Daniel Wan，霍尼韋爾安防集團英國渠道營銷負(fù)責(zé)人，從視頻監(jiān)控和其他一些安防應(yīng)用中的一個認(rèn)知的附加費用到成本效益以及關(guān)鍵組成部分來評估視頻分析的發(fā)展。在明周界防范中已經(jīng)顯示出視頻分析好處，增加安保服務(wù)的效率和它傳遞的投資回報率，特別在預(yù)防性的安防，表明視頻分析技術(shù)的時代確實是終于到來了。

　　視頻分析為大眾所接受的假曙光

　　 自最早的所謂“視頻分析”系統(tǒng)推出以來，作為一個傳統(tǒng)視頻監(jiān)控的可靠的和具有成本效益的附加品，該技術(shù)經(jīng)歷了若干假曙光。雖然視頻分析被預(yù)測作為一個安防產(chǎn)業(yè)多年來顯著的增長領(lǐng)域，但視頻分析技術(shù)還沒有成為廣泛采用的理由。

　　早期的制造商曾在他們的解決方案中，夸大其詞地聲稱系統(tǒng)能夠解決問題，但在過程中卻越來越讓用戶感到不滿。這些解決方案通常是創(chuàng)業(yè)公司創(chuàng)造的，僅僅專注于提供視頻分析而已。因此，他們無法有效地與現(xiàn)實生活中的安防應(yīng)用中的其他組件或規(guī)則來整合。這導(dǎo)致許多早期的系統(tǒng)被以一個唯一的基礎(chǔ)來進行安裝，而客戶往往不愿意為一個新的技術(shù)而花費多余的錢?！　?/P>

　　視頻分析在運輸行業(yè)應(yīng)用中的效益

　　視頻分析在運輸部門應(yīng)用的演變，除了實現(xiàn)履行基本的安全需要外，現(xiàn)在還包涵了若干高級特征，如交通監(jiān)控功能和對運輸樞紐中的留守行李進行檢測。

　　在基本層面上，視頻分析能力，確定運動和車輛(行人)的速度，而不是其他雜散物體，借用該技術(shù)來檢測定義區(qū)域周邊和可疑行為。顯然，在有汽車和行人進入危險區(qū)或管制區(qū)的時候，有安全警報的要求，比如鐵軌。

　　即使是一個袋子，如果單獨被留在機場，也可能成為一個嚴(yán)重的安全威脅。

　　在機場和火車站里無人認(rèn)領(lǐng)/孤單的行李，構(gòu)成明顯的安全威脅，這已經(jīng)成為運輸經(jīng)營者和政府越來越感興趣的領(lǐng)域。再而，民政廳的智能檢測系統(tǒng)(i-LIDS)影像庫認(rèn)證包括一個涉及到物體檢測和跟蹤的特定腳本。視頻分析能夠提醒運輸經(jīng)營者包裹在某處遺落，而這些包裹可能在多個CCTV屏幕中都沒有被發(fā)現(xiàn)。簡單來說，該軟件能夠在物品從某個場景移動后，發(fā)出盜竊警示。

　　視頻分析在CCTV市場中的成長

　　最近，有許多因素促進著人們對視頻分析的興趣和需求的增加，這使得視頻分析被重新考慮來作為一個可行的視頻監(jiān)控選擇。從模擬過渡到網(wǎng)絡(luò)化的CCTV監(jiān)控已顯現(xiàn)出數(shù)字解決方案已經(jīng)被更廣泛的接受，包括視頻分析分析，使得能夠從遠(yuǎn)程來監(jiān)控和管理系統(tǒng)，從而帶來了長遠(yuǎn)的好處。

　　更重要的是，終端用戶和安防系統(tǒng)集成商們過去都無法指出一個智能分析系統(tǒng)標(biāo)準(zhǔn)或認(rèn)證，以便判斷他們采購的有效性。這樣的情況直到民政廳的i-LIDS認(rèn)證的推出才得以改變。

　　i-LIDS是英國政府的視頻分析技術(shù)基準(zhǔn)標(biāo)準(zhǔn)，它由民政部科學(xué)發(fā)展部(HOSDB)授給符合政府使用的基于視頻檢測系統(tǒng)開發(fā)嚴(yán)格標(biāo)準(zhǔn)的安防產(chǎn)品。

　　i-LIDS認(rèn)證是第一個真正獨立的視頻分析系統(tǒng)質(zhì)量和功能的措施，制造商們通過獲得i-LIDS認(rèn)證來達到最好的表現(xiàn)。

        提高安保服務(wù)的效能和效率

　　現(xiàn)在已經(jīng)到了把視頻分析作為提升你的安防人員表現(xiàn)的方法時候了

　　當(dāng)考慮到保安和運營者的成本時，誘人之處在于假如評估使用視頻分析技術(shù)，通過引入該技術(shù)可以立刻減少安防工作人員的人手?jǐn)?shù)量?；蛟S，視頻分析更適合被看作是一種提高運營商和保安的效率的方法。[nextpage]

　　事實證明CCTV運營商在監(jiān)控視頻錄像超過二十到三十分鐘后開始錯過重大事件。這種效果被現(xiàn)在運營商們因要求在控制室里進行監(jiān)控而安裝的大量的屏幕而加乘。視頻分析則能夠用來協(xié)助突出顯著的事件，從而使得運營商能夠做自己最擅長的事情：決定時間是否可疑。因此，安全功能使得在現(xiàn)有的工作人員基礎(chǔ)上就能夠更高效地操作更多CCTV攝像機。監(jiān)控工作也變得更加刺激，且不需要不斷地長時間檢查視頻監(jiān)控屏幕。

        智能分析的效果并不是用更多的攝像機來取代保安員，而是讓運營商能監(jiān)控更多的探頭和屏幕。用智能分析來支援保安人員，提供了基于分析事件來切換警衛(wèi)來進行流動巡邏的能力，從而更有效的覆蓋更大的區(qū)域并且更快地對真實事件進行反應(yīng)，而不僅僅是沿著巡視路線一次次例行巡更?？偠灾?，視頻分析還使得安保隊伍突出正在發(fā)生的潛在危險狀況，并已預(yù)先巡邏防范和預(yù)防事故，而不是僅僅在事后才來做出反應(yīng)。

　　周界防護，視頻分析的核心競爭力

　　今天市面上可以找到許多視頻分析軟件包，提供眾多的功能，但往往并不與安防相關(guān)。然而，大多數(shù)已經(jīng)安裝的系統(tǒng)都集中在一個視頻分析的核心競爭力上面：周界保護。

       總之，視頻分析還使得安防隊伍突出正在發(fā)生的潛在的危險情況，并已預(yù)先巡邏防范和預(yù)防事故。

       周界防護系統(tǒng)可以補充圍欄，或者更換外部紅外，地震探測器等，使用戶能夠查明定義的入侵者可能闖入的特殊區(qū)域。潛在地包括虛擬“圍欄”線，這在入侵者攀爬它時可以觸發(fā)警報(無需警衛(wèi)在其沿線巡邏)，當(dāng)穿過特定方向，或者警戒區(qū)域時，比如附近的道路，將觸發(fā)在這些地方的汽車大小的物體停留時間過長的警報。結(jié)合這些規(guī)則保證了只有可疑行為才引發(fā)安全警報，而不是像當(dāng)?shù)氐囊吧鷦游锘蛱鞖庀嚓P(guān)的影響這樣的無關(guān)對象。

        如何消除在這種“無菌區(qū)”內(nèi)的安全誤報是在HOSDB(英國內(nèi)政部科學(xué)開發(fā)部)對視頻分析系統(tǒng)評價的主要測試項目。隨著越來越多的終端用戶和集成商實現(xiàn)平臺并且在周界探測之類的應(yīng)用中視頻分析的成本效益逐漸顯現(xiàn)，這種特定的情況使得對 I-LIDS認(rèn)證的需求強烈。

　　投資視頻分析的回報

        雖然被視頻分析的好處所吸引，但安全管理人員總是一直面臨一個挑戰(zhàn)，那就是投資被視為需要額外費用的技術(shù)的理由。提供一個明確的投資回報(ROI)的一個方法是，看看一個企業(yè)現(xiàn)有的花費或損失。

        預(yù)防總是比治療便宜。視頻分析有突出可疑行為的能力以，它讓經(jīng)營者能做出反應(yīng)并防止發(fā)生事故，而不僅僅是查看視頻以幫助在事后抓住犯罪嫌疑人，因而可以避免損失和損害，并且見效快。

　　作者：丹尼爾萬　霍尼韋爾安防集團英國渠道營銷負(fù)責(zé)人

【英文原文】

　　Video analytics comes of age

　　Are we about to see the first truebreakthrough for video analytics?

　　Daniel Wan, Channel Marketing Leader UK at Honeywell Security Group, assesses the development of video analytics from a perceived premium add-on to a cost-effective and key component in video surveillance and several other security applications. Its proven benefits in perimeter protection, increasing the efficiency of manned guarding services and the ROI it delivers, particularly in preventive security, shows that video analytics is indeed a technology whose time has finally come.

        False dawns with mass acceptance of video analytics

        Since the very first so-called "video analytics" systems were introduced, the technology has experienced a number of false dawns in terms of its acceptance as a reliable and cost-effective addition to traditional video surveillance tools. Although predicted as a significant growth area in the security industry for many years, the wide adoption of analytics has not materialised for a number of reasons.[nextpage]

        Early manufacturers made impressive claims for their solutions, often over-promising on what systems were able to deliver and causing customer dissatisfaction in the process. These solutions were usually created by start-up companies and were focused only on delivering video analytics. As a consequence, they were incapable of being effectively integrated with other components or disciplines in real-world security applications. This also resulted in many early systems being installed on a trial-only basis, with customers often unwilling to pay a premium for the new technology.

        Benefit of video analytics in transport applications

        The use of video analytics in the transport sector has evolved beyond fulfilling basic security needs to include advanced features such as traffic monitoring functions and detection of left-behind luggage within transport hubs.

        On a basic level, the ability of video analytics to identify the movement and speed of vehicles (and people), as opposed to other spurious objects, lends the technology to monitoring of perimeters and suspicious behaviour within defined areas. Clearly, there is a requirement for security alerts where there is a danger of cars or pedestrians entering hazardous or restricted zones, such as railway tracks.

        Even a lone bag can become a significant security threat if left unattended at an airport

        Unclaimed/ orphaned luggage poses a significant security threat in airports and railway stations and, as such, has become an area of increasing interest for transport operators and governments. Again, the Home Office's Imagery Library for Intelligent Detection Systems (i-LIDS) certification contains a specific scenario relating to detection and tracking of objects. Video analytics can alert transport operators to packages left within a scene, which otherwise may go undetected in the midst of multiple CCTV screens. Similarly, the software can highlight instances of theft, where objects have been removed from a scene.

        Growth of video analytics in the CCTV market

        Recently, a number of factors have contributed to an increase in interest and demand, enabling video analytics to be reconsidered as a viable video surveillance option. The growing transition from analogue to networked CCTV surveillance has seen a wider acceptance of digital solutions, including analytics, leading to further benefits in terms of being able to monitor and manage systems from remote locations.

        More importantly, end users and security system integrators were historically unable to refer to a standard, or accreditation, for analytics systems in order to judge their effectiveness before purchase. This was the case until the introduction of the Home Office's i-LIDS certification.

        i-LIDS is the UK government's benchmark standard for video analytics technology and is awarded to security products judged to have met the stringent criteria specified by the Home Office Scientific Development Branch (HOSDB) in the development of video-based detection systems for government use.

        Manufacturers meeting the highest level of performance classification receive the i-LIDS accreditation, which is the first real independent measure of the quality and functionality of a video analytics system.

　　Improving efficacy and efficiency of manned guarding services

        The time has come to view video analytics as a way of improving the performance of your security staff

        When the costs of manned guarding and operators are considered, it's tempting when assessing video analytics to assume that the number of security staff can be immediately reduced by introducing the technology. Perhaps it's more appropriate to view video analytics as a way to improve the efficiency of operators and security guards.[nextpage]

        It has been proven that CCTV operators begin to miss significant events after monitoring video footage for more than twenty to thirty minutes. This effect is multiplied by the large number of screens that operators are now required to monitor in a control room environment. Analytics can be used to assist in highlighting notable events, thereby allowing operators to do what they do best: deciding whether an incident is suspicious or not. Security functions can therefore handle more CCTV cameras more effectively with their existing staff. The job of monitoring is also made more stimulating, removing the need to constantly survey a wall of video screens for long periods.

　　In replacing security guards with yet more cameras, the net effect is to create more footage and screens for the operators to monitor. Supporting the guards with analytics provides the ability to switch guards to mobile patrols based on analytics incidents, thus covering a larger area more effectively and responding to real incidents more quickly, rather than just following a routine guard tour. Above all, video analytics also enables security teams to highlight potential risk situations that are taking place and have patrols pre-empt and prevent incidents rather than just respond after an incident has occurred.

        Perimeter protection, video analytics' core competency

       There are numerous video analytics packages available in the market today, all offering a multitude of capabilities, and not always linked to security. However, the majority of installed systems have focused on one of video analytics' core competencies: perimeter protection.

　　Above all, video analytics also enables security teams to highlight potential risk situations that are taking place and have patrols pre-empt and prevent incidents

　　Perimeter protection systems can supplement fences, or replace external PIRs, seismic detectors, etc., allowing the user to identify specific areas where intruders will be identified. Potentially this includes virtual ‘fence' lines that will trigger an alarm when an intruder climbs over it (rather than when a guard patrols along it), tripwires that trigger when crossed in specific directions, and alert areas, such as nearby roads, which will trigger if a car-sized object dwells in them for too long. Combining these rules ensures that only suspicious behaviours trigger the security alarms and not spurious objects such as the local wildlife or weather related effects.

　　Eliminating such false security alarms in a sterile zone is a primary test in the HOSDB's evaluation of video analytics systems. This specific scenario continues to be a much demanded i-LIDS accreditation as increasing numbers of end users and integrators realise the performance and cost benefits of video analytics in perimeter detection type applications.

        Return on investment on video analytics

　　Although attracted by the benefits of video analytics, security managers have always faced challenges in justifying the investment in what has been perceived as a premium technology. One method of providing a clear return on investment (ROI) is to look at the existing costs for a business, or its losses.

　　Prevention is almost always cheaper than cure. Analytics' ability to highlight suspicious behaviour and let operators react to and prevent them from turning into incidents rather than just reviewing video to help catch suspects after the event, can provide quick returns in avoiding loss and damage.

        Daniel Wan

　　Channel Marketing Leader UK

　　Honeywell Security Group