Abstract
Summarizes our progress towards a fully transparent, flexible, and scalable thin-film image sensor. In contrast to conventional image sensors, it does not capture pixels in image space on the sensor surface, but makes integral measurements in Radon space along the sensor’s edges. Image reconstruction is achieved by inverse Radon transform. By stacking multiple layers, it enables a variety of information, such as color, dynamic range, spatial resolution, and defocus, to be sampled simultaneously. Lensless multi-focal imaging allows reconstructing an entire focal stack after only one recording. The focal stack can then be applied to estimate depth from defocus. Measuring and classifying directly in Radon space yields robust and high classification rates. Dimensionality reduction results in task-optimized classification sensors that record a minimal number of samples. This enables simple devices with low power consumption and fast read-out times.
© 2016 Optical Society of America
PDF ArticleMore Like This
Ronan Kerviche and Amit Ashok
CW5D.7 Computational Optical Sensing and Imaging (COSI) 2016
Nobukazu Teranishi
IMA1 Imaging Systems and Applications (IS) 2011
Boyd Fowler
IT2F.1 Imaging Systems and Applications (IS) 2016