摘要
以汽车为例,研究风格计算自动化的关键基础技术——产品形态特征识别,提出一套框架/模型和实现方法,建立一个比较完善的形态特征识别框架。从风格创新角度出发,结合汽车造型特征,定义3个与汽车特征识别密切相关的关键类集合。基于关键类集合,定义汽车关键特征点集合与局部特征组件集合之间的映射,以及局部特征组件集合与汽车车型集合之间的映射,在此基础上提出汽车形态特征识别模型,并给出关键特征点提取算法、局部特征分割算法以及车型识别算法的示例。最后通过一个实例系统对识别模型和方法做了介绍。
The automobile is taken as an example,the basis of the automatic style computation technology-the automatic segmentation and recognition of the product shape characteristics is studied.From the perspective of style innovation,three key classes which are closely related to characteristics recognition are defined.Based on the key classes,the mapping between the key feature points set and the car assembly set,as well as the mapping between the car assembly set and the vehicle type set are defined.On this basis,a recognition model about automobile shape characteristics recognition and put forward a few recognition algorithm samples are given,including the key feature points extraction algorithm,the car assembly set segmentation algorithm and the vehicle type recognition algorithm.At last,a prototype system about the recognition method is introduced.
出处
《机械工程学报》
EI
CAS
CSCD
北大核心
2011年第10期157-163,共7页
Journal of Mechanical Engineering
基金
江苏省自然科学基金资助项目(BK2004073)
关键词
产品风格
创新设计
汽车识别
关键类集合
Product style Innovation design Automobile recognition Key classes set