scispace - formally typeset
Open AccessJournal Article

Real FFT based on graphic processing unit for image processing

Reads0
Chats0
TLDR
By analysis of general real Fast Fourier Transform(FFT)algorithm, an improved real FFT algorithm is presented based on Graphic Processing Unit (GPU) and the experiment results show that the presented algorithm is practical for image processing.
Abstract
By analysis of general real Fast Fourier Transform(FFT)algorithm,an improved real FFT algorithm is presented based on Graphic Processing Unit(GPU).A general real FFT can reduce the computation consumption of the conventional complex FFT algorithm by about 40%,and the presented real FFT algorithm improves the reduction to 50%.On the basis of the specific configuration of data storage,the presented real FFT algorithm can be applied in two directions continuously to generate the 2D Fourier spectrum,also the parallel real FFT transform can be implemented by using the 1D complex FFT function from the CUFFT library.The precision and the efficiency of the presented real FFT are inspected by Wiener filter's application,image processing results demonstrate that the presented real FFT algorithm shows operating properly and working well.With the help of the presented real FFT,Wiener filter can restore a 2 048×2 048 8-bit image by motion speed of 19.26 ms,which is 2.34 times as compared with that of the Wiener filter based on complex FFT on GPU,and is 37.46 times that of the Wiener filter based on real FFT on CPU.The experiment results show that the presented algorithm is practical for image processing.

read more

Citations
More filters
Proceedings ArticleDOI

An Image Enhancement Method Based on Gamma Correction

TL;DR: The novel Gamma correction curve is presented to enhance the target image of the wheel alignment system and increases the contrast ratio of the chessboard target image which is benefit for feature extraction, points matching and vision measurement.
Proceedings ArticleDOI

Implementation and performance of image filtering on GPU

TL;DR: Compared with the traditional method of image filtering which was performed on CPU, the implementation of image filters on GPU has a speedup of approximately 10 times, which has great potential as high-performance co-processor.
Proceedings ArticleDOI

A Novel Corner Point Detector for Calibration Target Images Based on Grayscale Symmetry

TL;DR: The novel corner detection algorithm based on grayscale symmetry, designed to illustrate the symmetrical feature around the optimal corner, and the Gassian weighted symmetry detector, which can be used to test the corners of the calibration target precisely.
Journal ArticleDOI

Joint inspection in X-ray #0 belt tire based on periodic texture

TL;DR: Experimental results show that the algorithm can accurately locate the joint and quantify the size of joint with 3.4% false positive rate and 2% false negative rate, which meets the industrial requirements.
Proceedings ArticleDOI

Research on the Model and Finite Element Analysis of Vehicle Chassis Detection System Based on COSMOSWorks

TL;DR: Simulation results show that the dangerous place of the base assembly is the nearby regions of the double-rocker’s fixing holes where the limit stress is 2.398E+7N/m2 and the safety factor is about 26.2%, and the problem of detecting chassis gap accurately is resolved.
Related Papers (5)