WO2016076841A1 - Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo - Google Patents

Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo Download PDF

Info

Publication number
WO2016076841A1
WO2016076841A1 PCT/US2014/065039 US2014065039W WO2016076841A1 WO 2016076841 A1 WO2016076841 A1 WO 2016076841A1 US 2014065039 W US2014065039 W US 2014065039W WO 2016076841 A1 WO2016076841 A1 WO 2016076841A1
Authority
WO
WIPO (PCT)
Prior art keywords
video
video data
kpl
data
vms
Prior art date
Application number
PCT/US2014/065039
Other languages
English (en)
Inventor
Alex Sternberg
David Nelson-Gal
Jason Banich
Eric Green
Manqing Liu
Original Assignee
Viakoo, Inc.
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 Viakoo, Inc. filed Critical Viakoo, Inc.
Priority to PCT/US2014/065039 priority Critical patent/WO2016076841A1/fr
Publication of WO2016076841A1 publication Critical patent/WO2016076841A1/fr

Links

Classifications

    • 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
    • H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N17/00—Diagnosis, testing or measuring for television systems or their details
    • H04N17/004—Diagnosis, testing or measuring for television systems or their details for digital television systems

Definitions

  • the present application relates to video surveillance in general. More specifically, the present application discloses monitoring quality of video data in a video surveillance infrastructure.
  • a typical video surveillance application comprises many cameras, camera network switches, recording servers, storage system and video management software (VMS).
  • VMS video management software
  • the goal of this kind of application is to provide situational awareness not only allowing a small number of individuals (e.g., guards) to monitor a broad expanse of physical plant and/or property, but also to provide a recoverable record of events that might have occurred. This capability helps to better understand what happened and facilitates recovery, arbitration of disputes and improvement of flawed procedures. Therefore, one of the most important goals is to make sure that video streams from cameras are recorded properly onto storage systems. Often, video streams are not recorded due to component failures or software errors, or configuration mistakes.
  • the present disclosure illustrates methods of creating definite metrics for video surveillance applications that address these problems, leading to better operational awareness and better efficacy.
  • the disclosure also provides a mechanism that can proactively alert users when there are real issues associated with these video surveillance applications, helping people responsible for the infrastructure to focus only when the applications need servicing attention.
  • KPI Key performance indicators that are necessary to properly measure the health of video surveillance applications and the supporting infrastructure.
  • the performance indicators include: Video Path Uptime (VPU), Video Stream Delivery Index (VSDI), and, Video Retention Compliance (VRC), From these metrics, it is possible to calibrate whether the surveillance infrastructure is operating properly. These metrics can be used to properly alert video network administrators of problems that are actually affecting the video surveillance application. It is also possible to use these metrics to build better analytics to determine root cause of problems as well as build prediction models for potential problems before they occur.
  • VPU Video Path Uptime
  • VSDI Video Stream Delivery Index
  • VRC Video Retention Compliance
  • Figure 1 illustrates a block diagram of an example camera network, whose performance can be monitored utilizing the embodiments of the present invention
  • Figure 2 illustrates a single data stream path for device C I shown in Figure 1 ;
  • Figure 3 illustrates a flowchart of a method of measuring quality of video in a video surveillance infrastructure, according to an embodiment of the present invention.
  • VPU Video Path Uptime
  • VSDI Video Path Delivery Index
  • VRC Video Retention Compliance
  • the three metrics, VPU, VSDI, and VRC are generated for each camera.
  • the metrics can be aggregated together to create overall metrics for a digital video recording server, a site, or a collection of sites, or an entire enterprise.
  • the site is typically a single physical location where all components are located. This maybe a single store, a building or a collection of buildings and grounds located in a common geographic area.
  • the methods and systems disclosed here are suitable for multiple physical locations or sites as well.
  • the process requires measuring the current state of the infrastructure at regular intervals, Each sample measurement collects metrics about the infrastructure either as an instantaneous measure (live) or as an accumulated count since the last sample was taken. Trend metrics plot these metrics from one sample to the next, providing a long-term view of how surveillance infrastructure behaves over time.
  • VPU Video Path Uptime
  • VPU measures the end-to-end availability of end-to-end video data stream (sometimes also called a 'video data path' or 'camera stream') from camera to storage media.
  • a video data stream may have multiple components, such as an incoming video data stream coming from a camera to a server (may be via a switch), and an outgoing video data stream going from the server to a storage device via a storage path.
  • One of the major goals of VPU is to communicate whether camera streams are recording as designed. This metric is, at its core, an aggregation of measures of a distributed relationship where every element of that relationship has to be working for the overall measure to considered working correctly (i.e, the logic outcome is 'TRUE').
  • the example camera network in Figure 1 there is a collection of devices (cameras labeled CI -CI O), which are the points of entry in the respective video paths.
  • the cameras are streaming data through one or more switches (such as, switch 1 , switch 2) to one or more servers (such as server 1), Within the server(s), there are applications that process the data and store it in storage devices such as Disks 1 -3.
  • a device like a camera can have one or more streams of data. To understand the performance of the overall surveillance infrastructure, it is needed to aggregate a metric for each camera stream. Also, persons skilled in the art will appreciate that any number of switches, servers, applications, and/or storage devices can be used.
  • Figure 2 shows a single illustrative data stream path for camera CI , going through switch 1 (via path 202) to server 1 (via path 204). The processed data then is routed to storage device Disk 2 via storage path 206.
  • the data paths may dynamically vary for efficient load balancing and processing. For example, path 202 may lead to switch 2 if switch 1 is overloaded, switch 2 will send the data stream to an appropriate server, and path 206 may lead to Disk 1 or Disk 3 if Disk 2 is temporarily full, Other possible data paths are 'greyed out' in Fig. 2 to highlight a particular data path used in a configuration. However, the idea is that the data stream path is configurable and changeable without moving away the scope of the current disclosure.
  • VPU is defined as the end-to-end stream-uptime percentage.
  • An aggregated VPU score for a collection of cameras is expressed as follows:
  • the VPU of a server X would have a VPU X equal to the sum of all VPU, values of all the cameras (e.g., 'n' number of cameras) recording to that server divided by the number of cameras recording to that server.
  • the VPU for an entire site would be the sum of all camera stream VPU values in that site divided by the number of camera streams within that site.
  • VSDI Video Stream Delivery Index
  • VSDI measures the performance impact of saturation or decay of a video network on video quality. This is different from VPU in that VPU measures camera streams that are in a failed state. VSDI measures the health of camera streams that are still recording data but due to problems, are decaying in the quality of video that is getting recorded. This is because, unlike typical network traffic which provides some guarantee of delivery of data sent along a TCP/IP socket connection, video streams end data using protocols that are tolerate some loss of data to favor keeping up, in real time, with the data stream.
  • this decay can result in dropped packets, which in turn can cause one or more frames if video to be lost due to inter-frame dependencies. Additionally, system performance and storage performance can lead to greater and greater queue depth on 10 paths, eventually leading to dropped frames as well.
  • normal operation for a camera might be configured to generate a video stream at 8 frames a second at a resolution of 4 Mega pixels. At maximum resolution and throughput, this could produce a data stream of well over a Gigabyte of data per minute. Compression and motion detection can significantly reduce this. However, it can be quite significant. Compound this with a surveillance infrastructure connected to dozens if not hundreds of cameras and the stress on networking, compute and storage resources can be tremendous.
  • VSDI is a measure that reflects these problems.
  • each camera stream has a VSDI measure and then this VSDI measure can be aggregated in logical groups of camera streams, either by server, by site, by company or any other collection users choose to evaluate their infrastructure.
  • VSDI is a percentage, its lowest possible value being 0, which implies the system has detected frame loss for that sample. Furthermore, a value of 100% implies all the frames of video have been transmitted successfully.
  • VSDI for a single video stream could have other values that are greater than zero (0%) but less than 100%. This is to reflect the property that there may not be frame loss yet, however, the system is detecting varying degrees of risk to an individual video stream path.
  • VSDI value 80%
  • a rising storage queue depth over the course of several measures could also reduce the VSDI as well or drop packet events from the network that are still within acceptable ranges.
  • a VSDI of 20% implies that the infrastructure is getting pushed to its capacity and dropped frame events are eminent and may have already occurred.
  • VSDI becomes another PI which can be used to correlate information from multiple sources to get to root cause and develop predictive analytics about what the problem might be in a user's infrastructure and perhaps early detection or prediction of failures so users can take steps to ameliorate the problem or add more resources to their infrastructure to accommodate the actual traffic before they start losing data. Given as a percentage between 0%- 100%, these values can be aggregated to create overall metrics for a collection of camera streams recording to the same media, or flowing through the same server or across an entire site.
  • VRC Video Retention Compliance
  • Any video surveillance recording system has physical capacity, which sets some upper limits as to the amount of video data can be recorded. For practical reasons, when these systems run out of space, they must delete older video data to create room for newer video data. This is a necessary and acceptable strategy since most of the video data is somewhat worthless if nothing important has happened. For example, a video stream of an unused or rarely used back door to a facility doesn't need to be saved in perpetuity if nothing has ever happened that is worth watching. The time between when the video stream data is first captured to the moment it has to be deleted to make room for new video is called the retention period.
  • This video retention period is a key dimension of the design of the system and has a significant impact on costs, i.e., more retention time implies more storage.
  • the retention period represents the time an organization has to evaluate whether something has happened that would warrant archiving video clips of interest to be saved more permanently. For example, if a system saved video data for only 48 hours, one may find that a break-in that happened to a warehouse on Friday evening would already have been overwritten when it is discovered on Monday morning.
  • Different camera streams may also have different objectives. For example, for physical security, exterior camera views of a facility may only need to be retained for as long as it takes to discover perimeter security breaches and then perform an investigation, anywhere from 2 to 4 weeks. However, regulations around a pharmaceutical manufacturing line may require retention of data for up to six months, which allows for bad lots to be discovered and then traced back to the manufacturing sequence that might have produced it. Cameras monitoring access to Data Centers that store financial information can have retention windows of several months. A typical retention period may be six months. Saving all camera streams for the maximum retention period simply because some camera streams need to be saved for that long is extremely expensive.
  • VSDI Video Stream Delivery Index
  • traditional measures do not help. Free space is inadequate, Furthermore, measuring the oldest file at any one moment in time fails because some camera streams maybe in compliance with their retention goals while others may not be. Moreover, for certain systems, the presence of an errant file can make the system look like it is retaining data for a long time when actually on going storage associated with a particular video stream could be well below its requirements. As a result, the only way for even the most sophisticated organizations to verify that they are in compliance with goals is to have employees periodically verify each camera, one-at-a-time, to make sure it is still retaining data according to the stated goals. This is expensive and prone to human errors in
  • VRC R / /RG
  • VRC VCR
  • a single video stream can actually have a VCR; value which is greater than 100%, which reflects a situation where the data exceeds the retention goal.
  • VRC doesn't just aggregate in averages as a camera stream that significantly exceeds its retention goal could start hiding problems in other streams not meeting their retention goals. Rather, when normalizing, one uses a maximum value of 100% for any camera stream that is exceeding its goal when the streams are aggregated together. This keeps the aggregate measure between 0% and 100% which facilitates subsequent roll-ups at server, site or company levels. Therefore, for a collection of cameras,
  • the VRC for a site achieves 100% only if all of the video streams are retained for a least as long as their respective retention goals. This measure also allows us to alert on individual streams that are exceeding or missing their goals.
  • VRC VRC of a camera stream by understanding where the VMS is storing the data, often referred to as "Storage Path.”
  • Storage Path Different VMS softwares use different conventions but often times it is a directory on a mounted volume that is either explicitly given or can be derived from volume, path and camera name conventions. Extracting this information from the VMS then allows us to calculate each camera's actual retention period, R, using the methods indicated above. Retention Goals maybe explicitly indicated in the VMS software or may require operators to manually give it to us. From the two, one can calculate VRC for each camera and, in turn, for arbitrary collections.
  • VRC as with VPU and VSDI, provide one or more PIs to determine root cause and predictive analytics to help customers achieve their goals and address problems before they become disasters.
  • FIG. 3 shows a flowchart 300 that summarizes the method of video data monitoring according to the teaching of this disclosure.
  • VPU is calculated for each video stream.
  • each camera may generate multiple video streams.
  • VSDI is calculated for each video stream.
  • VRC is calculated for each video stream.
  • aspects of the disclosure can be implemented in any convenient form.
  • an embodiment may be implemented by one or more appropriate computer programs which may be carried on an appropriate tangible carrier medium.
  • Embodiments of the disclosure may be implemented using suitable apparatus which may specifically take the form of a programmable computer running a computer program arranged to implement a . method as described herein.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Closed-Circuit Television Systems (AREA)

Abstract

L'invention concerne des indicateurs-clés de performances qui sont nécessaires pour mesurer correctement l'état de santé d'applications de surveillance vidéo et de l'infrastructure qui les soutient. Parmi les indicateurs de performances figurent: le temps de disponibilité des trajets vidéo (VPU), l'indice d'acheminement des flux vidéo (VSDI), et la conformité de la conservation des vidéos (VRC). À partir de ces métriques, il est possible d'étalonner le caractère satisfaisant du fonctionnement de l'infrastructure de surveillance. Ces métriques peuvent être utilisées pour alerter correctement les administrateurs de réseaux vidéo sur des problèmes qui affectent réellement l'application de surveillance vidéo. Il est également possible d'utiliser ces métriques pour construire de meilleurs éléments d'analyse afin de déterminer les causes profondes de problèmes ainsi que pour construire des modèles de prédiction de problèmes potentiels avant que ceux-ci ne surviennent.
PCT/US2014/065039 2014-11-11 2014-11-11 Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo WO2016076841A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/US2014/065039 WO2016076841A1 (fr) 2014-11-11 2014-11-11 Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/US2014/065039 WO2016076841A1 (fr) 2014-11-11 2014-11-11 Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo

Publications (1)

Publication Number Publication Date
WO2016076841A1 true WO2016076841A1 (fr) 2016-05-19

Family

ID=55954762

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2014/065039 WO2016076841A1 (fr) 2014-11-11 2014-11-11 Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo

Country Status (1)

Country Link
WO (1) WO2016076841A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107465564A (zh) * 2016-06-03 2017-12-12 德科仕通信(上海)有限公司 Vod服务质量监测系统及方法
US10944993B2 (en) 2018-05-25 2021-03-09 Carrier Corporation Video device and network quality evaluation/diagnostic tool
US11290707B2 (en) 2018-12-21 2022-03-29 Axis Ab Method for carrying out a health check of cameras and a camera system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7519504B2 (en) * 2004-03-31 2009-04-14 Emc Corporation Method and apparatus for representing, managing and problem reporting in surveillance networks
US7523092B2 (en) * 2004-12-14 2009-04-21 International Business Machines Corporation Optimization of aspects of information technology structures
US20100208064A1 (en) * 2009-02-19 2010-08-19 Panasonic Corporation System and method for managing video storage on a video surveillance system
US20130024429A1 (en) * 2010-04-29 2013-01-24 Hewlett-Packard Development Company, L.P. Multi-Jurisdiction Retention Scheduling For Record Management
US8548297B2 (en) * 2008-09-11 2013-10-01 Nice Systems Ltd. Method and system for utilizing storage in network video recorders

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7519504B2 (en) * 2004-03-31 2009-04-14 Emc Corporation Method and apparatus for representing, managing and problem reporting in surveillance networks
US7523092B2 (en) * 2004-12-14 2009-04-21 International Business Machines Corporation Optimization of aspects of information technology structures
US8548297B2 (en) * 2008-09-11 2013-10-01 Nice Systems Ltd. Method and system for utilizing storage in network video recorders
US20100208064A1 (en) * 2009-02-19 2010-08-19 Panasonic Corporation System and method for managing video storage on a video surveillance system
US20130024429A1 (en) * 2010-04-29 2013-01-24 Hewlett-Packard Development Company, L.P. Multi-Jurisdiction Retention Scheduling For Record Management

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107465564A (zh) * 2016-06-03 2017-12-12 德科仕通信(上海)有限公司 Vod服务质量监测系统及方法
CN107465564B (zh) * 2016-06-03 2020-05-19 德科仕通信(上海)有限公司 Vod服务质量监测系统及方法
US10944993B2 (en) 2018-05-25 2021-03-09 Carrier Corporation Video device and network quality evaluation/diagnostic tool
US11290707B2 (en) 2018-12-21 2022-03-29 Axis Ab Method for carrying out a health check of cameras and a camera system

Similar Documents

Publication Publication Date Title
US10750126B2 (en) Systems and methods of measuring quality of video surveillance infrastructure
JP6690011B2 (ja) ストリーミング分析法を用いてネットワーク問題の実効顧客影響度をリアルタイムで測定するためのシステム及び方法
EP3197097B1 (fr) Corrélation kpi multiple dans des protocoles sans fil
CN102387038B (zh) 基于视频检测及综合网管的网络视频故障定位系统及方法
CN202282837U (zh) 视频质量诊断系统
US20120260277A1 (en) Multimedia Test Systems
US20110122773A1 (en) Method, system, and computer program product, for correlating special service impacting events
US9456190B2 (en) Systems and methods of determining retention of video surveillance data
EP3197100A1 (fr) Corrélation à causes multiples dans des protocoles sans fil
US11711509B2 (en) Early video equipment failure detection system
US20090292720A1 (en) Service Model Flight Recorder
CN103577273B (zh) 用于捕获诊断跟踪信息的方法和计算机系统
CN103391425A (zh) 基于时分复用视频分析的监控智能化改造方法及系统
WO2016076841A1 (fr) Systèmes et procédés de mesure de la qualité d'une infrastructure de surveillance vidéo
CN104202576B (zh) 一种智能视频分析系统
CN105119767A (zh) 一种数据自检及自清理的软件运行状态监测方法及系统
US10187250B2 (en) System and method for capacity planning on a high speed data network
CN111178241A (zh) 一种基于视频分析的智能监控系统及方法
CN107317708B (zh) 一种法院业务应用系统的监测方法及装置
WO2024131461A1 (fr) Procédé, appareil et dispositif d'analyse de cause profonde pour un fichier vidéo à saut de trame
CN107465652B (zh) 一种操作行为检测方法、服务器及系统
CN104038736B (zh) 视频数据动态传输方法
US10944993B2 (en) Video device and network quality evaluation/diagnostic tool
CN112203166A (zh) 一种多模型用户健康档案评分方法及装置
US20210021461A1 (en) Switch monitoring system and method of use

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14906033

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14906033

Country of ref document: EP

Kind code of ref document: A1