Abstract We propose the obtaining of the 3D shapes by scanning with a light line of generated digitally Gaussian, triangular or sinusoidal profile. Algorithms based on the Bezier curves, linear regression, and nonlinear regression methods were implemented, which allowed to obtain the skeleton of each line of light captured with sub-pixel accuracy. The height measurements obtained with each projected profile were compared with a coordinate measuring machine (CMM). Finally, reconstruction efficiency was determined for each projected light profile in terms of accuracy and processing time.

中文翻译：

---

使用白光轮廓来勾勒物体的轮廓

摘要 我们建议通过扫描生成的数字高斯、三角形或正弦曲线的光线来获得 3D 形状。实施了基于贝塞尔曲线的算法、线性回归和非线性回归方法，从而能够获得以亚像素精度捕获的每条光线的骨架。将每个投影轮廓获得的高度测量值与坐标测量机 (CMM) 进行比较。最后，根据精度和处理时间确定每个投影光剖面的重建效率。