Vehicle detection and classification in the Scala sensor by using binary classification

Abstract

In this paper, we present approach for the detection and classification of multiple vehicles using a vehicle mounted laser scanner. The sensor, which is placed in front of the vehicle, is Scala 1403 and its characteristic is that it scans data twice -upper and lower direction- in one period. Consequently, it is difficult to detect and classify vehicles continuously. For solving this problem, another method for classification is needed. In this paper, binary classification is proposed for classification of Scala sensor. It can show better classification result than by using SVM (support vector machine) in case of occlusions. Experimental results carried out with laser range data illustrate the robustness of our approach.