Resolution analysis in computational imaging with patterned illumination and bucket detection

Abstract

In computational imaging by pattern projection, a sequence of microstructured light patterns codified onto a programmable spatial light modulator is used to sample an object. The patterns are used as generalized measurement modes where the object information is expressed. In this Letter, we show that the resolution of the recovered image is only limited by the numerical aperture of the projecting optics regardless of the quality of the collection optics. We provide proof-of-principle experiments where the single-pixel detection strategy outperforms the resolution achieved using a conventional optical array detector for optical imaging. It is advantageous in the presence of real-world conditions, such as optical aberrations and optical imperfections in between the sample and the sensor. We provide experimental verification of image retrieval even when an optical diffuser prevents imaging with a megapixel array camera.

Dual-mode optical microscope based on single-pixel imaging

“…To boost up the image acquisition, we take advantage of the ground-breaking theory of compressive sensing (CS), which makes it possible to recover an N -pixels image from M < N measurements [24,25]. In CS, one takes 145 samples of the N-dimensional transmission vector T using a matrix M × N matrix S , obtaining undersampled measurements Y = S T. Along this section, and with no lack of generality, we will refer to the transmission signal, where an identical relation holds for the reflection mode.…”

“…Other structured illumination approaches have also been implemented for microscopy during the past few years. The single-pixel imaging (SPI) scheme [13] has gained considerable attention as a very effective sensing mechanism and has triggered diverse applications where conventional cameras equipped with 25 millions of pixels fail to give an adequate response, including optical microscopy [14]. As a matter of fact, SPI allows imaging at previously undeveloped spectral bands [15,16], at the photon-counting regime [17], in presence of strong turbulence or scattering [18,19], and also through multimode optical fibers [20,21].…”

“…If many different masks are used, their shapes and the intensity 40 signals are combined to retrieve the sample. This generalized mask scanning offers several advantages, like signal-to-noise ratio enhancement and the possibility to reduce the acquisition time through compressive sensing, over the traditional raster scan used in confocal microscopy where a single bright pixel is scanned to build up the image [24,25].…”

Super-Resolution Scanning Laser Microscopy Based on Virtually Structured Detection

“…62 Combining patterned illumination with the single-pixel detection strategy, it has been shown that the resolution of the reconstructed image is limited only by the NA of the projecting optics, regardless of the quality of the collection optics. 63 An apodization filter without adjustable parameters based on the application of the Lukosz bound was proposed to achieve the artifact-free reconstruction. 64 With 3D SIM, an additional 2-fold increase in the axial resolution can be achieved by modulating the excitation light along the z -axis using 3-beam interference and processing a z -stack of images accordingly.…”

Speckle-Shifting Ghost Imaging

“…The reference arm has been removed in computational ghost imaging (CGI) and replaced with sequential illumination patterns generated by a spatial light modulator (SLM) [6][7][8]. The patterned illumination and bucket detection in CGI significantly improve the performance of photon counting imaging [9], scattering imaging [10] and Terahertz imaging [11].…”