Key Factors for High Quality Surveillance Video

Getting high quality surveillance video can be challenging. Indeed, many reports or surveys indicate that most users are not happy with the quality of their surveillance video.  This report identifies the key factors that impact producing high quality surveillance video.

The traditional three (well known factors) are:

• Providing sufficient Image Detail, or Resolution
• Handling low light or night-time
• Handling wide variance in lighting like bright sunlight coming through a front door

Image Detail

Image detail, or resolution, is the most well known of the factors impacting video surveillance quality.  Most viewers can readily see the difference in increasing quality from youtube videos to VCR tapes to DVDs. The same pattern exists in video surveillance with video resolution ranging from as low as CIF (320 x 240 pixels or .08 megapixels) to 5 megapixel (2560 x 1600 pixels) or higher.

The move to megapixel cameras offers the most important path to increasing image detail.  Increased image detail can reduce the total number of cameras deployed, saving costs as well as providing evidence impossible with traditional cameras. See a demonstration of megapixel camera abilitites.

Low Light/No Light

Surveillance cameras routinely need to operate in conditions with low or no light. This is very common in security as surveillance cameras are used to spot intruders attempting to attack a facility.

A variety of night vision cameras are used to see when natural or artificial light is missing. The four major types of night vision cameras (in order of cost) are infrared (uses lamps emitting invisible light e.g., Extreme CCTV), SWIR (does not require any illumination as it uses the Earth's own glow e.g.,NoblePeak), thermal (uses heat  e.g., Flir) and laser-based (generates a targeted beam of light e.g., Vumii).

Wide Variances in Light

Even when it's daytime and plenty of sunlight is available, variances in the amount of light can cause significant problems. The most well known example is opening your front door. A flood of light enters behind you.  This can be a significant issue for surveillance cameras. Less dramatic but also important is light coming from windows as well as the placement of lamps and light bulbs within a facility.  All of these can create significant problems for cameras.

The most common solution to this is the use of wide dynamic range cameras. Historically, cameras adjusted the amount of light for the entire scene. You could reduce or increase the amount of light but it equally impacted the whole image. With wide dyanmic range cameras, different parts of a single image can receive different levels of light. This means that a dark area can receive more light making the image clearer while a very bright area can receive less light so it does not look washed out.  The most well known providers of WDR cameras are Pixim and the Panasonic SDIII.

Restrictions on Camera Placement

When you are taking someone's photograph you can pretty much position the camera anywhere you want - on the ground, on a ladder, from a side angle, etc.  On the contrary, surveillance cameras can usually only be placed in very limited places.

Here are the main issues:

- Aesthetic constraints: Most organizations want the cameras to be incospicuous which usually means they need to be high and out of the way. Unfortunately, high and out of the way usually produces poorer quality images because of the steep angles this creates.

- Operational constraints: Since cameras need to be permanently mounted, they need to be put in places that do not interfere with cabinets, bookshelves, doors, etc. They also cannot be placed in places which are hard to reach or hard to run networking/power because of the financial costs.

Restrictions on Lighting Additions

A professional taking photographs can use a flash or bring lights. This is rarely possible with video surveillance.  Most organizations do not want to add new lighting indoors or outdoors as the expense is high.  Where you most likely see lights being added are outdoors but that is only when there is no street lights or the street lighting is bad.

Issues that could be 'simply' solved with the addiiton of aritifical light and generally not feasible for video surveillance.

Cannot Restrict How People Move

With a photograph, you can have the subject tilt their head or move left or right.  With surveillance video, not only can you not have the subject/suspect move, you rarely can constrain the flow of traffic.

Unless you are a prison or similar type environment, most organizations do not want to restrict movement or change facility layouts to accomodate surveillance cameras.

As such, depending on how people tilt their head or one on what side of the hallway they walk, the image can look significantly better or worse.

Conditions Change 24/7

With a photograph, the camera can be manually optimized for the specific situation then and there. You can adjust your shutter speed, the size of your aperture, the speed of your film, etc.  All of this can precisely optimize the shot for that specific moment.

However, with a surveillance camera, the lighting issues you have at 8 am are often completely different than at 3pm or at 7pm. You can go from frontlight to sidelight to backlight to no light as the day progresses. It is very difficult to get a camera to instantly and continously adjust to such a wide variety of lighting conditions.

Cost Constraints are a Big Deal

When taking photographs, the cost of equipment (even if it's thousands of dollars) is inexpensive relative to the photographer's time. Video surveillance is the opposite. It takes minimal ongoing labor cost to maintain a video surveillance system but the equipment cost is very significant.

Using a better lenses or superior camera, could cost an organization $100 - $600 for each camera. With systems today of hundreds to thousands of cameras, this results in significant increases in cost.

While the quality of images is improving (through megapixel cameras, night vision and WDR cameras), video surveillance faces a number of fundamental challenges far greater than that of photography. These challenges may diminish but it is unlikely that surveillance video will ever come close to the quality of the films and photography we are used to seeing on TV and in print.