Vehicle Feature Extraction and Application Based on Deep Convolution Neural Network Zhang Hua*, Liu Xiang** College of Electronic and Electrical Engineering, Shanghai University of Engineering Science, Long Teng road 333#, Songjiang District, Shanghai, 201620, China *zhanghuashine@163.com
**2509020608@qq.com
Online published on 2 July, 2018. Abstract In view of the existing vehicle image features extraction of large amount of calculation, slow speed, extract the characteristics of the uniqueness is not strong, the complex extraction process and so on, is put forward based on the depth of the convolution of the neural network vehicle feature extraction method, and is applied to extract the depth of the characteristics of vehicle brand recognition and deck vehicle identification. The experiment shows that the proposed method can extract the depth characteristics of the vehicle effectively, and obtain good vehicle identification and identification accuracy. Top Keywords Feature extraction, Deep learning, Fake plate vehicles, Vehicle recognition. Top |
|
Access denied
Your current subscription does not entitle you to view this content or Abstract is unavailable, the access to full-text of this Article/Journal has been denied. For Information regarding subscription please click here.
For a comprehensive list of other publications available on IJour.net please click here
or, You can subscribe other items from IJour.net (Click here to see other items list.)
Top