Feature-Aware Adaptive
Denoiser-Selection for Compressed Image
Reconstruction
Mengke Ma, Dongqing Li, Shaohua Wu
(B) , and Qinyu Zhang
Harbin Institute of Technology, Shenzhen, Guangdong, China
{mamengke, lidongqing}@stu.hit.edu.cn
{hitwush, zqy}@hit.edu.cn
Abstract. Compressed sensing (CS) has been extensively studied in
image processing; however, the ill-posed inverse problem in the decoder
is complicated, still leaving room for further improvement. Denoisingbased approximate message passing (D-AMP) is a fast CS reconstruction
algorithm that can transform the complex reconstruction problem into
a classic denoising problem. But the existing denoisers are favorable on
images with special features, so the D-AMP algorithm which has a fixed
denoiser cannot universally achieve the best reconstruction quality for
all types of images. In this paper, we propose a CS image reconstruction
framework which can adaptively select the proper denoiser, and a featureaware denoiser-selection method which extracts features by fast Fourier
transform (FFT). Results show that the proposed feature-aware adaptive
denoiser-selection-based CS image reconstruction method can obtain the
best reconstruction quality for different types of images.
Keywords: CS reconstruction · Adaptive denoiser-selection · Feature
extraction
1 Introduction
Compressed sensing (CS) [1,2] is a new compression method of signal sampling
which breaks through the limitation of Nyquist sampling theorem. CS reduces
the sampling complexity of the encoder by increasing the complexity of decoder.
Over the last decade, the application of CS in image processing has gained considerable popularity, among which the reconstruction algorithm of decoder is very
important, which directly determines the application prospect of CS [3]. Typical CS reconstruction algorithms can be cast as iteratively estimating signals
from perturbed observations [4]. Denoising-based approximate message passing
(D-AMP) algorithm can integrate various existing denoisers in the iteration and
c
Springer Nature Singapore Pte Ltd. 2020
Q. Liang et al. (Eds.): Artificial Intelligence in China, LNEE 572, pp. 537–545, 2020.
https://doi.org/10.1007/978-981-15-0187-6_64
Denoiser-Selection for Compressed Image
Reconstruction
Mengke Ma, Dongqing Li, Shaohua Wu
(B) , and Qinyu Zhang
Harbin Institute of Technology, Shenzhen, Guangdong, China
{mamengke, lidongqing}@stu.hit.edu.cn
{hitwush, zqy}@hit.edu.cn
Abstract. Compressed sensing (CS) has been extensively studied in
image processing; however, the ill-posed inverse problem in the decoder
is complicated, still leaving room for further improvement. Denoisingbased approximate message passing (D-AMP) is a fast CS reconstruction
algorithm that can transform the complex reconstruction problem into
a classic denoising problem. But the existing denoisers are favorable on
images with special features, so the D-AMP algorithm which has a fixed
denoiser cannot universally achieve the best reconstruction quality for
all types of images. In this paper, we propose a CS image reconstruction
framework which can adaptively select the proper denoiser, and a featureaware denoiser-selection method which extracts features by fast Fourier
transform (FFT). Results show that the proposed feature-aware adaptive
denoiser-selection-based CS image reconstruction method can obtain the
best reconstruction quality for different types of images.
Keywords: CS reconstruction · Adaptive denoiser-selection · Feature
extraction
1 Introduction
Compressed sensing (CS) [1,2] is a new compression method of signal sampling
which breaks through the limitation of Nyquist sampling theorem. CS reduces
the sampling complexity of the encoder by increasing the complexity of decoder.
Over the last decade, the application of CS in image processing has gained considerable popularity, among which the reconstruction algorithm of decoder is very
important, which directly determines the application prospect of CS [3]. Typical CS reconstruction algorithms can be cast as iteratively estimating signals
from perturbed observations [4]. Denoising-based approximate message passing
(D-AMP) algorithm can integrate various existing denoisers in the iteration and
c
Springer Nature Singapore Pte Ltd. 2020
Q. Liang et al. (Eds.): Artificial Intelligence in China, LNEE 572, pp. 537–545, 2020.
https://doi.org/10.1007/978-981-15-0187-6_64