High-speed compressed sensing reconstruction on FPGA using OMP and AMP

Bai, Lin; Maechler, P.; Muehlberghuber, Michael; Kaeslin, Hubert

doi:10.1109/icecs.2012.6463559

Cited by 84 publications

(52 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For a fairly comparison, the sparsity as suggested in the corresponding paper is used as the number of iterations so as to indicate the total recovery time for a 512 Â 512 image. It can be seen that the proposed hardware is 5:87Â faster than [9], 4:25Â faster than [10] and 1:92Â faster than the state-of-the-art implementation [11]. It also should be pointed out that the reconstructed image quality from the proposed hardware is better than OMP as it is indicated in Fig.…”

Section: Results and Comparisonsmentioning

confidence: 74%

“…It reaches 85.69 MHz in Virtex-5 and takes 7.13 us to reconstruct 128 samples (K ¼ 0:25; S ¼ 5). Two algorithms (OMP and AMP) have been implemented in Virtex-6 in [10]. The maximum frequency is 100 MHz and 165 MHz for OMP and AMP, respectively.…”

Section: Introductionmentioning

confidence: 99%

“…Previous hardware reports of CS have been mainly focused on the algorithms of split bregman algorithm (SBA) [7], OMP [8,9,10,11], and approximate message passing (AMP) [10]. In [7], it takes 38.2 ms to reconstruct a 128 Â 128 image using SBA.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Hardware efficient architecture for compressed imaging

Luo

Huang

Chang

et al. 2014

IEICE Electron. Express

View full text Add to dashboard Cite

Compressed sensing has gained a wide application in image acquiring and reconstructing. Separable linear reconstruction has been shown to be effective in compressed imaging. This paper presents efficient hardware architecture based on adaptive sampling and separable reconstructing. By exploiting parallel properties in the architecture and timing scheme, high performance hardware has been proposed for both encoding and decoding sides. High performance Cholesky based matrix inversion has been implemented to solve the least square problem. Besides, high precision arithmetic element functions (reciprocal and square root reciprocal) have been presented by using of table interpolation and single iterated Newton-Raphson method. Experiment results show that the proposed hardware architecture can efficiently reduce the process time in encoding and decoding of a 512 # 512 image. The speedup is about 58# compared with the software-based approach (using LAPACK), and it is at least 1.92# faster than the state-of-theart implementation.

show abstract

Section: Results and Comparisonsmentioning

confidence: 74%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hardware efficient architecture for compressed imaging

Luo

Huang

Chang

et al. 2014

IEICE Electron. Express

View full text Add to dashboard Cite

show abstract

“…Various applications, such as signal de-noising, satellite remote sensing, image sensing for SAR echo mode and mobile communication, are aiming to use CS. Current research is focused on both sampling techniques and enhancing the reconstruction algorithms in terms of accuracy [1], [2] and complexity reduction on different platforms [3], [4], [5], [6], [7], [8], [9], [10].…”

Section: Introductionmentioning

confidence: 99%

Accelerating compressive sensing reconstruction OMP algorithm with CPU, GPU, FPGA and domain specific many-core

Kulkarni

2015

2015 IEEE International Symposium on Circuits and Systems (ISCAS)

View full text Add to dashboard Cite

Compressive Sensing (CS) signal reconstruction can be implemented using convex relaxation, non-convex, or local optimization algorithms. Though the reconstruction using convex optimization, such as the Iterative Hard Thresholding algorithm, is more accurate than matching pursuit algorithms, most researchers focus on matching pursuit algorithms because they are less computationally complex. Orthogonal Matching Pursuit (OMP) is a greedy algorithm, which solves the problem by choosing the most significant variable to reduce the least square error. In this paper, we propose an efficient parallel architecture for OMP CS reconstruction. For architecture implementation, we perform measurement and sparsity analysis to reduce the complexity. The proposed architecture is platform independent and is implemented on 7 different platforms including general purpose CPUs, GPUs, a Virtex-7 FPGA and a domain specific many-core. The implementation results indicate that reconstruction time on FPGA is improved by 3× compared to previous FPGA implementation, whereas GPU implementation is 4× faster than the previously proposed GPU-based OMP architecture. The CPU implementation is 6× faster, compared with previous CPUbased implementation. The domain specific many-core acheives 24 times faster reconstruction time when compared to both GPU and CPU implementations.

show abstract

“…Patric Machler, Lin Bai, Hubert and Michel published paper [21] as "High Speed Compressed Sensing Reconstruction on FPGA Using OMP and AMP". Compressed sensing is mathematically modeled as y=Φx, where x is i/p signal, Φ is non-adaptive linear projections.…”

Section: Inputmentioning

confidence: 99%

A Survey and Theoretical View on Compressive Sensing and Reconstruction

Bujari¹,

Siddamal²

2016

IJIGSP

View full text Add to dashboard Cite

Abstract-Most of the current embedded systems operate on digital domain even though input and output is analog in nature. All these devices contain ADC (Analog to Digital converter) to convert the analog signal in to digital domain which is used for processing as per the application. Images, videos and other data can be exactly recovered from a set of uniformly spaced samples taken at the Nyquist rate. Due to the recent technology signal bandwidth is becoming wider and wider. To meet the higher demand, signal acquisition system need to be improved. Traditional Nyquist rate which is used in signal acquisition suggests taking more numbers of samples to increase the bandwidth but while reconstruction most of the samples are not used. If samples are as per Nyquist rate then, this increases the complexity of encoder, storage of samples and signal processing. To avoid this new concept Compressive Sensing is used as an alternative for traditional sampling theory. This paper presents a survey and simplified theoretical view on compressive sensing and reconstruction and proposed work is introduced.

show abstract

High-speed compressed sensing reconstruction on FPGA using OMP and AMP

Cited by 84 publications

References 11 publications

Hardware efficient architecture for compressed imaging

Hardware efficient architecture for compressed imaging

Accelerating compressive sensing reconstruction OMP algorithm with CPU, GPU, FPGA and domain specific many-core

A Survey and Theoretical View on Compressive Sensing and Reconstruction

Contact Info

Product

Resources

About