Forensic Enhancement

Forensic Enhancement – Increase safety, expose the truth, reduce costs and prove your case

Unlike movies you might see in a theatre, security camera recordings have low resolution, slow speed and reduced clarity. These settings extend recording time, but they also reduce the forensic value of the stored media. Quality may be further sacrificed due to cheap equipment, faulty installation and a resolution mismatch between the cameras and the recording unit.

Movies and television have also fostered unrealistic expectations about forensic enhancement. This is commonly referred to as "the CSI effect.” You’ve probably seen a movie or show in which a single grainy image is enhanced to the point that the investigators can clearly see the criminal in the reflection of the victim’s eye, even though the victim started as a small object on the screen. If this type of technology actually existed, a disposable camera could be enhanced to view the night sky with the same magnification as the Hubble space telescope.

Depending on how a security video was recorded -- and extracted, an enhancement company can improve that video to exceed the camera’s "live view” quality. For maximum improvements, each color channel must be isolated before the enhancement process begins. Unfortunately, nearly all of the commercial enhancement products and service providers skip this step. They also may lack the ability to separate sequenced video (where several cameras take turns being viewed) or multiplexed video (where several cameras simultaneously share the same recording screen).

Resolution is the term that describes how many dots (pixels) of information are visible, expressed by their horizontal and vertical values. Digital Video Recorders (DVR) typically record at QVGA (320x240) or VGA (640x480) resolution. Network Video Recorders (NVR) typically record at CIF (352x288), 2CIF (704x288) or 4CIF (704x576).

For all five of these resolution settings, the recorder repetitively saves only the odd numbered rows in one image, and then all the even rows in the next. During playback, the recorder interlaces these rows to display a complete image. Since nearly all security recorders are set to capture below 30 images per second, fast-moving objects may look streaked because the odd and even image lines aren’t matching up. This can actually make the images appear worse during playback.

Another problem is that none of these common recording resolutions match the resolution of the cameras connected to the recorder. In America, cameras transmit using the NTSC (National Television System Committee) standard, which was set 70 years ago. The NTSC standard has a viewable resolution of 486x220 (486x440 after interlacing). This represents the best resolution that can be received by the recorder, regardless of the recorder’s settings.

This means that under the best circumstances, the recorded quality will not exceed 0.1 megapixels (486x220). In contrast, any digital camera that you’re likely to use at home or work probably provides a resolution that’s a hundred times greater.

To overcome this limitation, newer cameras are designed to connect through a standard computer network or their own proprietary protocol. This allows the full capability of the camera to reach the DVR, without interlacing or most of the other challenges associated with analog camera systems. If you’re thinking about using this set-up, just be aware that measuring megapixels does not tell the complete story when gauging the expected clarity of a camera’s picture.