点击次数：

论文类型： 期刊论文

发表时间： 2018-04-01

发表刊物： MACHINE VISION AND APPLICATIONS

收录刊物： SCIE、EI

卷号： 29

期号： 3

页面范围： 467-476

ISSN号： 0932-8092

关键字： Forging; Feature lines; Rapid extraction; Thermal geometric parameter measurement; Computer vision

摘要： The online dimensional measurement of large hot forging is an important procedure in the forging process. Because of different production demands, the final geometrical shapes of large forgings are usually different (e.g. cylindrical columns, rectangular prisms). Forgings of different geometric shapes need to be measured along different dimensions. For rectangular forgings, the lengths primarily need to be measured. A generalized measurement system for different geometric shapes of forgings cannot provide the accuracy of measurement systems targeted at measuring a known shape. Based on the characteristics of the rectangular forgings, a thermal dimensional measurement system is proposed in this paper. The localization, rapid extraction of feature points and method for measuring the dimensions of rectangular forgings are presented. The proposed methods can easily and efficiently extract the feature points of the forging. The experiment results show that the method proposed in this paper has the advantages of high precision and high efficiency, which is appropriate for online measurement.