How to optimize your Security Guard Service Perimeter Intrusion Detection
Transcription
How to optimize your Security Guard Service Perimeter Intrusion Detection
d ve er es R ht s ig A ll R – al e C op yr ig ht © 20 0 9 -M au riz io C ar di n How to optimize your Security Guard Service integrating the latest technology in Perimeter Intrusion Detection Copyright © 2009 - Maurizio Cardinale – All Rights Reserved – al e ar di n Monitor and Control: A figure occupies at least 5% of the screen height and the scene portrayed is not unduly cluttered. From this level of detail an observer should be able to monitor the number, direction and speed of movement of people across a wide area, providing their presence is known to him; i.e. they do not have to be searched for. A ll R ig ht s R es er ve d After the introduction in the market of the latest product of Southwest Microwave, the Micropoint Cable, you have the possibility of detecting an intruder with a precision of up to 3 meters all along the protected perimeter. The aim of this document is to compare the performance of MicroPoint Cable with one of the competitors with respect to the CCTV assessment. We adopt the methodology developed by J. Aldridge and C. Gilbert (see References), which is also the basis of the EN 50132-7, “CCTV surveillance systems for use in security application”. Using this methodology provides 1) objective indications of what it is possible to see in a frame, avoiding thus subjective judgements and 2) the possibility of obtaining the same findings when the tests are repeated. This methodology identifies five types of possible cognitive activities when one looks at the screen of operating CCTV system. They are: monitor&control, detect, observe, recognise and identify. Each of these activities is possible depending on what one can see in the screen. In what follows you may read the indications about the percentage which the figure (that is, the potential intruder) has to occupy on the screen height in order to carry out each of those activities. These indications suppose the use of a camera with resolution PAL standard: riz io C Detect: The figure now occupies at least 10% of the available screen height. After an alert an observer would be able to search the display screens and ascertain with a high degree of certainty whether or not a person is present. 20 0 9 -M au Observe: A figure should occupy between 25% and 30% of the screen height. At this scale, some characteristic details of the individual, such as distinctive clothing, can be seen, whilst the view remains sufficiently wide to allow some activity surrounding an incident to be monitored. yr ig ht © Recognise: When the figure occupies at least 50% of screen height viewers can say with a high degree of certainty whether or not an individual shown is the same as someone they have seen before. C op Identify: With the figure now occupying at least 100% of the screen height, picture quality and detail should be sufficient to enable the identity of an individual to be established beyond reasonable doubt. Copyright © 2009 - Maurizio Cardinale – All Rights Reserved SXGA- 720p WSVGA SVGA Height 4001 1080 960 720 600 600 Width 720 1920 1280 1280 1024 800 VGA 2CIF CIF QCIF 480 288 288 144 640 704 352 176 ig 1080p – A ll R PAL ht s R es er ve d After the introduction of the digital technologies, the cameras can now have different type of resolutions. So we need to know what the percentages required are for each activity when the resolution is different from PAL standard. In the following tables1 you will find the different types of resolutions (tab. 1) and the equivalent percentages screen height needed to maintain the required resolution for each cognitive activity (tab. 2): SXGA- 720p Identify 100 38 42 56 Recognise 50 19 21 28 Observe 25 10 11 Detect 10 4 5 Monitor 5 2 2 WSVGA VGA 2CIF CIF QCIF 67 67 84 139 139 278 34 34 42 70 70 139 14 17 17 21 35 35 70 6 7 7 9 14 14 28 3 3 3 5 7 7 14 io 1080p 20 0 9 -M riz PAL C SVGA au Category ar di n al e Table 1: Commonly encountered resolutions, SXGA- is the resolution of the Megapixel camera used in the test ht © Table 2: Equivalent percentage screen heights for different digital resolutions. Green boxes indicate that it is reasonable to achieve the appropriate camera view. Red boxes indicate it may be unreasonable or difficult to achieve an appropriate camera view. op The resolution being compared reflects the lowest resolution in the chain, not necessarily the display screen resolution. There is no significant image compression being applied to the image. The person imaged is of average height (5’4” to 5’8” or 1.64m to 1.76m) C • • • yr ig Caveats: Copyright © 2009 - Maurizio Cardinale – All Rights Reserved 1 These tables, slightly modified, are from reference n° 4. C op yr ig ht © 20 0 9 -M au riz io C ar di n al e – A ll R ig ht s R es er ve d We have chosen to use a Megapixel camera because people usually think that this camera is enough to obtain a high performance CCTV system. This belief underestimates that it is also essential to be able to locate the intruder and zoom on him quickly, before the intruder hides himself in the site. Yet when the field of view is wide, the guard cannot detect where the intrusion is occurring: the search has to be done by other two guards going out to look for the intruder. MicroPoint Cable protection system has been built to be able to locate the subcell in which the intrusion is occurring. Then a camera zooms on it and allows collecting all the essential information for the decision making of what to do next. We have used the Rotakin (see reference n° 2) to simulate an individual moving by the fence of a long perimeter. Then we have placed the Rotakin (the individual) at four distances from the camera, 25, 47, 69 and 113 meters, and we have taken pictures. The competitors use a camera with fixed focal lens, while the Southwest Micropoint Cable system uses a PTZ camera with a 10x zoom lens. The pictures have been taken in perfect conditions of light. The camera has been placed along the perimeter. The pictures on the left side are from the performance of competitors systems. The pictures on the right side are from the Southwest Micropoint Cable system. Each picture reports on the top-right box the type of resolution (always Megapixel 1280x960), the focal length, the distance from which the Rotakin is seen, how many pixel are needed to detect (in % and value), how many pixel are needed to observe (in % and value) and how many pixel the Rotakin (target) occupies of the screen height (value). Let's read the first picture in which the Rotakin is seen from a distance of 25 meters (Fig. 1). The box says that the Rotakin (Target) occupies 103 pixel. Since 48 pixel are needed in order to detect and 106 pixel are needed in order to observe, we can conclude that a guard watching that frame can only detect the intruder, but cannot observe him. This first finding shows that what is usually thought as optimal, that is, placing a camera every 25 meters on the perimeter border, only offers a very poor assessment of what is occurring along the fence despite its costs. Let's examine the first pair of pictures. Here the Rotakin has moved from the camera and it is now 47 meters away from it. The picture on the left has been taken with a camera with fixed focal lens, simulating what a guard sees on the monitor with a standard system. The picture on the right has been taken with a PTZ camera with 10x zoom lens, simulating what a guard sees on the monitor with a MicroPoint Cable system. We can see some differences between competitors' systems and the Southwest Micropoint Cable one. The former can hardly detect the intruder, but it cannot observe his actions. On the contrary the Southwest Micropoint Cable system allows the guard to detect and observe the intruder: the guard can now sees very clearly his actions. In the second pair of pictures the Rotakin has moved 69 meters away from the camera. The competitors' systems cannot detect the intruder. On the contrary the Southwest Micropoint Cable system keeps performing very well and allowing detection and observation. In the third pair of pictures the Rotakin has moved 113 meters away from the camera. The comparison between the two systems is even more striking: the guard watching the monitor of the competitors cannot see any sign of intrusion; whereas the guard looking at a monitor of the Southwest Micropoint Cable system can detect and observe the intruder, and is able to intervene quickly and efficiently. Copyright © 2009 - Maurizio Cardinale – All Rights Reserved d ve er es R ht s ig A ll R – al e ar di n C op yr ig ht © 20 0 9 -M au riz io C Figure 1: Distance 25 meters Figure 2: Distance 47 meters Figure 3: Distance 47 meters 10X Zoom In d ve er es R ht s ig A ll R – al e ar di n Figure 5: Distance 69 meters 10X Zoom In C op yr ig ht © 20 0 9 -M au riz io C Figure 4: Distance 69 meters Figure 6: Distance 113 meters Figure 7: Distance 113 meters 10X Zoom In A ll R ig ht s R es er ve d The following tables summarize what the test has shown, that is, the different chance of monitoring, detecting, observing, recognising and identifying at different distances of the intruder from cameras, when one uses Megapixel 1280x960. We have also added the same type of analysis when one uses VGA 640x480. The standard VGA IP camera is included because it is the most affordable and obtainable in the current CCTV market compared with the expensive and bandwidth demanding Megapixel camera. Depending on the type of lens (fixed or varifocal), we obtain very different CCTV assessment. Table 4 indicates that when a VGA 640x480 PTZ with 10x optical zoom is adopted, the activity of detecting is possible under all conditions. This type of activity may be considered as the Minimum Operational Requirement (MOR). However one should take into account that surveillance guards work in conditions which are not so optimal as those supposed in these tests. Visibility and lighting are not always good and the type of work (the guards are often involved in multiple tasks) takes the guards busy. This means that when the alarm rings the guard has to move his attention to the monitor and quickly understand what happened. We would suggest the building of a system which allows the activity of observing, because this permits to adjust easily the perception and see the intruder and his actions also in bad conditions of visibility and lighting. Megapixel 1280x960 VGA 640x480 47 m 69 m 113 m 25 m 47 m 69 m 113 m MONITOR n.a. n.a. n.a. n.a. YES YES NO NO DETECT (MOR) YES YES NO NO YES NO NO NO OBSERVE NO NO NO NO NO NO NO NO RECOGNISE NO NO NO NO NO NO NO NO IDENTIFY NO NO NO NO NO NO NO NO au riz io C ar di n al e – 25 m -M Table 3: Competitor's performance with fixed lens camera VGA 640x480 - PTZ 47 m 69 m 113 m 25 m 47 m 69 m 113 m MONITOR YES YES YES YES YES YES YES YES DETECT (MOR) YES YES YES YES YES YES YES YES OBSERVE YES YES YES YES YES YES YES YES RECOGNISE YES YES YES YES YES YES NO NO YES YES NO NO YES NO NO NO ht yr ig op C IDENTIFY © 25 m 20 0 9 Megapixel 1280x960 - PTZ Table 4: MicroPoint system performance with PTZ camera with 10x optical zoom Copyright © 2009 - Maurizio Cardinale – All Rights Reserved ve d The tables clearly show that the concentration of resources on the use of Megapixel cameras is mistaken and not economical. Cameras alone cannot provide protection. Cameras have to be integrated in a Perimeter Intrusion Detection system (PID). The best option is adopting a high performance PID such as MicroPoint Cable combined with inexpensive but an effective VGA dome camera. This integrated system is the most cost-effective option that brings excellent and reliable outcomes. C op yr ig ht © 20 0 9 -M au riz io C ar di n al e – A ll R ig ht s R es er As mentioned above, visibility is an issue. Following the methodology of J. Aldridge, C. Gilbert (reference n° 3), we carried out a visibility test using a camouflaged Rotakin. In the test we have placed the Rotakin as an intruder crawling along the fence. This simulates a situation in which the intruder acts with circumspection, trying to hide himself, as usually skilled intruders do. We have used a PTZ camera. Figure 8 is the standard frame from 100 m, supposing a camera placed on a building. As you can see, the skilled intruder is completely out of sight for the surveillance guards watching the monitor. In figure 9, with a field of view of 33 meters, you can see something but you cannot understand whether it is an intruder or something else such as an animal, just causing a nuisance alarm. Only in figure 10, it is possible to locate and detect the intruder as such. We point out that the frame represented in figure 10 is possible only with a PID based on MicroPoint Cable, which can locate and zoom on the area where the intrusion is occurring, at a reasonable cost. This means that MicroPoint Cable is the only system that provides an excellent protection of perimeters from skilled intruders. Figure 8: Field of view 100 meters Copyright © 2009 - Maurizio Cardinale – All Rights Reserved d ve er es R ht s ig A ll R – al e ar di n Figure 10: Field of view 10 meters (10x Zoom) riz io C Figure 9: Field of view 33 meters (5.7x Zoom) -M © 20 0 9 EN 50132-7 CCTV surveillance systems for use in security application Part:7 Application guidelines July 2007 J. Aldridge, The Rotakin – A Test Target for CCTV Security Systems, Home Office P.S.D.B. Publication 16/89 J. Aldridge, C. Gilbert, Performance Testing of CCTV Perimeter Surveillance Systems, Home Office P.S.D.B. Publication 14/95 Ver. 1.0 N. Cohen, J. Gattuso, K. MacLennan-Brown, CCTV Operational Requirements Manual 2009, Home Office P.S.D.B. Publication 28/09 Ver.5.0 C op yr ig ht 1. 2. 3. 4. au REFERENCES Copyright © 2009 - Maurizio Cardinale – All Rights Reserved