The Ultimate Integration — Video Motion Detection

Oct. 27, 2008
Motion detection has overcome many of its early problems, and now there's a universe of effective product available.

Not many years ago, video motion detector (VMD) technologies were receiving a lot of attention from the sharpest electronic security developers. Outdoor motion detectors were especially in vogue, with prestigious groups such as Sandia National Laboratories conducting tests for their suitability in protecting our most sensitive national security sites. Single- and multi-channel suppliers were numerous, and many major CCTV manufacturers carried at least a private-label version of a VMD.

A review of the VMD category at the recent ISC West event in Las Vegas indicates a robust number of VMD suppliers (83 companies listed), but only a handful are dedicated VMD manufacturers. What happened? What caused this upheaval? The answer is one word: integration.

Motion detection is arguably the most integrated technology in video today. It is included in almost every alphabet-soup surveillance solution—DVRs, DVMS's (digital video management systems), and IVS's (intelligent video systems) are just a few. How did we get here, and what is the impact upon users and their security solution options?

Two major factors drove the increased integration of VMD:

1) The initial technical shortcomings of the DVR, and

2) Mother Nature's ability to outfox the smartest engineer when it came to outdoor reliability.

The DVR Factor

When they were first introduced, DVRs were expensive and offered limited storage. The first DVRs using high-speed, high-cost, low-density disks could only compete effectively with tape technology if we could somehow minimize the amount of space required to store quality digital video.

Engineers found two obvious solutions to this problem: integrate outside alarm devices with the DVR, or record via a time schedule. These techniques would restrict the continuous recording that would otherwise burn up the disk space. Somewhere along the line, a very enterprising engineer realized that if we only recorded when a camera “saw” activity or movement, then we could both guarantee capture of desired scenes and conserve expensive and limited storage space.

Technological advances have now almost eliminated disk size and cost as a factor in recording capabilities, but motion detection still holds its own as a security-enhancing, space-saving feature.

Mother Nature's Impact

The second driving force in the rapid integration of motion detection into other products has been the effect of Mother Nature on outdoor applications. We should note here that even the most rudimentary VMD works well indoors, where environmental factors such as lighting, shadows, and unusual movements can be controlled rather easily. In areas such as warehouses, indoor motion detection may even be considered for a primary detection technology. Outdoors, however, is a different matter.

Having been involved with some of the most advanced (and expensive) VMDs of their time, I can attest to Mother Nature's success in defeating the most complex algorithms. Moving shadows from cloud formations, moving shade areas under trees, reflections from snow alongside a plowed roadbed … all of these can result in the most sophisticated VMD being ignored as a false-alarm generator in very short order.

Integrated Solutions Enable Outdoor Use

While storage capacity is no longer an issue, Mother Nature is still the bane of most outdoor VMDs. The good news is that many integrated motion detection products can assist today's security professional in taming Mother Nature's impact. The bad news is that it will require some research and comparison to identify the best mixture of features to apply to each application. However, there are some common options that any responsible security professional can consider when it comes to outdoor applications.

As early as the mid 1990s, perimeter protection experts recommended that VMDs be used in conjunction with other alarm technologies in “gated” alarms. This technique required two different technologies (say a VMD and a fence-mounted sensor) to alarm within a pre-set time window to actually trigger an alarm to an operator.

Today's most effective integrated technology almost always has the gated alarm technique built in. For example, “trip wire” capability can be added to traditional movement detection to serve the same function as multiple sensors. A trip wire in this context is a virtual line drawn onto the video scene as viewed on a monitor. For example, if we use a mouse to draw a line in front of a hallway door, an alarm is generated if anyone crosses the line to enter the room. Virtual trip wires aren't sensitive to wind, shadows or other acts of nature, which means the environment is stabilized for reliable detection.

Other techniques for outdoor motion detection include object detection, tracking and behavior detection.

Object detection is the ability to filter objects by size. In outdoor applications we may wish to ignore items smaller than vehicles that could carry explosives, so we implement object detection and define the minimum size requirements to initiate an alert.

Tracking is the ability to follow an individual or object through the video scene. Tracked objects are highlighted on screen by a colored box that remains with them as they move. Sophisticated systems can track multiple objects in a single scene. The most sophisticated systems (usually integrated video systems) can even track objects from one camera scene to the next. Tracking can also include directional capability. We can track all objects moving towards our building and ignore those moving away.

Many times these various factors can be combined to provide very specific rules for declaring an alarm. Object size, speed and direction can be combined to ensure that a vehicle entering an exit lane will cause an alarm, while a bicycle doing the same thing will not. Since these are software-driven features, we can expect even more advanced capabilities in the future.

Besides the multiple technology examples, integrated VMDs and IVS's can provide myriad special functions useful in precisely defining alarm conditions. Alerts can be triggered by people counts for occupancy, the lingering or loitering of people, a fallen person, or gathering crowds. Objects that are abandoned or removed can cause alerts. Vehicles can be counted and alerts declared for stopped vehicles or unauthorized u-turns.

Pyramid of Solutions

Solutions abound today, and your application and budget will dictate your options.

DVR. At the DVR level, the number and sophistication of features is somewhat tied to product price. Most basic DVRs have some form of VMD. In many cases, though, the VMD is basic, allowing only a few areas to mask (areas within the video scene not part of the VMD). These can be effective for indoor use.

Higher-priced and more feature-laden DVRs may include more sophisticated detection with up to 10 masked areas per camera, and even some directional capabilities. These units may be suitable for more severe applications including some limited outdoor use.

The drawback to DVRs is that they are hardware-based and may include proprietary solutions not conducive to flexible expansion or interface with other products. I must note here that we are speaking only in general terms.

DVMS. At the DVMS level, we generally have a higher level of feature sets and complexity within the VMD. We have also gone to another price bracket in most cases. These management systems may include features such as object tracking. As more individual parameters are included within the VMD program, it will become more flexible and more able to work successfully in a broader range of applications. Again, a hardware solution may include proprietary technology that is difficult to integrate. Additionally, in larger applications, the hardware costs can become burdensome.

IVS. The IVS solution tops the pyramid, since almost all IVS products include feature-laden video motion detection. Combinations of many of the techniques discussed in this article are found in IVS software solutions. Drawbacks can include software costs (including yearly maintenance and upgrade fees) and inflexibility if the software is tied to specific hardware modules.

Evaluating Solutions

As with all advances in technology, some products are more elegant and well positioned for future advances than others. What should a security director look for in integrated VMDs?

I would suggest software-based open architecture as a primary decision-making factor. Open-architecture software uses formats such as ODBC, which allows communication with other databases within your security systems (such as badging and access control) to maximize integration and flexibility. Some IVS software solutions may require an API to interface to a specific manufacturer's subsystem. You can find out from your existing legacy manufacturers if APIs are available prior to investing in a new software product.

Avoid products that require software customization or extensive added hardware, since this can limit your architecture, build in obsolescence and add unneeded costs. Carefully weigh your needs, your budget and the products being offered.

In some applications, a robust DVR motion detector may be a better choice than an extensively featured program from which you will require only a small percentage of use. On the other hand, sophisticated programs that are hardware independent may be the answer for flexibility and cost-effective growth.

While there are no simple templates to apply, there is a large universe of effective VMD available today. With some definition of needs, research and evaluation, highly effective security can be applied both indoors and outdoors using video motion detection as one of the tools.

Dale S. Duda is a security industry consultant with Sandra Jones & Company. He assists companies with channel strategies and bringing new products to market. Mr. Duda has 27 years of experience in card access, biometrics, smart cards, digital CCTV, DVRs and perimeter detection for government and industry. He has previously written for ST&D.