基于Hough变换的实验数据坏点剔除方法

A Method of Experimental Defect Rejection Based on the Hough Transform

在实验测试中,除了获得包含误差的正常数据,实验员常常也可能观察到一些明显错误的点,我们称之为"坏点"。坏点的存在可能会严重影响实验最终结果的可靠性,所以在数据分析前,应该被处理掉。以往的常规坏点处理方法多为肉眼查看,手动排除。这种方法工作量大,并且判断依据不够明确,只靠"感觉",另外,在接收到大量数据的时候,手动法无能为力。本文在使用计算机编程的基础上,应用Hough变换,提出了一种坏点剔除的新方法,可处理符合直线拟合特征的实验数据中的坏点;作为扩展应用,该方法也可以处理以下两种数据:可通过函数变换转化为符合线性分布的数据和Hough变换可处理的服从曲线分布的数据。仿真实验和实际应用表明,本方法具有较好的性能。

In experimental tests, apart from obtaining the normal data with allowable errors, the experimenters usually get some unexpected wrong data called ＂defect marks＂. Following the routine method of experimental data processing, the method of bad point exclusion based on automatic programming is seldom taken into consideration by experimenters. This article presents a new method based on the Hough transform to reject bad points. The method is fit for processing data with linear characteristics and can be extended to deal with the data that is possible to be translated into a linear form through functional transformations; curved lines, which can well be processed by the Hough transformation, can be its application too. Simulation experiments and practical applications manifest that the method raised in this paper performs robustly.