摘要
车辆进气格栅区域具有显著纹理特征,因此格栅区域特征可用于车型自动识别。本文提出一种车辆进气格栅区域的定位方法,并对格栅区域进行特征提取,为解决车型自动识别问题提供思路。首先,使用水平梯度的水平投影确定进气格栅区域的上下边界,然后,利用水平梯度和垂直梯度的垂直投影确定进气格栅区域的左右边界;最后,利用灰度梯度共生矩阵和Gabor小波变换方法提取进气格栅特征。仿真实验结果表明,本文提出的方法能够准确地定位进气格栅区域,基于灰度梯度共生矩阵和Gabor小波变换方法提取的格栅纹理特征能够有效地应用于车型识别。
The area of vehicle grille has remarkably texture feature,so the grille features can be used for automatic identification of vehicle types. In this paper, a method for locating the area of the vehicle grille is proposed, and the features of the grille area are extracted. Firstly, the horizontal projection of the horizontal gradient is used to locate the grille boundary of top and bottom,and then, the left and right boundary of the grille area is located by vertical projection of horizontal and vertical gradients. Finally, the features of the grille are extracted by using the gray gradient co-occurrence matrix and the Gabor wavelet transform. The simulation results show that the proposed method can precisely locate the grille area,the grille feature based on the gray gradient co-occurrence matrix and the Gabor wavelet transform can be effectively applied in identification of vehicle types.
出处
《电脑知识与技术》
2016年第1X期191-194,共4页
Computer Knowledge and Technology
基金
国家自然基金(71203229)
2014北京市支持中央高校共建项目公安视听专业建设
中国人民公安大学2015年中央高校基本科研业务费专项资金资助项目(2015JKF01245)
关键词
车型识别
灰度梯度共生矩阵
GABOR小波变换
格栅定位
纹理特征
vehicle type identification
gray gradient co-occurrence matrix
Gabor wavelet transform
grille localization
texture features
