Smoothing Method for Shape Measurement by Using a Low-Cost 3D Scanner
Hiroshi MARUO, Ayumu Saitoh, Tomohisa ISHIBASHI, Yuichi Tamura · Plasma and Fusion Research · 2011
A 3D scanner is a device that measures 3D objects in the real world and converts them to 3D digital data. However, much noise and several loss areas (holes) are observed in the outputted data, especially when using a low-cost 3D scanner. The purpose of this study is to propose a method for noise removal, burying holes, and outputting a smooth 3D model. We propose a smoothing method employing the multi-level partition of unity implicit (MPU) method. This method can compute a continuous surface by using implicit functions, enabling noise and loss area removal. However, the output 3D data generated by this method is not often accurate because the MPU method is sensitive to noise. Therefore, before applying the MPU method, we applied a noise reduction process. In this study, we propose a system that can grasp smooth 3D object data by using a low-cost 3D scanner.