Vehicle detecting in traffic scenes with introduction of subtractive clustering algorithm

This paper proposes a novel vehicle detecting approach for surveillance scenes with single stationary camera. Difference accumulative based background modeling method is used for background modeling. Background subtraction operation is used for detecting moving vehicles and Otsu method is used to threshold the background difference image. Subtractive clustering algorithm is applied for vehicle locating in proposed vehicle detecting system. Experiments on real scenes show that the proposed approach is effective for vehicle detecting and subtractive clustering is useful for vehicle locating problems.