Open AccessOn learning based compressed sensing for high resolution image reconstruction
Author(s) -
Islam Sheikh Rafiul,
Maity Santi P.,
Ray Ajoy Kumar
Publication year - 2020
Publication title -
iet image processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.401
H-Index - 45
eISSN - 1751-9667
pISSN - 1751-9659
DOI - 10.1049/ipr2.12029
Subject(s) - compressed sensing , computer science , convolutional neural network , artificial intelligence , image (mathematics) , deep learning , transmission (telecommunications) , iterative reconstruction , signal (programming language) , artificial neural network , matrix (chemical analysis) , computer vision , pattern recognition (psychology) , telecommunications , materials science , composite material , programming language
Compressed sensing (CS) or compressive sampling has shown an enormous potential to reconstruct a signal from its highly under‐sampled observations. A high dimensional image processing system can adopt the CS paradigm to reduce the storage and the transmission burden. However, a large sensing system (matrix) is required to capture high dimensional images. The CS reconstruction algorithms demand heavy computational requirement due to the use of large sensing matrix. It sometimes becomes impractical to implement a sensing system of the desired size due to the limited access to the CPU memory. To address this issue, the present work proposes a deep learning based CS framework that uses a convolutional neural network to enable capturing of a high dimensional image by utilizing a multi‐layer subsampling and filtering operations. The proposed approach uses another convolutional neural network that reconstructs the original image without depending on the sensing network at a significantly reduced computational cost. Extensive simulation results show that the proposed method reconstructs a high dimensional image with an improved PSNR value by 1.5 ± 0.63 dB, SSIM value by 0.04 ± 0.02 and FSIM value by 0.02 ± 0.02 compared to the other state‐of‐the‐art methods in less than 1.4 seconds.