Resolution Loss by Pointwise Degradations

Pointwise image degradations are caused by several sources; added specular or indirect reflections, additive scattering, radiance attenuation in haze and flash photography. It is found through a linear-systems analysis that in these cases there is a cutoff frequency induced by noise despite having no optical blur in the imaging model. This low signal-to-noise ratio (SNR) results in resolution loss similar to image blur. The intention of this research is to perform image recovery on images which suffer from this phenomenon, specifically in hazy images.

While image recovery under noise is widely studied, there is little emphasis on performance as a function of object size. This research results in a tool to assess the ability to recover an object or feature in an image depending on its size, distance from the camera and radiance difference from its nearby background per attenuation coefficient of the medium. This analysis also suggested the potential of polarization in visibility enhancement and allowed for a quantization of its viability.

Presentations

POLARIZATION - BENEFICIAL ​FOR VISIBILITY ENHANCEMENT?

Polarization - beneficial for visibility enhancement? slides.

Screenshot from 2021-01-20 14-52-19.png

RESOLUTION LOSS WITHOUT OPTICAL BLUR

Resolution loss without optical blur slides.

Screenshot from 2021-01-20 14-52-05.png
 

Publications

2012resolutionLossa.png

Resolution Loss Without Imaging Blur. Tali Treibitz and Yoav Y. Schechner. J. Optical Society of America A, 2012.

2009polarizationa.png

Polarization- Beneficial for Visibility Enhancement? Tali Treibitz and Yoav Y. Schechner. oral in CVPR, 2009.

2009pointwiseDegradationa.png

Recovery Limits in Pointwise Degradation. Tali Treibitz and Yoav Y. Schechner. Proc. IEEE ICCP, 2009.