Full reference image quality assessment based on Gabor filter

Abstract

To resolve the problem that existing image quality assessment methods cannot assess the images as accurate as human visual systems,a full reference image quality assessment method based on Gabor filter was proposed.Gabor filter uses complex waveforms to modulate Gaussian function.It can simultaneously reach the theoretically limited lower bound in time and frequency on the product of its bandwidth and duration.The parameters of Gabor filter are optimized to simulate the human visual system effectively.After extracting the features from the Gabor filter and comparing them,better effects of image quality assessment are achieved.Experimental results show that the proposed method can correlate well with the human subjectivity in all kinds of distortions listed in the reference database.