Abstract: One of the most popular areas of unmanned aerial vehicles (UAV) use is the monitoring of
roads and highways. UAVs are considered to be a low-cost and rapidly growing platform that can
provide effective mechanisms for data collection and processing, especially in case of long distances.
In this paper a new method for automated road monitoring with the help of UAV is offered. The method
is suitable for determining image areas, where the heterogeneity compared with the general road
structure is spotted. Also, algorithms for finding road cover from video shots and for determining the
direction of the road are offered. The method is applicable for automated control of UAVs in order to
find and track roads, as well as registration of various types of objects on the road. All the methods and
algorithms were tested on a model and the results are shown.
Keywords: image processing, road tracking, road monitoring, similarity measure, UAV
ACM Classification Keywords: Image Processing and Computer Vision
TECHNIQUE FOR ROAD AUTOMATED TRACKING WITH UNMANNED AERIAL