Luo Wei, Zhang Yibo, Göröcs Zoltán, Feizi Alborz, Ozcan Aydogan
Electrical Engineering Department, University of California, Los Angeles, CA, 90095, USA.
Bioengineering Department, University of California, Los Angeles, CA, 90095, USA.
Sci Rep. 2016 Mar 11;6:22738. doi: 10.1038/srep22738.
To achieve high-resolution and wide field-of-view, digital holographic imaging techniques need to tackle two major challenges: phase recovery and spatial undersampling. Previously, these challenges were separately addressed using phase retrieval and pixel super-resolution algorithms, which utilize the diversity of different imaging parameters. Although existing holographic imaging methods can achieve large space-bandwidth-products by performing pixel super-resolution and phase retrieval sequentially, they require large amounts of data, which might be a limitation in high-speed or cost-effective imaging applications. Here we report a propagation phasor approach, which for the first time combines phase retrieval and pixel super-resolution into a unified mathematical framework and enables the synthesis of new holographic image reconstruction methods with significantly improved data efficiency. In this approach, twin image and spatial aliasing signals, along with other digital artifacts, are interpreted as noise terms that are modulated by phasors that analytically depend on the lateral displacement between hologram and sensor planes, sample-to-sensor distance, wavelength, and the illumination angle. Compared to previous holographic reconstruction techniques, this new framework results in five- to seven-fold reduced number of raw measurements, while still achieving a competitive resolution and space-bandwidth-product. We also demonstrated the success of this approach by imaging biological specimens including Papanicolaou and blood smears.
为了实现高分辨率和宽视场，数字全息成像技术需要应对两个主要挑战：相位恢复和空间欠采样。以前，这些挑战分别通过相位检索和像素超分辨率算法来解决，这些算法利用了不同成像参数的多样性。尽管现有的全息成像方法可以通过依次执行像素超分辨率和相位检索来实现大的空间带宽积，但它们需要大量数据，这在高速或经济高效的成像应用中可能是一个限制。在此，我们报告一种传播相量方法，该方法首次将相位检索和像素超分辨率结合到一个统一的数学框架中，并能够合成具有显著提高的数据效率的新全息图像重建方法。在这种方法中，孪生图像和空间混叠信号以及其他数字伪像被解释为由相量调制的噪声项，这些相量在解析上取决于全息图和传感器平面之间的横向位移、样本到传感器的距离、波长和照明角度。与以前的全息重建技术相比，这个新框架将原始测量数量减少了五到七倍，同时仍能实现有竞争力的分辨率和空间带宽积。我们还通过对包括巴氏涂片和血涂片在内的生物样本成像证明了这种方法的成功。
