|
|
|
The Geutebruck factory at Windhagen near Bonn in Germany. (Photo by Geutebruck) |
History
Geutebruck was founded in 1970 by Thomas Geutebruck, father of the current Managing Director. In the early days it produced camera devices using un-perforated 35mm film for transmitting images of cheques in banks and credit institutions. Its XY photographic surveillance camera, which became the world market leader, was produced until 2005.
Meantime the company was making other ground-breaking developments which paved the way for the DVR as we know it today. Geutebruck inventions included: the first video motion detector for outdoor use which could reliably differentiate between environmental changes and real alarm situations; the first ever digital alarm memory which saved the pre-alarm history and the alarm picture, and which worked together with the video motion detector; the first digital recording system for alarm picture sequences with camera/ date/ time indexing and rapid retrieval; and in 1993 the world¡¯s first commercial multi-channel color DVR, the MultiScope. This was the start of a highly successful product series. Networkable, scaleable with client/server architecture and sophisticated alarm management, the MultiScope in various guises proved very adaptable, reliable and easy to integrate into other higher level systems. It maintained its place as flagship product until 2005 when the GeViScope arrived.
Open and Compatible
|
.jpg)
|
|
The GeViScope¡¯s Digital Signal Processing (DSP) technology enables the use of many new virtual devices which are flexible and easily updated. (Photo by Geutebruck) |
The Geutebruck system philosophy has always been to be open and compatible with as many other systems and devices as possible in order to allow customers to benefit from whatever technology is available to them, whether new and cutting-edge or long-standing, tried-and-tested. When it came to developing the next generation of CCTV systems, Geutebruck came up with a high performance digital platform which exploits the ¡®intelligence¡¯ of digital while retaining the valuable characteristics of analogue, and which is designed to be updated with new technologies long after its contemporaries have become obsolete -- the GeViScope.
In a significant departure from conventional design, the GeViScope handles a large number of channels, each with its own Digital Signal Processor (DSP). This means that besides being able to dual stream (simultaneously produce recording and live viewing streams of different qualities from the same source), the processors enable a GeViScope-based CCCTV system to dispense with many conventional CCTV hardware devices (matrix, encoders, VMDs and etc.) and replace them with virtual ones. Virtual devices are easy to install and update, and provide flexibility, enabling each signal to be processed precisely with the way which best suits the operation. Some virtual devices Geutebruck has developed for the GeViScope and other DSP-based hardware support new and unique processes: MPEG4 CCTV encoding, Dynamic Live Streaming (DLS) and Dual-Sensor detection.
MPEG4CCTV Compression
Standard video compression formats adopted from the TV and multimedia world were designed primarily for creating one stream at a time. But, in the CCTV business, encoders need to handle many video-audio channels at once, and have to function in real time, i.e., without any perceptible delay. This is why many CCTV systems using classical compression formats look fine with a few video channels, but have problems with speed, image quality and/or system stability at full load. Geutebruck¡¯s MPEG4CCTV compression process, on the other hand, avoids these problems because it was specially designed to generate large numbers of different video-audio channels simultaneously, without overstretch at high loads or speeds. It is an MPEG4 process but it incorporates variable quality and variable compression. It lets the user define the image quality -- including the GOP sizes -- for each individual recording and each individual live streaming channel from each camera -- whether analogue or IP (or megapixel). Its variable compression process maintains this desired picture quality, while automatically changing the frequency of reference frames in real-time, in accordance with the overall movement in the picture, the motion in selected areas, other event triggers and any scheduled routines.
Focus on User¡¯s Needs
|
.jpg)
|
|
With the display server telling the encoder what size and quality pictures to encode, much less data is generated and the user makes hardware and infrastructure savings. (Photo by Geutebruck) |
As a result of focusing the process on delivering what is important to the CCTV user, and only generating data which are relevant to his or her needs, MPEG4CCTV minimizes data volume, bandwidth and storage costs, and has such a low latency (120ms) that it can be used to control fast dome and PTZ cameras. Users enjoy exceptional display characteristics: smooth, multi-camera synchronized playback, forwards and backwards at any speed (from 1 picture/ 10s to 25fps and even image by image), and, with a single keystroke they can jump back seconds or minutes to view sets of synchronized recordings. Easily to copy for back-up and evidential purposes, MPEG4CCTV footage can be protected with individual password encryption and is not open to manipulation in multimedia editors.
Avoiding Waste
The volume of data routinely generated in a digital CCTV system usually corresponds to what is required to deliver the biggest and highest resolution pictures ever needed in the control room. In most systems, this high volume has to be generated even though only a small proportion of it is used. The rest is waste so Geutebruck has developed a way of avoiding it. This development puts the display server in real-time control of the size and quality of the picture data as it is generated so that only what is really required by the display windows at that moment is ever generated and transmitted -- a fraction of the norm.
Dynamic Live Streaming
Geutebruck¡¯s DSP-based system could already generate vast quantities of compressed data; record a camera at one quality while streaming the same pictures at another (dual streaming); and could vary the quality and picture rates for live streamed channels independently of the recorded ones. But a couple of other new innovations were needed. The first was a kind of speed regulator for monitoring the display PC and ensuring that picture data is never transmitted to it faster than it can process. And the second was a new process called Dynamic Live Streaming (DLS) in which the software running on the display PC provides continuous feedback to the compression hardware about what size window, and what format the viewer requires for each channel. Thus enabling the encoder to match the live channel picture data precisely to the needs of the current display windows.
Given that control rooms normally only need 2 CIF resolution for maximized windows, and CIF is usually adequate elsewhere (or even QCIF in very small windows), automatic switching to CIF reduces data volume by about 50 percent, and any switching to QCIF makes another 10 to 20 percent reduction. Over the whole system this represents a very large reduction of data without any appreciable loss of quality. Switching between formats is fast and imperceptible, and of course the highest quality dominates if multiple PCs access the same camera.
Using DLS cuts out waste and delivers hardware and infrastructure savings. No redundant data is transmitted so less bandwidth is used; and since each display server has less data to process, it can run lots more screens. A single display server running Geutebruck¡¯s display software on a Core2Duo processor can feed 100 screens with live MPEG4 video at 25fps. So DLS is attractive for large new systems and for existing installations where capacity is becoming an issue.
Detecting in a Class of Its Own
|
.jpg)
|
|
Dual-Sensor mode usually halves the number of unwanted alarms. (Photo by Geutebruck) |
Geutebruck¡¯s Dual-Sensor is a detector in a class of its own. It combines the particular strengths of video motion detection with those of video analytics to identify threats more accurately and eliminate almost all unwanted alarms.
Geutebruck has a long and enviable track record of producing high-end video motion detectors for securing extensive and high security sites. In prisons, power plants, airports, defence establishments and others, its VMDs have proved extremely efficient at detecting movement in specified zones and very accurate at eliminating the effects of global environmental changes -- weather or mast movement, for example. Its latest VMD is a virtual one. This software package for its DSP-based CCTV platforms can be run on its own or in tandem with video analytics software, as a so-called Dual-Sensor. The advantages of the latter strategy become more obvious if we look at the similarities and differences between these two types of system.
|
.jpg)
|
|
Figure 1. Dual-Sensor runs VMD and VA independently on the same scene then combines their output so that an alarm is only generated when the criteria for both systems are met at the same time. (Source: Geutebruck) |
Tried and Tested Technology
Video motion detectors work by analyzing the composite video signal from the CCTV camera, or to be more precise, by analyzing an electrical representation of the instantaneous grey shades in the picture and monitoring this for changes in brightness or contrast. In a picture from a fixed camera in an environment with fixed illumination, this change is logically deduced to be the result of activity in the scene. High sensitivity enables the VMD to analyze each individual detection zone in the scene for subtle differences. And comparing all the results enables it to deduce which changes are global, and therefore benign, and which are localized and therefore more likely to represent a hazard. A fast operating cycle enables it to compare and store information from successive images captured by the same camera during the time that a fast-moving target passes across the scene. When small detection zones in the far field are set up to be much more sensitive than those in the near field, the VMD can correct for perspective and an object¡¯s apparent size difference when nearby or far away.
Object Recognition
Video analytics software is good at recognizing different kinds of specified objects in a video scene. It continuously analyzes the pixels in the whole video picture and deduces for itself the basic underlying model, updating it for global changes such as weather and lighting. It continuously compares current pictures with the updated background model, and compares any unnatural changes with the user¡¯s target specification. It allows the user to specify target objects by likely size, speed, direction of movement and status, and lets him link their properties. So the system recognizes ¡®person¡¯ ¡®car¡¯ fast¡¯ ¡®slow¡¯ ¡®wrong direction¡¯, ¡®object appearing¡¯ and ¡®object disappearing¡¯ separately, or only when logically combined with each other. It can therefore trigger alerts when a person moves fast, or a car goes in the wrong direction, but if desired, it will ignore a car going fast or a person going in the wrong direction. The user can also define different areas of interest in the scene and associate them with different target characteristics and alarm response behaviors.
Altogether Better
|
.jpg)
|
|
With Geutebruck¡¯s Interactive Specifier Library (ISL) planners produce tender documents for CCTV systems in minutes instead of days. (Photo by Geutebruck)
|
Geutebruck has designed its software so that both detection systems can run independently to monitor the same scene, but their output can also be logically combined so that an alarm is only passed on when the precise pre-determined criteria for both systems are met at the same time. So for instance, the bird which might trigger a VMD alarm because it meets the system¡¯s movement and location criteria, would not match the VA¡¯s specific size or direction criteria, so no combined system alarm would be issued. And, on the other hand, the wind deflecting the leaves in a shrubbery might deceive the VA into registering the shape of a person or a car, but the VMD would not react so no alarm would be reported.
The Dual-Sensor exploits the fact that by working together the systems¡¯ strengths are enhanced and their weaknesses reduced. Each system alone would produce a small, but (possibly highly) undesirable number of unwanted alarms. However, with two finely-tuned, compatible systems running in tandem this number is halved and the Dual-Sensor has proved able to get very, very close to the ideal situation of no unwanted alarms.
Reaching a Wider Market
In its development program, Geutebruck is currently focussing on bringing the high-end benefits of DSP technology and MPEG4CCTV within reach of a wider market. Not every customer or project needs the full capabilities of the GeViScope, so other more modest platforms and devices are being developed to meet the needs of other groups of users. Hence the introduction of re_porter in 2007 and CAM2IP in early 2008, both of which are aimed at the middle of the market but still offer the benefits of DSP technology, MPEG4CCTV compression and DLS.
By Name and by Nature
The re_porter is a very cost-effective, compact, hybrid DVR which comes with operator software and digital matrix functionality already on board. For use with 4, 8, 12 or 16 cameras, it offers permanent and event-controlled recording at up to 25 pps per channel in any picture format, as well as fully independent simultaneous streaming of up to 25 pps per channel into the network. It supports a synchronized audio channel, any mix of analogue and IP sources, including megapixel cameras from major suppliers and has comprehensive backup functions. Any number of re-porters can networked together to create large scalable CCTV systems.
Where Less Means More
The CAM2IP encoder is for CCTV users who want to convert existing analogue cameras into intelligent network cameras and use them with MPEG4CCTV compression in a GeViScope or re_porter-based CCTV system. CAM2IP not only enables easy network integration, but it puts the intelligent video management next to the camera where it can relieve the network most. Permanent or event-controlled encoding can be managed by the CAM2IP¡¯s built-in activity detection function, via TCP/IP alarm trigger from the host server, and/or directly through its alarm input. The result being that CAM2IP only feeds pictures into the network when they are needed and in the form they are needed -- minimizing bandwidth and network load.
Help for Specifiers
With the aim of easing the lot of beleaguered CCTV planners, Geutebruck recently made a new online tool available free on its Website. The Interactive Specifier Library enables tender documents to be produced in a fraction of the normal time. It provides databases with all the background information and tender texts needed to prepare documentation for a new CCTV project. The planner only has to mark the relevant items and the tool compiles the texts in a file in a standard layout. All the products listed in the library are of course Geutebruck products, but the user can make changes in MS Word once the completed file has been downloaded.
Business strategy
Geutebruck is, and intends to remain, primarily a developer and producer of specialist CCTV solutions but it has cultivated co-operative relationships with several other specialist partners to enable the development of integrated solutions. These solutions deliver additional operational efficiency and cost benefits for specific industries, for example, by integrating CCTV with solutions for cash management in the money transport industry and with high-end home automation systems for the exclusive end of the residential market.
Steady Growth
Geutebrck¡¯s business growth has been steady and sure over its three and a half decades. Currently employing around 200 people world-wide, the company markets its products internationally through eight overseas subsidiaries and through trusted sales partners in more than 30 other countries.
Interview
|
|
We have not blindly followed
popular trends, but have looked for
ways to provide
more effective solutions.
|
|
Katharina Geutebruck
Managing Director of Geutebruck
(Photo by Geutebruck) |
|
More than 30 years of experience in developing and producing CCTV systems, Geutebruck has led the global video surveillance market with its talent for innovation and ability to design open and compatible solutions. The company¡¯s system philosophy has always been to be open and compatible with as many other systems and devices as possible in order to allow customers to benefit from whatever technology is available to them, whether new and cutting-edge or long-standing, tried-and-tested. To learn more about the company¡¯s product development strategies and technical breakthroughs, Sunny Kim, Editor of SecurityWorld INT¡¯L, has interviewed Katharina Geutebruck, Marketing Director of Geutebruck GmbH.
Geutebruck has gained a strong foothold as a leading player in the world¡¯s video surveillance market. Please give us a brief account of how the company has made the success.
By ensuring that we focus on providing effective solutions to our customers¡¯ security problems. By investing in the development of our staff, and as a result providing exceptional levels of customer service. By pioneering products and technologies which provide real value, and by continuously striving for improvement.
Could you tell us about Geutebruck¡¯s product development strategy? How does Geutebruck identify the potential applications or products?
Geutebruck has more than 30 years of experience in the industry. We have seen many fads come and go. What proved to be sustainable over the years is a focus on products/ technologies which solve security problems. We have not blindly followed popular trends, but have been (and always will be) looking for ways to provide more effective solutions. We have also sought to maximize the use of our customers¡¯ existing investments by ensuring backward integration of our products, using open architecture to enable integration with third party systems, and finally, designing our products to be flexible to adapt to (even unknown) future developments.
What do you think are the greatest challenges in the video surveillance market? How does Geutebruck address them?
I see three main challenges:
¡¤Ensuring that our focus remains on solving real-live security problems. Not just racing towards higher specifications or new technologies, but ensuring that such advancements actually have a real impact on improving the security of people and assets.
¡¤Improving the man-machine interface: making it easier and simpler to use the surveillance technology effectively and in the best possible way.
¡¤Improving capacity for identifying real security threats, for instance, by using analytics to reduce false alarms. Enabling the user to focus on situations of real risk /threat only.
Which area do you think is the focal point of the market and what are the emerging technologies in this sector?
This ties in with my last point. The greatest focus at the moment seems to be in developing analytics. But here it¡¯s important not to be distracted by the marketing hype. Users shouldn¡¯t expect science fiction type results; and our development focus has to remain on analytics which actually make surveillance more effective -- not just on what sounds exciting in brochures.
Likewise with the hype surrounding hardware component specifications and other technical improvements, it is always wise to consider whether the so-called improvements offer real benefits in reducing risk or minimizing threats. Sometimes they don¡¯t.
Please tell us if there have been any technical breakthroughs your company has made recently and thus contributed to the security industry.
It is no secret that we are at the forefront of many breakthrough developments in the industry. Just take our GeViScope CCTV platform for instance. Its unique DSP technology opens up new avenues for new kinds of software functionality. Not just through substituting flexible software for traditional hardware, but also through developing new security-focused functions. MPEG4CCTV encoding and Dual-Sensor, our combined VA and VMD system, are examples of what are already available.
What do you think will be the next ¡®Big things¡¯ in the video surveillance industry?
My predictions for the future include: improvements in software; the transfer of analytics intelligence nearer to the cameras; the emergence of consulting expertise in distinguishing between what is a fad and what provides effective and reliable solutions; and the integration of CCTV with other systems into an easily manageable whole!
What will actually reduce risk and minimize threats?
The answer in my opinion relies more on a solid understanding of the challenges and logical application of sensible technology. In my view man is still master of machines!
For more information, please send your e-mails to swm@infothe.com.
¨Ï2007 www.SecurityWorldMag.com. All rights reserved.
|
