Verification of Characteristics of Gaussian Filter Series for Surface Roughness in ISO and Proposal of Filter Selection Guidelines  

作　　者：Yuki Kondo Ichiro Yoshida Dan Nakaya Munetoshi Numada Hiroyasu Koshimizu 

机构地区：[1]HOSEI University,3-7-2 Kajino-cho,Koganei-shi,Tokyo,184-8584,Japan [2]Chukyo University,101-2 Yagotohonmati Showa-ku,Nagoya,Aichi,466-8666,Japan [3]YYC Solution,3-102-3F Kamiyashiro Meito-ku,Nagoya,Aichi,465-0025,Japan

出　　处：《Nanomanufacturing and Metrology》2021年第2期97-108,共12页纳米制造与计量（英文）

基　　金：We would like to thank Editage(www.editage.com)for English language editing.

摘　　要：In surface roughness measurement,if spikes are included in the primary profile,a problem occurs wherein the Gaussian filter(GF)is unable to extract the shape components.To address this problem,the use of a robust filter is proposed.However,ISO16610-31:Gaussian regression filters(GRF)only provide a single method and a few examples,and does not specify the conditions under which the primary profile can be covered.Moreover,the data presented in the example on robustness in ISO16610-31 do not contain roughness components.In actual roughness measurements,no primary profile exists that does not include a roughness component.Because the characteristics of GRFs are unknown,it is not yet clear which filter should be used for which primary profile,and this is an issue that has been raised at ISO and JIS conferences.In addition,the establishment of filter selection guidelines is necessary at measurement sites.Therefore,this paper clarifies the characteristics of GF-series filters,summarizes the points to be considered when using them,and identifies the filter that should be selected according to different situations.Based on the results,a figure that visualizes the characteristics of filters and a flowchart regarding which filter should be used are created;these tools,to the best of the authors’knowledge,did not exist prior to this study.It is believed that these results will help fulfil the needs of measuring job sites and also aid in filter selection.

关 键 词：Surface roughness Gaussian filter Robust filter M-estimation method Fast M-estimation method 

分 类 号：TN7[电子电信—电路与系统]