摘要
在灰色理论基础上提出了一种自动车型识别与流量预测技术。该技术实现可分为三个步骤:第一步,测量车辆的三维信息,获得车辆的长、宽、高度特征;第二步,利用灰色关联分析对车辆进行分类识别;第三步,建立一阶单变量车辆预测模型即GM(1,1),用于有关部门统计指标的预测。实验表明,该方法在车型识别中,具有比较高的识别精度,而灰色预测模型较传统的预测方法更具科学性与实用性。
A technology of vehicle model recognition and traffic forecast based on gray system is introduced. The tech- nology could be divided into three major steps： firstly, measuring the three-dimensional information of vehicle, obtain the features of length, width and height. Secondly, classifying and recognizing vehicle by gray correlation analysis, thirdly, es- tablishing a first-order single variable traffic forecast model, which is gray early warning model （GM（1,1））. Experiment result indicated that the novel method obtained high identification precision in vehicle model recognition, and compared with traditional forecasting methods, the GM（1,1） is more scientific and practical.
出处
《计算机与数字工程》
2010年第2期130-135,共6页
Computer & Digital Engineering
关键词
车型识别
灰色理论
灰色关联分析
GM(1
1)
vehicle model identification, gray theory, gray correlation analysis, GM（1,1）
