Key words
vision systems, noncontact measurements, image processing, aerial reconnaissance, unmanned aircrafts
articles
The author:
Roshchin D. Published in
№ 5(71)
31 october 2017 year
Rubric: Models and Methods
Rubric: Models and Methods
The method of vision systems application for contactless measurements of objects on the example of decision of aerial reconnaissance
The continuous development of computer technology opens more opportunities for its application
in creating smart and effective computer vision systems (VS), aimed at solving problems related to
digital image processing and object recognition.
VS is especially needed in the areas of science and technology requiring the use of contactless methods
for measuring various parameters of the objects. The main task for the improvement of VS is the
improvement of existing and development of new methods for facilitating the expansion of functionality
of these systems. The urgency and relevance of VS are considered on the example, which presents
a model of conducting aerial reconnaissance.
Tactical unmanned reconnaissance aircraft makes for the departure shot of a certain area. The shooting
is done at height and with the resolution necessary for confident recognition of targets like «tank»
or «car». The area of goals that you want to detect and recognize then on the received intelligence, is
very small compared to the whole area of the earth’s surface, entering the frame. It was found that on
analysis of the entire survey, by visual inspection of time you will need many times more than it spent
on its implementation. Thus, the main problem lies in the fact that the volume of intelligence information
obtained through imagery intelligence is huge and it requires considerable time for processing.
Expediting the processing of intelligence information is only possible by parallelization of the
processing and/or automation using the technical means like VS.
Accordingly, it is necessary to the development of contactless measurement methods based on the
methods and means aimed at the solution of similar tasks that require processing of large amount of
optical data over a short period or in real time.
