For years the security community has come to know video surveillance as the “silent witness.” In many court cases, the legal community has trusted video footage as the perfect replay of a crime event and the jury will typically declare, “I can see what is happening and that is good enough for me.” However, much of this history of the silent witness was built up through the years of analog video. Those who read my column on a regular basis know that ever since digital video started taking off I have pointed out how it may be harming our professional image. How can that be? It is not so much about us having an exciting new digital video palette to work with, but rather how images afforded by that technology may be technically compromised.

A comment from video forensic expert Grant Fredricks echoes this observation: “The recent migration from analog to digital video security systems has made it easier for system owners to manage the images at less cost, but the convenience of digital increases the responsibility to ensure accuracy.”

My question is, “Are you relaying this important information to your customer or prospect when laying out and selling a video system?” A video security system should be laid out with the customer for one main purpose: to achieve the best video evidence possible. 

More Cameras, Better Results

So what is the best way to specify video cameras for a good surveillance system? My first rule of thumb would be to take what the customer thinks he/she needs and double or triple the number of cameras and recording capacity. If you end up with something in the middle then you have done everybody a better service, especially with reference to video evidence.

OK, so how many cameras and where? A key question would be asking your customer what they want the camera(s) to identify. Here are some guidelines to satisfy the answer to that question (all pixel measurements are before compression has taken place):

Facial identification — Person’s height should cover 100 percent of the screen, with the head being 15 percent of the body and digitally not less than 90 pixels high

Facial recognition — Person’s body should fill at least 50 percent of the screen height, and digitally the body should not be less than 290 pixels

Intruder detection — Entire person should fill 10 percent of the screen, and digitally the body should not be less than 60 pixels

Crowd control and monitoring — Entire person should not be less than 5 percent of the screen and digitally 30 pixels high

Vehicle number recognition and identification — License plate should not be less than 5 percent of the screen height and digitally 30 pixels high
Compression and Other Key Points

All digital video is compressed at some point; that is just the nature of the beast. There are two types of compression: lossy and lossless. The challenge is to find a good middle ground as most video forensic people will favor lossless, as very little of the image data is lost. In lossy, as the name implies, much can be lost. However, many manufacturers favor lossy because various compression codecs can be used to derive a decent-looking image with very little distortion.

Another consideration is if a camera is placed too far away from a target area and/or a lens is selected to cover a large area (typically the case when a customer wants save money by having a single camera cover a large area) then digital enlargement is required and the quality of the video evidence will suffer. That begs the question: Was it such a good investment after all?

Always be sure to take the time to go through the views with customers by bringing along some demo cameras and lenses. You will end up selling more cameras.  (Sales tip: For on-site camera sales, rig up a 20-foot extension pole with a simple camera swivel mount on the top. Place your best camera with a good varifocal lens on top, combine it with a portable monitor/DVR setup and take it out as a demo unit to show the customer what they will get.)

Remember, you will never get a better digital picture than what is coming out of your camera. Trying to compress video too much will cause small anomalies in the digital video called artifacts.

Video forensic experts and criminal defense attorneys are having a field day in this area. I have seen video artifacts of a gun in a robber’s hand one moment and gone the next, or a case of a criminal moving toward a police officer one moment and then suddenly moving away. Which was it, and can the rest of this video evidence be trusted beyond a reasonable doubt?

Again, the above are areas where we are technically falling short. We are not placing cameras or enough of them, in strategic places for good identification. We are trying to compress images and criminal activity too much, causing strange behavior from our silent witnesses.

Courtroom Credibility Is Suffering

There was another incident recently in the news that showed the public how digital video cannot be trusted. Some of you may have seen the video. It was the legal case of the State of Florida vs. Claudio Muro.

The exhibit was time-lapsed video displayed from a nanny cam recording. While those of us in the industry understand how time-lapse recording shows slices of time rather than continuous video, the public, and worse yet the jury, saw this jerky time-lapse video as someone violently rattling a baby.

The person was initially found guilty and it took 29 months in jail to have a video expert prove the footage was misleading and get the charges dropped. I have heard rumors the accused may be suing the manufacturer on claims that more information should have been revealed as to how digital time-lapse recordings work and are viewed. I have to wonder where the dealer or integrator might fit into this liability. 

Working With Police Worthwhile

One good source I have found for specifications on commercial video camera placement is from the International Association for Identification (IAI) Scientific Working Group on Imaging Technology (SWGIT). A Google search will lead you to its white paper, “Recommendations and Guidelines for Using Closed-Circuit Television Security Systems in Commercial Institutions” Section 4, Version 2.1 2004.07.22 The paper gives some good examples of video equipment selection and camera placement.

As a final note, I want to repeat an invitation I put out a few years ago to the trade. If you are truly sincere about coming up with the best video evidence for your customers, you may want to pay a visit beforehand to your local police department.

Almost all now have video forensic officers that will be viewing those holdup video recordings from your customers. Many of these officers are also members of a forensic professional group called the Law Enforcement & Emergency Services Video Association (LEVA). You might get some good tips from them and they, I am sure, will be interested in your video experience as well. Let me know how it goes.         

Tech Talk Tool Tip

The lens calculator wheel device is a valuable portable video tool for anyone who is serious about the proper selection of camera lenses with reference to distance, field of view (FOV) and size of object to view. Look for additional tips on this at Security Sales & Integration’s Web site. The best part of getting this tool is it can typically be had for free just by asking major CCTV lens manufacturers. Distributors often have some at the counter as well.

If you are looking for an interesting and very low-cost portable electronic version of the lens calculator, then take a look at the PCam program at < a title='opens new window' href='http://www.davideubank.com' target='_blank'>www.davideubank.com.

This program can run from a Palm™ PDA or with a converter program for your PocketPC such as Style Tap (www.styletap.com). I have used the PCam program in the past and it is pretty slick.

If you enjoyed this article and want to receive more valuable industry content like this, click here to sign up for our FREE digital newsletters!

Related Content