Data fusion methods based on fuzzy measures in vehicle classification process

Abstract

The paper presents results of analysis and properties comparison of five different data fusion methods in process of vehicle classification. The fusion process has been realized on basis of signals features. The used signals comes from inductive loop and piezoelectric sensors placed in surface of the road. The models of vehicle classes have been defined by using fuzzy measures with triangular and gaussian shapes. The paper presents the construction method of such models and advantages of data fusion methods.