The Effect of Noise on the Optimal Code Transmittance for Compressive Sensing
Abstract
Many compressive sensing architectures perform well in simulation with low transmittance codes. These low levels are deceptive, however, since they result in low signal levels that are more susceptible to noise. Accounting for imaging noise, however, cannot be accurately modeled by simple Gaussian noise. A more complex noise model includes both read and shot noise. Read noise is caused by the system hardware and is independent of signal level. Shot noise is based in the light itself and is related to signal level. This note explores the effect of these noise sources on optimal code transmittance.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 20, 2018
- Accession Number
- AD1065141
Entities
People
- Michael Don
Organizations
- United States Army Research Laboratory