集成成像同名像点自筛选技术

Self-screening of the corresponding points in integral imaging

从集成成像三维获取技术出发,利用光线追迹的方法,理论分析了元素图像阵列中同名像点之间的关联性,得到了同名像点之间的等差关系,并利用此关系式对光学获取的元素图像阵列提取的同名像点的结果进行筛选,去除误差大的配准结果,从而提升同名像点的配准精度。光学实验结果显示,对于同一物体进行三维获取,经本文的集成成像同名像点自筛选技术得到的结果是500mm,相对误差为0.4%;而传统技术重构得到的结果为502mm,相对误差为0.8%。我们所提的这种方法有效提升了再现精度,提升幅度为50%,为集成成像三维形貌获取技术实现高精度获取提供了有效的技术支持。

In order to realize precision of the passive 3D shape measurement technology,a new technology by integral imaging is proposed in this paper based on the connection of the corresponding points. A ray tracing method is used to analyze the processing of pick-up information by the integral imaging system. The relationship of the corresponding points is shown under the theoretical analysis, which is determined by regular lens array in the integral imaging. The coordinate difference between the neighbor correspond- ing points is the same with each other in an ideal world. This relationship is used to screen the points from the elemental images which have bigger errors in the actual optical experiments, and used to reduce the whole errors of the reconstruction. Optical experiment is done to confirm the feasibility of the pro- posed method. Optical results show that the length is 500 mm which is captured by the proposed meth- od, The relative error is 0. 4~, But the same length captured by the conventional method is 502 mm whose relative error is 0.8 %. This self-screening technology increases the precision of the results. This result confirms that our proposed method opens a new way for the application of the passive 3D shape measurement technology.