VLSI Architecture for OMP to Reconstruct Compressive Sensing Image (original) (raw)
Advances in Science, Technology and Engineering Systems Journal
A real-time embedded system requires plenty of measurements to fallow the Nyquist criteria. The hardware built for such a large number of measurements, is facing the challenges like storage and transmission rate. Practically it is very much complex to build such costly hardware. Compressive Sensing (CS) will be a future alternate technique for the Nyquist rate, specific to some applications where sparsity property plays a major role. Software implementation of Compressive Sensing takes more time to reconstruct a signal from CS measurements using the Matching Pursuit (MP) algorithm because of fetching, decoding, and execution policy. It is necessary to build hardware in CS. The author proposes one such VLSI Architecture (Hardware) for 256 256 and 512 512image. Various random matrices like Bernoulli, Partial Hadmard, Uniform Spherical, and Random Matrix are used to build hardware. FHT (Foreward Transform) with ±2 6 threshold is applied to get CS measurements. The reconstruction time, Signal to Noise ratio (SNR), and Mean Square Error (MSE) are measured. Multiple time experiments are carried out and results show that for an image of size 256 256, SNR is 25 and MSE is 166. For the image of size 512 512, the values are 27 and 182. However, both the input images are resized to 256 256 so the reconstruction time is 2.62µ which is less is compared to software implementation.
Sign up for access to the world's latest research
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Related papers
Compressed sensing recovery algorithms and VLSI implementation
2013 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, 2013
This paper proposes two recovery algorithms modified from subspace pursuit(SP) for compressed sensing problems. These algorithms can reduce the complexity of SP and maintain high recovery rate. Complexity analysis and simulation results are provided to demonstrate the improvements. Additionally this work has implemented the VLSI circuit APR of the proposed algorithm using TSMC 90 nm process. The target clock frequency is 100MHz, and the corresponding APR dimension is 11.69 2. Based on the post-layout simulation the average power consumption is 431 mW.
Signal reconstruction processor design for compressive sensing
This paper presents a very-large-scale integration (VLSI) design to reconstruct compressively sensed data. The proposed digital design recovers signal compressed by specific analog-to-digital converter (ADC). Our design is based on a modified iterative hard threshold (IHT) reconstruction algorithm to adapt unknown and varying degree of sparsity of the signal. The algorithm is composed empirically and implemented in a hardware-friendly fashion. The reconstruction fidelity using fixedpoint hardware model is analyzed. The design is synthesized using Synopsys Design Compiler with TSMC 45nm standard cell library. The post-synthesis implementation consumes 165 mW and is able to reconstruct data with information sparsity of 4%, at equivalent sampling rate of 1 gigasample-per-second (GSPS).
Compressive sampling hardware reconstruction
2010
Compressive Sampling reconstruction techniques require computationally intensive algorithms, often using L 1 optimization to reconstruct a signal that was originally sampled at a sub-Nyquist rate. In this work we present a VLSI implementation of a computationally efficient algorithm named Orthogonal Matching Pursuit. We further optimize the algorithm to meet typical hardware constraints and describe the different block units of our design. We synthesize our design for the Xilinx Virtex 5 FPGA and give timing and area results. We summarize our work with a short discussion of the possible uses for our system.
Image Reconstruction Using Modified Orthogonal Matching Pursuit And Compressive Sensing
—Compressive sensing system merges sampling and compression for a given sparse signal. It can reconstruct the image accurately by using fewer linear measurements than the original measurements. Hence, it is able to achieve reduction in complexity of sampling and number of computations. Since existing algorithms for implementation of sampling for the whole image are time consuming and it requires huge storage space, greedy approaches are used commonly to recover sparse signals from fewer measurements. One of the commonly used greedy approaches is Orthogonal Matching Pursuit (OMP), which can iteratively reconstruct the image. In this paper, modified form of OMP is presented in which stopping condition specified by the Recovery condition and Mutual incoherence property is used on the low frequency coefficients of the image. The simulation result using modified OMP shows that the reconstructed image achieves better PSNR and uses lesser number of measurements.
IET Circuits, Devices & Systems, 2021
Compressed sensing-based radio frequency signal acquisition systems call for higher reconstruction speed and low dynamic power. In this study, a novel low power fast orthogonal matching pursuit (LPF-OMP) algorithm is proposed for faster reconstruction of sparse signals from their compressively sensed samples and the reconstruction circuit consumes very low dynamic power. The searching time to find the best column is reduced by reducing the number of columns to be searched in successive iterations. A novel architecture of the proposed LPF-OMP algorithm is also presented here. The proposed architecture is implemented on field programmable gate array for demonstrating the performance enhancement. Computation of pseudoinverse in OMP is avoided to save time and storage requirement to store the pseudoinverse matrix. The proposed design incorporates a novel strategy to stop the algorithm without consuming any extra circuitry. A case study is carried out to reconstruct the RADAR test pulses. The design is implemented for K = 256, N = 1024 using XILINX Virtex6 device and supports maximum of K/4 iterations. The proposed design is faster, hardware efficient and consumes very less dynamic power than the previous implementations of OMP. In addition, the proposed implementation proves to be efficient in reconstructing low sparse signals. 1 | INTRODUCTION High-frequency radio frequency (RF) signals, such as RADAR pulses, are sparse in nature in the transform domain. Exploiting this sparsity property, modern signal measurement systems use compressed sensing (CS) [1,2] in place of other existing sampling techniques [3] to acquire RF signals. CSbased acquisition systems can work with low speed analog-todigital converters due to sampling at sub-Nyquist rate [4]. In CS-based sampling paradigm, random measurements are taken from the signal and then the original signal is recovered from the measurement samples using signal recovery algorithms. Orthogonal matching pursuit (OMP) [5,6] is a well known recovery algorithm. Unlike the other greedy pursuit algorithms, OMP provides better performance with moderate computational complexity. OMP estimates a sparse signal by executing two steps in every iteration, viz., perform the atom searching (AS) and solve a least squares (LS) problem. In AS step, OMP identifies an atom or a column of the sampling matrix which gives maximum correlation with the current residual. Subsequently the signal is estimated by solving an LS problem. The timing complexity of the AS step is very high as it is a linear function of the signal sparsity and the number of samples. Many techniques are reported in literature to reduce the timing complexity of AS step. In [7] authors applied clustering algorithms to group the similar columns and reported a tree-based pursuit algorithm. But such algorithm has no reports of implementation. Researchers reported parallel selection of multiple columns to address the timing complexity problem in [8,9]. Multiple selection of columns reduces the timing complexity but with greater chances of choosing wrong columns. Many implementations are reported based on either field programmable gate arrays (FPGAs) or application-specific integrated circuits. The LS problem is solved in different ways in current research works. The implementations of OMP reported in [9-15] used modified Cholesky factorization [16] to solve the LS problem. The LS problem is solved by lowerupper decomposition [16] in [17]. QR decomposition [16] is another powerful matrix factorization technique which is used This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
Reconstruction of Compressive Sensing Signal Using Orthogonal Matching Pursuit Algorithm
This paper represents the reconstruction of sampled signal in CS by using OMP algorithm. We have used the concept of compressive sensing for sub Nyquist sampling of sparse signal. Compressive sensing reconstruction methods have complex algorithms of l1 optimisation to reconstruct a signal sampled at sub nyquist rate. But out of those algorithm OMP algorithm is fast and computationally efficient. To prove the concept of CS implementation, we have simulated OMP algorithm for recovery of sparse signal of length 256 with sparsity 8.
A Survey and Theoretical View on Compressive Sensing and Reconstruction
International Journal of Image, Graphics and Signal Processing
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.
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.