点模式匹配在羽毛弯度测量中的应用

Application of Point-Pattern Matching in Camber Measurement of Feather

点模式匹配是计算机视觉和模式识别领域中的重要课题,点模式匹配有效地避免了图像匹配计算量很大,匹配效率、匹配精度低和匹配时间长,不利于工业在线测量的缺点;采用了优化搜索路径的逐步匹配的点模式匹配方法进行羽毛弯度测量,根据不同测量精度要求进行算法上的调整,有效地提高了羽毛的匹配速度以达到提高识别速度的目的,增强了机器视觉在工业产品识别的使用价值,在羽毛自动分拣系统上实验证明,这种优化的点模式匹配算法是可行的。

Point-pattern matching is an important problem for the fields of computer vision and pattern recognition. Point-pattern matching have avoided effectively the disadvantages of conventional matching, such as large amount of computation, low effect of matching and precision, it is disadvantage to on-line measurement in industry, This paper adopts the method of matching step by step with optimal searching path to measure camber of feather, on the basis of different precision, it adjusts different arithmetic, speeds up the identification to feather and boosts up use of machine vision. Experiment results in automatic feather selecting system proved that this method is practicable.