Quadrox WebCCTV, Installation Manual

Page: 61 / 90

WebCCTV Installation Manual
61
Version 4.3 Series
5.2.2 Resolution
When the compression technique is based on JPEG images, a second big factor in data size
reduction is the resolution. As mentioned before, compression works by selectively reducing
the amount of information in the image. One way to do that is to reduce the size of the image.
Example :
A VHS quality image (PAL) of 384x288 pixels can be considered as a SVHS+ image of
768x576 pixels where you throw away every other pixel in either direction. This reduces the
data size by a factor of 4.
Of course, this influences the quality of the image. Especially fine details are more likely to get
lost, like face characteristics or details of clothing.
The resolutions that are mentioned, except for the mega-pixel, are for the PAL
standard. NTSC images and pure digital images might have a slightly different
resolution, but in general the same steps/magnitude apply.
Postage stamp format (192x122). This resolution is offered by some of Quadrox competitors.
At this image size however, most of the information in the image is lost. At best, one would be
able to count the number of people in the image, but there is no hope for e.g. identifying faces.
VHS quality (384x288). Small and thus favourable for storage and transmission, this
resolution gives a good quality in most cases.
SVHS quality (768x288). In this image format, only half of the image is lost. The image is
only decimated in the vertical direction. For analogue based cameras, this is the maximum
resolution possible without motion artefacts (see below). Best choice for an analogue camera.
Notice that this resolution doesn’t have the classical 4:3 proportions. This can be compensated
for, but some media players might display a “squished” image.
SVHS+ quality (768x576) : This is the maximum resolution for all analogue and most network
cameras currently in the market. Most detail is preserved.
This resolution has a big disadvantage though. Because of the video sensor technology inside
the camera (interlacing), the image shows horizontal lines at the edges of moving objects. This
makes the object (e.g. a face) unrecognizable in most cases and as such seriously diminishes
the usability of the camera.
For more detailed technical explanation about interlacing and its consequences,
please contact Quadrox support.
Megapixel quality (1280x960 and higher) : Network cameras with more than a million pixels
are becoming the new type of camera’s in today's marketplace. Because of their high pixel
count, they can preserve much more detail than a regular camera, making them useful in many
circumstances. This comes at a very considerable storage and network bandwidth cost!
As a final remark on resolution, we should reiterate that resolution has quite a different effect
when using inter-frame codecs for compression (MPEG, WMV). In these cases, choosing a
smaller resolution means less quality (as above), but not less data, since the bit rate is fixed.