基于过曝连通域投影强度自适应分布的三维成像(特邀)

3D Imaging Based on the Adaptive Distribution of Projection Intensity in Overexposed Connected Domains(Invited)

针对高反光表面三维成像,提出了一种简单、稳健的自适应条纹投影方法。该方法首先通过投影白色均匀图案帮助确定过曝光区域,并提取该区域的边界;然后将边界映射到投影仪像素坐标系,并降低该区域内部的像素强度,重新生成一组亮度分布适应样品亮度分布的条纹图;最后迭代重复上述步骤,直至所采集的图中无过曝光区域。与以往采用多次曝光、自适应投影、需要捕获大量条纹图像、操作繁琐、计算复杂的成像方法相比,所提出的方法需要捕获的图案更少且更容易实现。实验结果表明,所提出的方法能够准确测量生物组织高亮表面,并最终重建出完整、高精度的三维轮廓。

This study proposes a simple and robust adaptive fringe projection method for the 3D measurement of highlighted surfaces.The method initially projects a uniform white pattern to help identify overexposed regions and extract boundaries.These boundaries are then mapped to a projector pixel coordinate system.The pixel intensities within the boundaries are reduced to regenerate new fringe maps with brightness distributions adapted to the sample.Finally,this process is repeated until no overexposed regions remain in the acquired images.Conventional methods use multiple exposures and adaptive projection,have complicated operations and calculations,and must capture a large number of fringe images.By contrast,the proposed method just needs to capture fewer patterns and is easier to implement.Experimental results demonstrate that the proposed method can accurately measure the highlighted surface of biological tissue and reconstruct complete and high-precision 3D profiles.