eduzhai > Applied Sciences > Engineering >

Multi-element microscope optimization by a learned sensing network with composite physical layers

  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Standard microscopes offer a variety of settings to help improve thevisibility of different specimens to the end microscope user. Increasingly,however, digital microscopes are used to capture images for automatedinterpretation by computer algorithms (e.g., for feature classification,detection or segmentation), often without any human involvement. In this work,we investigate an approach to jointly optimize multiple microscope settings,together with a classification network, for improved performance with suchautomated tasks. We explore the interplay between optimization of programmableillumination and pupil transmission, using experimentally imaged blood smearsfor automated malaria parasite detection, to show that multi-element "learnedsensing " outperforms its single-element counterpart. While not necessarilyideal for human interpretation, the network s resulting low-resolutionmicroscope images (20X-comparable) offer a machine learning network sufficientcontrast to match the classification performance of correspondinghigh-resolution imagery (100X-comparable), pointing a path towards accurateautomation over large fields-of-view.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×