Applications designed to detect movements work with a
huge volume of image data that may disproportionately
increase the demands on the hardware configuration. Hence
the need to continuously improve their methods of reducing
the volume of image data is necessary for processing video
recording. Any saving, though small, may appear in one
image as insignificant item, but for the sum of thousands of
images that make up the video footage, it has very positive
effects. Problems in motion are not only caused by a huge
amount of data, but also by the enemy of image quality,
which is noise. This enemy contributes to false positives,
namely, the erroneous output. Usually, the sensitivity level
is set because of the reasons of existence of noise, when the
detector begins to consider changes in the image as a
movement. Generally, the greater the noise, the greater the
likelihood of false detection, and thus it is also necessary to
set the sensitivity level. This, of course, can lead to the fact
that the sensitivity level is set at a threshold where certain
types of movements are detected as noise.