Fast Reconstruction of Three-Quarter Sampling Measurements Using Recurrent Local Joint Sparse Deconvolution and Extrapolation
Authors:
Simon Grosche,
Andy Regensky,
Alexander Sinn,
Jürgen Seiler,
André Kaup
Abstract:
Recently, non-regular three-quarter sampling has shown to deliver an increased image quality of image sensors by using differently oriented L-shaped pixels compared to the same number of square pixels. A three-quarter sampling sensor can be understood as a conventional low-resolution sensor where one quadrant of each square pixel is opaque. Subsequent to the measurement, the data can be reconstruc…
▽ More
Recently, non-regular three-quarter sampling has shown to deliver an increased image quality of image sensors by using differently oriented L-shaped pixels compared to the same number of square pixels. A three-quarter sampling sensor can be understood as a conventional low-resolution sensor where one quadrant of each square pixel is opaque. Subsequent to the measurement, the data can be reconstructed on a regular grid with twice the resolution in both spatial dimensions using an appropriate reconstruction algorithm. For this reconstruction, local joint sparse deconvolution and extrapolation (L-JSDE) has shown to perform very well. As a disadvantage, L-JSDE requires long computation times of several dozen minutes per megapixel. In this paper, we propose a faster version of L-JSDE called recurrent L-JSDE (RL-JSDE) which is a reformulation of L-JSDE. For reasonable recurrent measurement patterns, RL-JSDE provides significant speedups on both CPU and GPU without sacrificing image quality. Compared to L-JSDE, 20-fold and 733-fold speedups are achieved on CPU and GPU, respectively.
△ Less
Submitted 5 May, 2022;
originally announced May 2022.
