Parallel 2D FFT implementation on FPGA suitable for real-time MR image processing

Liu, Limin; Wyrwicz, Alice M.

doi:10.1063/1.5019846

Cited by 13 publications

(9 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in the figure, the vectorized DFT and IDFT kernels consume (3.6 ms, 3.8 ms) and (2.43 ms, 2.59 ms) on 192 cores Intel and 768 cores NVIDIA GPUs, which are comparatively lesser than the 2D FFT implementation using CUFFT library on 5120 cores NVIDIA GPU or on a FPGA XC7K410T at a clock frequency 40 MHz. [62,63]. Figure 12 exhibits the average creation times of filters' matrices are too high for Gaussian and LoG filters, i.e., the expected filtration times are low compared to Ideal and Butterworth filters.…”

Section: Resultsmentioning

confidence: 99%

Optimized OpenCL™ kernels for frequency domain image high-boost filters using image vectorization technique

Satapathy

Livingston

2019

SN Appl. Sci.

View full text Add to dashboard Cite

Image high boost filtering uses high-boost filters to enhance the quality of an image, which has also seen in remote sensing, satellite broadcasting, classroom monitoring, and many more real-time video processing applications and requires its faster implementation. OpenCL is a widely adapted parallel programming framework that provides core level data parallelism, and dedicated for heterogeneous parallel devices like from low cost DSP to high-end CPU, GPU and FPGA. In this article, we have considered mostly used Ideal, Gaussian, Butterworth, and Laplacian of Gaussian frequency domain high-boost filters and implemented channelized OpenCL kernels for their rapid execution. In addition to that, these kernels are modified using image vectorization technique to optimize their time utilization by reducing the execution time of these OpenCL kernels to half. At last, performance analysis is carried out for these two types of OpenCL kernel implementations to determine their effectiveness with regard to time consumption and accuracy. Here, different image performance evaluation metrics like entropy, standard deviation, mean absolute error, percentage fit error, SSIM, correlation, and peak signal to noise ratio are applied to measure rightness of the above high-boost filters. From the results, we have concluded that a vectorized Butterworth high-boost filter kernel is the suitable one to provide better results among those filters, which might be highly adaptable in time bound real-time applications using various embedded devices.

show abstract

Section: Resultsmentioning

confidence: 99%

Optimized OpenCL™ kernels for frequency domain image high-boost filters using image vectorization technique

Satapathy

Livingston

2019

SN Appl. Sci.

View full text Add to dashboard Cite

show abstract

“…Main tasks are data sorting, optical link handling, multiplexing and forwarding of data from different interfacing links, embedding control and trigger information, etc. [14]. These versatile functionalities require RU to be designed on custom electronics boards with re-programmable functionality [15].…”

Section: Data Aggregation and Processingmentioning

confidence: 99%

Optimization of multi-gigabit transceivers for high speed data communication links in HEP Experiments

Khan

Mitra

Das

et al. 2019

Nuclear Instruments and Methods in Physics Research Section A:

View full text Add to dashboard Cite

The scheme of the data acquisition (DAQ) architecture in High Energy Physics (HEP) experiments consist of data transport from the front-end electronics (FEE) of the online detectors to the readout units (RU), which perform online processing of the data, and then to the data storage for offline analysis. With major upgrades of the Large Hadron Collider (LHC) experiments at CERN, the data transmission rates in the DAQ systems are expected to reach a few TB/sec within the next few years. These high rates are normally associated with the increase in the high-frequency losses, which lead to distortion in the detected signal and degradation of signal integrity. To address this, we have developed an optimization technique of the multi-gigabit transceiver (MGT) and implemented it on the state-of-the-art 20nm Arria-10 FPGA manufactured by Intel Inc. The setup has been validated for three available high-speed data transmission protocols, namely, GBT, TTC-PON and 10 Gbps Ethernet. The improvement in the signal integrity is gauged by two metrics, the Bit Error Rate (BER) and the Eye Diagram. It is observed that the technique improves the signal integrity and reduces BER. The test results and the improvements in the metrics of signal integrity for different link speeds are presented and discussed.

show abstract

“…The image reconstruction module was largely based on one of the recently developed 2D FFT modules 24 with small modifications. In the present work, we considered only a single image with pixel resolution of 128 × 128.…”

Section: Image Reconstruction Modulementioning

confidence: 99%

A multifunction digital receiver suitable for real-time frequency detection and compensation in fast magnetic resonance imaging

Liu

Wyrwicz

2019

Review of Scientific Instruments

Self Cite

View full text Add to dashboard Cite

We describe the development and implementation of a multifunction digital receiver suitable for magnetic resonance imaging with capability of real-time frequency detection and compensation. The digital receiver consists primarily of firmware modules that combine the functionalities of signal acquisition, frequency detection and compensation, and data correction and image reconstruction. The receiver was developed based on a single multiple-input multiple-output radio-frequency electronic board equipped with a reconfigurable Field Programmable Gate Array (FPGA) device. A simple and practical algorithm was developed and implemented on the FPGA to accelerate the data processing for frequency determination. The simplified frequency detection and the higher system integration enable the receiver to reduce dramatically the time for frequency detection and compensation. With this receiver, we are able to detect the frequency of short-duration signals in the bandwidth of 10 MHz centered at 400 MHz within 75 ns after the signal acquisition. We describe the designs of the key FPGA modules and how these modules integrate into a multifunction receiver. We also present testing data that validate the simplified algorithm for frequency determination, demonstrate frequency detection and compensation, and demonstrate how real-time data correction is performed during image acquisition and reconstruction.

show abstract

Parallel 2D FFT implementation on FPGA suitable for real-time MR image processing

Cited by 13 publications

References 37 publications

Optimized OpenCL™ kernels for frequency domain image high-boost filters using image vectorization technique

Optimized OpenCL™ kernels for frequency domain image high-boost filters using image vectorization technique

Optimization of multi-gigabit transceivers for high speed data communication links in HEP Experiments

A multifunction digital receiver suitable for real-time frequency detection and compensation in fast magnetic resonance imaging

Contact Info

Product

Resources

About