Title :
Vehicle classification based on the fusion of deep network features and traditional features
Author :
Hua Qian ; Yaying Zhang ; Chunmei Liu
Author_Institution :
Key Lab. of Embedded Syst. & Service Comput., Tongji Univ., Shanghai, China
Abstract :
Consider to the defect of traditional features which have high empirical components. A vehicle classification algorithm based on the fusion of higher-layer features of a deep network and traditional features was proposed. Firstly, the traditional features of PHOG and LBP-EOH were extracted. Secondly, the higher-layer features excavated from the vehicle pictures by deep belief networks were added, making these three kinds of features together by feature fusion. Finally, support vector machine is used to train and classify the vehicle. When the number of training samples is large enough, the algorithm has a significant effect compared to those with traditional features. It can achieve the accuracy of 95% in the six categories of vehicles.
Keywords :
belief networks; feature extraction; image classification; image fusion; road vehicles; support vector machines; traffic engineering computing; LBP-EOH feature; PHOG feature; deep belief networks; deep network feature fusion; feature extraction; support vector machine; traditional feature fusion; vehicle classification algorithm; Algorithm design and analysis; Feature extraction; Support vector machines; Vehicles; Deep Belief Networks; LBP-EOH; PHOG; Support Vector Machine; Vehicle Classification;
Conference_Titel :
Advanced Computational Intelligence (ICACI), 2015 Seventh International Conference on
Conference_Location :
Wuyi
Print_ISBN :
978-1-4799-7257-9
DOI :
10.1109/ICACI.2015.7184788