摘要
运用深度卷积神经网络中的AlexNet模型,对在各个角度下拍摄的具有复杂背景的汽车图片进行网络训练和测试,以对车辆进行识别和检测。在进行识别检测时,运用了基于深度学习的卷积神经网络,它不仅大大提高了识别准确率,同时在特征提取方面也由于其他传统算法,在车辆定位及分类识别领域具有广阔的应用前景。
This paper uses the AlexNet model in deep convolutional neural network to conduct network training andtesting on car images with complex backgrounds taken at various angles to identify and detect vehicles. In the processof identification detection, a deep learning-based convolutional neural network is used, which not only greatly improvesthe recognition accuracy, but also has broad application prospects in the field of vehicle location and classificationidentification than other traditional algorithms in feature extraction.It has broad application prospects in the field ofvehicle location and classification.
作者
贾瑞
Jia Rui(College of Electrical Engineering,Suzhou Chien-shiung Institute of Technology,TaicangJiangsu 215400)
出处
《现代工业经济和信息化》
2018年第12期24-26,30,共4页
Modern Industrial Economy and Informationization
基金
苏州健雄职业技术学院"三级联动"科研基金项目"基于微信公众号的农作物病虫害识别系统开发研究"的阶段性研究成果
项目编号为2017SJLD20
关键词
车辆检测
车型识别
深度学习
卷积神经网络
vehicle detection
vehicle recognition
deep learning
convolution neural network