|
Pixelation, blur and additional noise of imaging system limit the resolution of final images acquired. Many pixel superresolution algorithms have been applied to enhance the resolution of imaging system by merging a sequence of lowresolution holograms with different type of imaging system, for example by shifting illumination source or using wavelength scanning. Most of these pixel super-resolution imaging systems can only be implemented to single-layer sample. For multi-layer imaging system and volumetric imaging scenarios, the relative displacement of various sample at different layers will disturb each other. Herein, we report a portable, cost-effective, lensless wide-filed digital in-line holographic microscopy imaging system based on in-line hologram segmentation and pixel super-resolution algorithm, which can separate target sample from the background and improve the resolution of the sample. We demonstrated the effectiveness of our system with numerical simulation and experiment with volumetric sample. In numerical simulation, we applied a very simple two-layer sample model that samples in two layers have various moving speed and directions and also did the volumetric imaging experiment with cuvette containing algae floating in. In our system, the sensor captured a sequence of low-resolution diffraction patterns. The target sample mix with background disturbance, which will invalidate the direct pixel super-resolution technique. We applied segmentation algorithm to the reconstructed images from holograms, separating target sample from background and generating a sequence of sub-images containing only target sample with same resolution and numerical aperture as original holograms. Finally the enhanced resolution reconstructed image of target sample was obtained with pixel super-resolution algorithm, which can go beyond pixel limitation and get sub-pixel perspective microscopy. This imaging system has the advantages of wide-field, portable and lensless.
|
