ation 
ing 
18 
essing 
A 
4 
the 3.0 
ation 
th 
ie of 
las 
tions, 
ing , 
d 
n be 
ime 
or 
E ox ow 
mss cem 
With the availability of very high speed RAM refresh memories, it has 
now been possible to build large data bases of 4096 x 4096 pixels 
which can be partitioned and randomly accessed by multiusers for 
different applications. One of the most significant aspects of such 
System is intheir architecture design and feedback path that allows 
the actual viewed imagery to be read back into the data base at TV 
rates, The feedback also allows interative processing. Use of pipeline 
processors, image manipulation, and processing circuits, all updata- 
ble at video rates, allow a number of image processing techniques 
to be carried out in realtime. Another significant aspect is the use of 
firmware which allows turnkey operations which are the need of bulk 
data processing. In true sense, these are parallel array processors 
where à cycle time is 1/25th of à second and the array is 512 x 512 
pixels x 8 bit deep. 
Because of Several advantages ; these systems have found widespread 
uses in thermal mapping, hydrology, forestry, agriculture, weather 
prediction and analysis, non-destructive testing, quality control and 
inspection, X-ray fluroscopy, and a host of other fields. 
REAL TIME VS OFF-LINE VIDEO IMAGE PROCESSING SYSTEMS 
A real time image processing system is required when (i) it is 
necessary to make an important decision faster; (ii) raw image from 
sensors or input device is inadequate in some way and requires 
corrections; (iii) input image is changing; and (iv) it is necessary to 
eliminate delays caused by off-line computer or film processing. 
For instance, a standard frame of 512 x 512 pixels in a video image 
contains 262,144 picture elements (pixels). This amount of informa - 
tion is so large that most computers do not accept through a single 
input (I/O) channel. Although there is an infinite amount of flexibility 
in computer processing, speed of operation is limited. 
Thus, a video-rate image processor with limited flexibility will 
process this massive information in just 1/25th of a second while most 
computers would need much longer time. In essence, a real time 
image processing provides the following advantages: 
(i) It allows an important decision to be made faster. 
(ii) Once some specific image processing technique has been 
selected, it can be used continuously or repeatedly as part of 
overall process. 
(iii) Before a particular function has been optimized, it is easy to 
experiment with various processing functions while viewing 
the results on a display. Thus, a technique can be easily 
optimized by interaction between the user and the processor. 
(iv) Because of easy user-processor (man-machine) interaction, 
it is easy to continuously optimize processing functions as 
external conditions change. 
109 
m ———À