6. Analysis Of Resultes
In this research, the image of forest region after fire occurrence is used. First, the
burnt region is extracted by binarization and then, the noises inside the burnt region and
surrounding it are reduced as much as possible by several median filters which had
various dimensions. Also, for better assessment and evaluation of the primary area of
the burnt region in the forest (in the selected image) in a machine vision system, two
method are implemented which the values of areas (in pixel) are presented in table 1;
Table 1. The comparison between the primary areas
NO. Method Primary area (in pixels)
1 Integration 77689
2 Interconnected components extraction 77922
As it is obvious, table 1 shows the difference between two obtained areas as equal to
233 pixels which is almost negligible