针对基于结构照明的三维面形测量中阴影和条纹裂断的存在会影响物体面形重建质量的问题,提出一种新的获得整个物体三维面形的方法。在同一参考面内,旋转被测物体,并对旋转前、后的物体分两次进行测量,利用调制度信息确定重建物体高度数据的可靠区域,用最小二乘法找出两次重建的高度数据的坐标变换关系并在可靠区域内拼接,从而得到整个物体的正确三维面形分布。该方法操作简单,有效解决了传统方法不可避免的阴影和条纹裂断问题,保证了傅里叶变换轮廓术(FTP)的测量精度,通过实验证明了此方法的有效性。
Based on structured light illumination, the measuring precision is much affected by the shadow and the discontinuous captured fringe in optical 3-D sensing. In order to solve the problem, a method for obtaining the full 3D depth information is proposed, in which the test object is measured twice before and after rotating on the same reference plane. After obtaining the height maps from two deformed fringe patterns, modulation information is employed to confirm the valid mosaic regions in two reconstructed height maps, and the coordinates transformation relation of the two resulting height maps is obtained by the least-squares method. Its advantage is not only to ensure the measurement precision of Fourier transform profilometry (FTP), but also to avoid the influence of shadow in fringe.