摘要
提出了一种在固定摄像头拍摄的车辆图像序列中检测出运动车辆并识别其车辆类型的系统结构和基于Sobel边缘细化提取车型特征的新方法,本算法处理过程分为数字图像获取、图像处理和车辆外观特征数据提取三个主要部分。现场实验表明本方法实用可行。
The systematic architecture based on serial videos and a fixed camera was suggested. A new algorithm to detect the traits of cars based on Sobel edge-subsection was also proposed. The algorithm consisted of digital image acquisition,image processing and vehicle exterior data extraction, i. e three main sections. Experimentation on site was given to have shown the method feasible.
出处
《辽宁工业大学学报（自然科学版）》
2010年第1期5-7,共3页
Journal of Liaoning University of Technology(Natural Science Edition)
关键词
车型识别
图像处理
二值化
边缘检测
vehicle type recognization
image disposal
binarization
edge detection
