3.12. Cameras and Video input. 3.12. Cameras and Video input. The image is focussed onto a photosensitive surface (such as a charge coupled device (CCD)) line by line to accumulate entire image. The CCD produces current (or simply digital output in newer systems) which is proportional to light intensity (filtered for each of three colour bands).
Different quality/functionality cameras are used:
Monochrome: offers users extremely high resolution (1024x1024 pixels)
full-frame async dynamic image which uses a progressive scan interline
transfer imager for outstanding full-frame shutter and integration
characteristics shutter speeds up to 1/16000 sec. may be achieved.
Colour: high resolution cameras that offer the user outstanding color
fidelity and full, broad-band response for color applications such
as teleconferencing, machine vision, object recognition,
medical research, process monitoring airborne imaging and remote
observation. Typically 768 (h) x 494 (v) with high speed variable
shutter (to 1/10,000 sec), and auto/manual white balance.
Smart: The imputer is a camera and programmable machine vision
processor combined in a miniature package that is less than the
size of a standard CCD camera. Designed for machine vision
applications, this seamlessly-integrated architecture replaces the
uneconomic and unwieldy combination of a PC + Frame Grabber +
Camera + Processor for each vision node in an inspection system.
The video signal must be digitised to some standard format
such as PAL or NTSC (or some required resolution). This is done
via a frame store or frame grabber. This will sample the incoming
signal at discrete intervals to produce the "pixels" taking into
account the horizontal and vertical resolution of the signal.
These samples are then formatted appropriately (number of bits,
colour, etc) and stored in the frame buffer which can then be
read into normal computer memory for saving as a frame.
Fig. 3.20 : Video capture from IndyCam
