基于机器视觉的鞋楦数字化及类似方法对比  

Shoe-last measurer based on machine vision and contrast with other digitizers

作　　者：罗胜[1] 

机构地区：[1]温州大学机电工程学院,浙江温州325035

出　　处：《工程设计学报》2007年第1期57-61,共5页Chinese Journal of Engineering Design

基　　金：上海市科技攻关重大专项基金资助项目(02D211019)

摘　　要：介绍了机器视觉的原理和目前主流的鞋楦数字化方法,讨论了机器视觉的研究发展现状、优缺点及技术上的瓶颈;重点将接触式测量、激光扫描和立体视觉这三种方法进行了对比.接触式测量可靠性高,但是精度低.激光扫描法测量精度相当高,可测量柔软易脆对象,不必作测头半径补偿,工作距离大、范围广,但是容易受物件表面光学性质影响,设备精密,价格昂贵.立体视觉方法也可测量柔软易脆对象,速度快,硬件简单,价格低廉,并且不会抹平模型上的尖点.立体视觉方法是目前正在发展的一种图像测量方法,同样能完成鞋楦数字化,与前两种方法相比,在鞋模的定制方面更为合适.Machine vision principle and currently prevailing last-measure method are introduced. Research situation of machine vision and its advantages and disadvantages as well as its bottleneck in technology are discussed. This paper focuses on the contrast between contact measure, laser scanning, and machine vision. Contact measure has high reliability but low accuracy. Laser scanning has high measuring accuracy and can measure soft and fragile targets; it does not need to compensate the radius of measure head and can cover a larger range. However, laser scanning is easy to be affected by surface optical properties of parts and its price is high. Machine vision can also measure soft and fragile targets; it has high speed, simple hardware and low price and will not smooth the sharp points of the model. Machine vision is a developing method to measure, and can complete digitization of shoe-last. Comparing to the former two methods, it is more suitable for customization of shoe-last.

关 键 词：机器视觉 接触式测量 激光测量 鞋楦 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]