Robust cell wall recognition of laser measured honeycomb cores based on corner type identification. (January 2021)

• Record Type: Journal Article
• Authors: Dong, Zhigang; Qin, Yan; Kang, Renke; Wang, Yidan; Sun, Jiansong; Zhu, Xianglong; Liu, Yao
• Is Part Of: Optics and lasers in engineering. Volume 136(2021)
• Publication Date: 2021-01
• Subjects: Measurement of honeycomb core -- Cell wall recognition -- Corner type identification -- Discontinuous surface measurement -- Measurement data processing -- Honeycomb feature recognition; Lasers in engineering -- Periodicals; Optical measurements -- Periodicals; Optics -- Periodicals; Lasers en ingénierie -- Périodiques; Mesures optiques -- Périodiques; Optique -- Périodiques; 621.36605
• DOI: 10.1016/j.optlaseng.2020.106321
• Languages: English
• ISSNs: 0143-8166

Abstract

Highlights: The cell wall recognition rate reaches above 98%. The method is robust to burr interference and honeycomb hexagonal shape deformation. A corner type identification method is proposed. Abstract: Scanning measurement with laser displacement sensor is an effective method to measure the surface precision of honeycomb cores. Due to the special discontinuous structure of honeycomb cores, cell wall recognition plays an important role in its measurement data processing. The existing method is not stable enough in actual application because it is susceptible to burr data owing to its poor accuracy of identifying the true vertex from the detected corners. In this study, a new cell wall recognition method was proposed based on the corner type identification considering the regular structural features of the cell walls. The corner type identification method is used to distinguish the 3 types of corners according to the directions of the cell walls surrounding the given corner. By selectively determining the endpoints from the identified corners, the cell walls are recognized. One endpoint is determined once the neighbouring cell wall is recognized, and the other one is determined through detailed local analysis around the determined endpoint based on the corner type identification method. The experimental results demonstrate that the proposed method is significantly robust to burrs. Quantitatively, the recognition rate has an accuracy above 98%, and the mean position accuracy is below 0.2 mm.