Analysis of the objects images on the sea using Dempster-Shafer theory

Abstract

The paper presents the concept of using aerial and satellite imagery or images coming from the marine radar to identify and track vessels at sea. The acquired data were subjected to a highly advanced image analysis. The development of remote sensing techniques allows to gain a huge amount of data. These data are useful information source however usually we have to use different data mining methods to gain interested information. In this paper there is described a special approach based on specific approach coming from advanced image analysis method created for the purpose of tracking objects. Unfortunately, object recognition using image analysis involves some level of uncertainty. Therefore, in the final identification process we have to take into account the belief and the plausibility functions.