simultaneously Audio video synchronization, in order to can know rapidly that when future Query the abnormal moment occurs the moment.
10.L_max: in general, anomalous event once occurs, and generally requires and continues for some time, so once detecting To anomalous event, after opening video monitoring, L_max frame can be monitored the most continuously, after L_max frame, stop video monitoring, and And re-start accident detection.The length of L_max selects to need according to monitoring application scenario and the length of monitoring objective event Spend and select, such as the monitoring system of monitoring highway traffic accident, the generation of traffic accident the most in a flash, can set the shortest L_max value.
11.L: Video coding enumerator, it is ensured that anomalous event once detected, needs lasting video monitoring record L_max Frame.
Whole video monitoring process such as following steps as shown in Figure 2 represent:
1.E_curr is initialized as 0 or known current background acoustic energy meansigma methods.
2. dual-channel audio PCM sampling input,
3. calculate left audio frame energy value E_curr_l, calculate right audio channel frame energy value E_curr_r.
4. calculate present frame dual-channel audio energy value E_curr=(E_curr_l+E_curr_r)/2.
5. judge whether E_curr/E_back exceedes threshold T hreshold, if it exceeds threshold value, then detect different Ordinary affair part, enters next step and carries out anomalous event process, otherwise update E_curr, return to step 2 and again detect.
6. record the time of origin T of anomalous event, and frame of video monitor counter L is reset, be ready for abnormal thing Part detection and video monitoring.
7. can cross the information analysis abnormal sound orientation such as double track energy difference and time difference, phase contrast.
8. adjust photographic head to obtain optimal monitoring angle, and open video monitoring probe and proceed by video monitoring figure As picked-up and coding.
If 9. current video monitoring record frame number reaches L_max frame, then exit video monitoring, enter monitoring post processing step Rapid 13.
10. photographic head picked-up video image encodes.
11. find anomalous event object to be correlated with by the difference of the video frame image before, during and after contrast anomalous event Image block.
Anomalous event object associated picture block, according to the monitoring objective affair character preset, is analyzed determining by 12. Whether it is monitoring objective event, meanwhile, refresh counter L, return to step 9.
13. if during anomalous event is analyzed, successfully be detected this anomalous event is monitoring objective event, then by pre- If mode process, such as report to the police;The most only it is recorded as an anomalous event to occur.
Anomalous event time of origin T and Video coding code stream are stored memorizer by 14., come back to step 2, carry out complete A new Voice Surveillance.
As shown in Figure 3, monitoring system includes audio-video monitoring synchronization module, Voice Surveillance module and video monitoring mould Block.
In view of synchronization and the determination in anomalous event generation moment of Voice & Video monitoring, both need to share one Audio-video monitoring synchronization module in signal source of clock, i.e. figure.Audio frequency can send identical clock information depending on monitoring synchronization module Enter Voice Surveillance module and video monitoring module.
Left and right acoustic channels PCM value one road is delivered to Voice Surveillance module and is carried out successional Voice Surveillance and (need not sound at some Frequently in the case of seriality monitoring, it is also possible to only just carry out Voice Surveillance coding record when detecting anomalous event), and by sound Frequency monitoring encoding code stream together recorded memory module A together with temporal information;Another road is then sent into video monitoring module and is carried out different Often sound detection.
In video monitoring module, video monitoring control module controls whole video monitoring process.In video monitoring module In, first in energy computation module, the audio frame inputted directly is calculated the energy value E_curr of present frame, and is not examining In the case of measuring abnormal sound, update background sound energy value E_back, enter in ensuing abnormal sound detection module The contrast of row E_curr and E_back numerical value processes.Abnormal sound once detected, then represent anomalous event and occur, then in exception In event angle detecting module, according to E_curr_l, E_curr_r, and combining left and right acoustic channels time difference, phase contrast etc. is additionally believed Breath, it is judged that the direction that anomalous event occurs, and notify video monitoring control module, adjust the direction of video monitoring probe, start Carry out anomalous event analysis and Video coding record.Analyze in module in video anomalous event, can according to before anomalous event, in, The difference of rear video picture frame finds the image block corresponding to anomalous event, in conjunction with set in monitoring objective affair character storehouse Record monitoring objective feature, is analyzed the image block corresponding to anomalous event processing contrast to be confirmed whether as monitoring objective Event, if monitoring objective event, then carries out reporting to the police or other respective handling according to the processing mode set;Failing to Detecting it is monitoring objective event, the most only report anomalous event occurs.In spite of monitoring objective event, send out in anomalous event After life, L_max frame video image all can be encoded and be stored into memory module B by video encoding module.Wherein memory module A Can be same module with memory module B, with can be as disparate modules.

Claims (2)

1. a safety monitoring system, it includes audio-video monitoring synchronization module, Voice Surveillance module and video monitoring module, its It is characterised by:
Audio frequency can be sent identical clock information depending on monitoring synchronization module and enter Voice Surveillance module and video monitoring module;
Voice Surveillance module carries out successional Voice Surveillance;
The sound control video monitoring that video monitoring module arrives according to Voice Surveillance module monitors.
2. a safety monitoring system as claimed in claim 1, wherein video monitoring module is according to Voice Surveillance module monitors To sound control video monitoring specifically include:
Calculate the audio energy values of current audio frame;
When described audio energy values exceedes described first threshold with background sound energy value variable quantity, it is determined that for there is abnormal thing Part, and start the video monitoring to described anomalous event.
CN201610457785.4A 2016-06-22 2016-06-22 Safety monitoring system Withdrawn CN106060467A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610457785.4A CN106060467A (en) 2016-06-22 2016-06-22 Safety monitoring system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610457785.4A CN106060467A (en) 2016-06-22 2016-06-22 Safety monitoring system

Publications (1)

Publication Number Publication Date
CN106060467A true CN106060467A (en) 2016-10-26

Family

ID=57168041

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610457785.4A Withdrawn CN106060467A (en) 2016-06-22 2016-06-22 Safety monitoring system

Country Status (1)

Country Link
CN (1) CN106060467A (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106454282A (en) * 2016-12-09 2017-02-22 南京创维信息技术研究院有限公司 Security and protection monitoring method, apparatus and system
CN110211316A (en) * 2019-06-05 2019-09-06 深圳市梦网百科信息技术有限公司 A kind of monitoring and early warning method and system
CN112992340A (en) * 2021-02-24 2021-06-18 北京大学 Disease early warning method, device, equipment and storage medium based on behavior recognition

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150081721A1 (en) * 2012-03-21 2015-03-19 Nikolay Ptitsyn Method for video data ranking
CN105338294A (en) * 2014-08-07 2016-02-17 富士通株式会社 Monitoring device and method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150081721A1 (en) * 2012-03-21 2015-03-19 Nikolay Ptitsyn Method for video data ranking
CN105338294A (en) * 2014-08-07 2016-02-17 富士通株式会社 Monitoring device and method

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106454282A (en) * 2016-12-09 2017-02-22 南京创维信息技术研究院有限公司 Security and protection monitoring method, apparatus and system
CN110211316A (en) * 2019-06-05 2019-09-06 深圳市梦网百科信息技术有限公司 A kind of monitoring and early warning method and system
CN110211316B (en) * 2019-06-05 2021-12-21 深圳市梦网视讯有限公司 Monitoring early warning method and system
CN112992340A (en) * 2021-02-24 2021-06-18 北京大学 Disease early warning method, device, equipment and storage medium based on behavior recognition

Similar Documents

Publication Publication Date Title
CN105338294A (en) Monitoring device and method
US10127452B2 (en) Relevant image detection in a camera, recorder, or video streaming device
CN107071342B (en) A kind of anomalous event automatic push and monitoring method and system based on historical operation
US10997423B1 (en) Video surveillance system having enhanced video capture
CN105915862A (en) Video processing method
JP5919365B2 (en) Video summarization method and apparatus
TW201724866A (en) Systems and methods for video processing
CN201114536Y (en) A video monitoring system
CN106060467A (en) Safety monitoring system
CN101089915B (en) Embedded high clarity, digital and network high speed pickup superspeed snapping system
JP2013127716A (en) Abnormal state detection system for congestion
WO2016201960A1 (en) System and method for traffic emergency intervention
CN105227918B (en) A kind of intelligent control method
CN201830388U (en) Video content collecting and processing device
CN107071337A (en) The transmission method and device of a kind of video monitoring image
CN210666820U (en) Pedestrian abnormal behavior detection system based on DSP edge calculation
CN106027981A (en) Storage control system
CN102622579B (en) Face early warning method and system
CN103188468A (en) Study on useful video recognition video tape recorder
US20140267724A1 (en) Recording system and method for vehicle
CN101808229A (en) Video stream rapid-playback system based on feature tag
CN201717988U (en) All-in-one machine with intelligent office monitoring system
CN104050785A (en) Safety alert method based on virtualized boundary and face recognition technology
CN101848369A (en) Method for detecting video stop event based on self-adapting double-background model
CN103905789B (en) The method and its device of the preservation, search and ex-post analysis of dynamic analyze data

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WW01 Invention patent application withdrawn after publication

Application publication date: 20161026

WW01 Invention patent application withdrawn after publication