Image Processing Reference
In-Depth Information
Fig. 2.20 A gray-scale image and part of the image described as a 2D array, where the cells
represent pixels and the value in a cell represents the intensity of that pixel
per second. That is quite a number even for today's powerful computers. So when
you choose your camera do not make the mistake of thinking that bigger is always
Besides picking a camera with a reasonable size you should also consider intro-
ducing a region-of-interest (ROI). An ROI is simply a region (normally a rectangle)
within the image which defines the pixels of interest. Those pixels not included in
the region are ignored altogether and less processing is therefore required. An ROI
is illustrated in Fig. 2.21 .
The ROI can sometimes be defined for a camera, meaning that the camera only
captures those pixels within the region, but usually it is something you as a designer
define in software. Say that you have put up a camera in your home in order to
detect if someone comes through one of the windows while you are on holiday. You
could then define an ROI for each window seen in the image and only process these
pixels. When you start playing around with video and image processing you will
soon realize the need for an ROI.
Further Information
As hinted at in this chapter the camera and especially the optics are complicated
and much more information is required to comprehend those in-depth. While a full
understanding of the capturing process is mainly based on electrical engineering,
Search WWH ::

Custom Search