Abstract
We report Giga-pixel lensfree holographic microscopy and tomography using color sensor-arrays such as CMOS imagers that exhibit Bayer color filter patterns. Without physically removing these color filters coated on the sensor chip, we synthesize pixel super-resolved lensfree holograms, which are then reconstructed to achieve ∼350 nm lateral resolution, corresponding to a numerical aperture of ∼0.8, across a field-of-view of ∼20.5 mm2. This constitutes a digital image with ∼0.7 Billion effective pixels in both amplitude and phase channels (i.e., ∼1.4 Giga-pixels total). Furthermore, by changing the illumination angle (e.g., ±50°) and scanning a partially-coherent light source across two orthogonal axes, super-resolved images of the same specimen from different viewing angles are created, which are then digitally combined to synthesize tomographic images of the object. Using this dual-axis lensfree tomographic imager running on a color sensor-chip, we achieve a 3D spatial resolution of ∼0.35 µm×0.35 µm×∼2 µm, in x, y and z, respectively, creating an effective voxel size of ∼0.03 µm3 across a sample volume of ∼5 mm3, which is equivalent to >150 Billion voxels. We demonstrate the proof-of-concept of this lensfree optical tomographic microscopy platform on a color CMOS image sensor by creating tomograms of micro-particles as well as a wild-type C. elegans nematode.
Highlights
Color opto-electronic sensors such as CMOS imagers that exhibit Bayer patterns form the main stream detector-arrays employed in digital electronic devices, including cell phones, webcams and digital cameras, with a sales volume of .5 billion per year [1]
Physical removal of such color filters is costly to implement in high-volumes, and could alter the optimized design of pixel structures. Another option is to use monochrome version of the same sensor-chip of interest, which, is not often released by CMOS manufacturers since the main application areas, i.e., cell phones, and webcams, strictly demand color sensorarrays. To better handle this challenge, here we introduce a new computational approach to utilize color sensor-arrays in lensfree microscopy and tomography for achieving Giga-pixel imaging on a chip
To quantify the lateral resolution of our lensfree holographic set-up based on color CMOS sensor-arrays and validate our modified pixel super-resolution method (Fig. 2), we imaged a grating with 350 nm lines etched on glass (i.e., 700 nm period), which was fabricated using focused ion beam (FIB) milling
Summary
Color opto-electronic sensors such as CMOS imagers that exhibit Bayer patterns (composed of one Red, two Green and one Blue pixels) form the main stream detector-arrays employed in digital electronic devices, including cell phones, webcams and digital cameras, with a sales volume of .5 billion per year [1]. Physical removal of such color filters is costly to implement in high-volumes, and could alter the optimized design of pixel structures Another option is to use monochrome version of the same sensor-chip of interest, which, is not often released by CMOS manufacturers since the main application areas, i.e., cell phones, and webcams, strictly demand color sensorarrays. Without physically removing the color filters installed on a sensor chip, based on a new reconstruction approach, as illustrated, we synthesize pixel super-resolved lensfree holograms of specimen using 45u rotated green pixel functions (refer to Methods Section for details) These super-resolved holograms are reconstructed to achieve ,350 nm lateral resolution, corresponding to a numerical aperture (NA) of ,0.8, across a field-of-view of ,20.5 mm. These results create an effective voxel size of ,0.03 mm across a sample volume of ,5 mm, which is equivalent to .150 Billion voxels
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
