压缩全息搭起Gabor全息和压缩感知(Compressed sensing,CS)理论之间的桥梁,特别适合从单帧二维全息测量数据中重建三维对象,是一种新兴的三维重建技术.本文将压缩全息方法从单波长情形推广到多波长,提出一种基于三维总变分稀疏模型的改进彩色全息压缩成像方法,建立多波长情形下的压缩测量模型.该方法利用对象的稀疏先验知识,从单帧二维彩色全息图中重建多波长三维对象,有效地实现孪生像的抑制和多层切片相互之间的散焦图像对重建质量的影响.数值实验结果验证了本文提出方法的有效性.
Compressed holography is an emerging 3D reconstruction technique, which bridges the gap between com- pressed sensing (CS) theory and Gaborts holography, especially for rebuilding 3D objects from a single-frame 2D holog- raphy measurement data. In this paper, the single-wavelength settings in compressed holography are extended to the multi-wavelength, and an improved compressed color holography imaging method is proposed, and a compressed mea- surement model in multi-wavelength case is established. Utilizing sparse prior knowledge of an object, a multi-wavelength 3D object can be reconstructed effectively from single-frame 2D color holography data of the object, so as to suppress the twin image and the defocus image due to multilayer slices and thus improve high quality reconstruction. Numerical results have demonstrated the effectiveness of our method.