TARGET IDENTIFICATION

HAVELSAN develops applications for exploitation of satellite images, which can either be performed manually by the operator within the scope of remote sensing or automatically in forms of operator-interactive or operator-non-interactive. The ultimate goal of remote sensing is classification of the related image or content and marking of objects automatically or by the user.

In this regard, satellite images and aerial photos are classified for identification of natural areas and urban centers. These categories can be such as earth, green areas, residential areas, seas, lakes, rocky areas and agricultural areas.

Manual extraction of road network is a time-consuming operation. HAVELSAN is capable of generating a high accuracy road map automatically with the algorithms it has developed. Objects such as buildings are also important in remote sensing activities.

HAVELSAN has developed high performance algorithms which automatically generate subparts of the regional targets such as airports, harbors, military facilities, dams etc.
In the following sample figure, remotely and automatically sensed buildings (Targets) can be seen:

Algorithm results that find similar target types, of which samples are automatically selected by the user. When an aircraft example is selected others can be found or when a forest is given as example, whole borders of forestland can be classified and marked with the semi-automatic/interactive method.

Photos