彩色结构光三维重建过程中,系统的非线性耦合以及待测曲面的拓扑结构等均会对结构光解码产生影响,从而导致调制条纹漏检和颜色码误识别。为解决这一问题,提出一种基于彩色编码结构光的三维重建方法。利用YUV颜色通道对调制条纹进行滤波差分投影处理,通过调制条纹的波形分布提取中心特征线;利用颜色聚类方法精确获取调制条纹的颜色码信息值;最后,为了建立编码条纹码字与调制条纹码字间的对应关系,提出了基于序列特征组合的优化匹配方法,结合双目视觉深度感知数学模型,求解出编码码字的空间三维信息值。通过实验分析可知,本文方法的漏检率低,颜色码识别率高,具有较强的抗干扰性和鲁棒性。

In the 3D reconstruction of color structured light, the non-linear coupling of systems, the topological structure of the test surface, and other factors will affect the decoding of structured light, which can lead to the missed detection of modulation fringes and the false recognition of color codes. To solve the problem, we proposed the 3D reconstruction method of color-coded structured light. First, filtering differential projection was performed on the modulation fringes through the YUV color channel, and then the center feature lines of the fringes were extracted after an analysis of the waveform distribution of the fringes. Furthermore, the information values of color codes of the fringes were accurately acquired by virtue of the color clustering method. Finally, in order to establish the corresponding relationship between the codewords of coded fringes and those of modulation fringes, we put forward an optimal matching method based on the combination of sequence features, and combined with the mathematical model of binocular depth perception, we solved the three-dimensional information value of the codewords of coded fringes. The experimental results show that this method has a low rate of missed detection, a high recognition rate of color codes, and strong anti-interference and robustness.