Wednesday, 29 October 2008


The future for CCTV ... Mega Pixels or Mega problems?

Apologies in advance for the somewhat sensationalist heading to this blog post, but I've been pondering recently the accelerating transition from conventional analogue (frequently described as legacy) CCTV cameras, towards 'how did we ever manage without them' digital models.

You'll perhaps have to forgive me for being ever so slightly flippant with that last comment, but as time and again I come across apparent CCTV "experts" promoting the quantum leap into the wonderful world of MegaPixel technology, as perhaps the best thing since crimp connectors, the realisation for me that things may not be quite so hunky dory has come as a bit of a shock.

Increasingly I find the suggestion that fitting extremely wide angle lenses as a possible killer application for MegaPixel cameras, unfortunately overlooks a glaringly obvious problem with what should be in the great scheme of things, a revolution in imaging technology. 

For the benefit of those unfamiliar with the concept of MegaPixel imaging, the theory is actually very simple, so I'll likewise try and explain the issues as straightforwardly as possible.

An everyday standard analogue or IP Video camera may produce an image resolution (horizontal res.) of let's say for arguments sake, perhaps 480 lines.

Now a jobbing 2 MegaPixel camera can normally produce a resolution which is roughly four times that of our conventional example, so unlike a normal camera, the MegaPixel image is actually  sufficiently sharp to allow a digital electronic zoom function to be used. This will allow the operator to examine a selected area of the picture for more detailed examination, without having to use the standard approach of increasing magnification with an optical zoom lens.

So far so good, but then recently I've noticed a growing tendency for some practitioners to promote the idea that the MP cameras are so sharp, you can easily fit an ultra wide angle or even borderline fisheye type lens, to cover extremely large areas with just the one camera.

Now this practice may seem o.k. on paper, but then the realities of trying to be all things to all men suddenly becomes somewhat less attractive, when the disadvantages are carefully considered.

Firstly, if you compare the performance of a normal camera fitted with a "standard" lens, against a 2MP camera fitted with an ultra wide angle optic, whilst the latter may cover a much greater area (and at a significantly greater cost), if you zoom in electronically to pick out a face from the crowd, the actual size of the persons head at a set distance, may well be equivalent to ... or dare I suggest even less than, the normal set up - so in this simple example, apart from covering a much wider area, the technical quality advantage presented by a much higher resolution camera, is to all intents and purposes completely lost.

Now the next problem relates to the difference between the two lenses. 

With a "standard" lens, a person can move away from the camera a few metres, and the reduction in their head image size on screen will in general terms be minimal. However, with the MegaPixel camera, if it's fitted with an ultra wide angle optic (i.e. a very short focal length lens), any movement away from the camera will result in a very rapid reduction in the size of the person on screen.  

So, in order to provide an equivalent size image, a greater degree of digital zoom would be required, which will effectively cause a reduction in image quality (probably through increased "pixellation"); so regrettably, in a purely practical sense, as a technical solution it may actually be far less effective for providing those vital evidential quality images.

This may all sound a bit strange compared to what many would have us believe, but the bottom line as with any video imaging technology, is that the more you ask a camera to do, then generally speaking the less efficient it becomes.

It would be a shame if all the potential quality improvements that are inherent using MegaPixel systems, were thrown away in the rush to adopt the technology in any situation whether optimised or not.

That said, if ever there were a clear case for the uninitiated of caveat emptor (buyer beware), then this would be as good an example as I could reasonably think of .... 

Labels: , , , , , ,

Comments: Post a Comment

Subscribe to Post Comments [Atom]

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]