Full-reference video quality evaluation method based on Log-Gabor similarity

Abstract

The invention discloses a full-reference video quality evaluation method based on Log-Gabor similarity. The full-reference video quality evaluation method comprises the following steps: firstly, performing Log-Gabor transformation on U and V components of one frame of a video to obtain the amplitude and phase of a transformation coefficient, calculating the amplitude similarity and phase similarity, and comprehensively obtaining the Log-Gabor similarity; constructing a three-dimensional LBP feature for the Y video components of three adjacent frames, extracting a three-dimensional LBP featurehistogram, and calculating the three-dimensional LBP feature similarity of the distorted video and the reference video; and finally, combining the log-Gabor similarity and the three-dimensional LBP feature similarity to obtain a total similarity as an objective video quality evaluation result. According to the full-reference video quality evaluation method, the transform domain and spatial domaincharacteristics of the Y, U and V components of the video are fully considered, and the time domain characteristics are extracted by adopting the three-dimensional LBP characteristics, so that the video quality evaluation precision is improved.

Description

technical field [0001] The invention belongs to the field of video processing, in particular to a full-reference video quality evaluation method based on Log-Gabor similarity. Background technique [0002] Video quality assessment is a key issue in the field of video processing. Video quality assessment methods can be divided into subjective video quality assessment methods and objective video quality assessment methods according to whether people participate. The subjective video quality evaluation method uses humans to score images, and the evaluation results are accurate, but the evaluation process is complex and time-consuming, and it is difficult to be applied in real time. The objective video quality evaluation method does not require human participation, and the image quality is automatically predicted by a specific computer algorithm. According to whether the original undistorted video is used as a reference, the video quality evaluation method can be divided into a ...

AI Technical Summary

Problems solved by technology

The traditional full-reference video quality evaluation method uses MSE (Mean Square Error) or PSNR (Peak Signal-to-Noise Ratio) for video quality evaluation. This method has a clear physical meaning and a simple algorithm, but it has disadvantages such as mismatching with the subjective visual characteristics of the human eye. Not widely used in practice

For video quality assessment, many scholars have proposed improved methods, Zhang[Y.Zhang,X.-B.Gao,L.He,W.Lu,R.He.Objective Video Quality Assessment Combining Transfer Learning with CNN.IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS), 2019] Using convolutional neural network combined with transfer learning for video image quality evaluation; Lu [W.Lu, R.He, J.Yang, C.Jia, X.-B.Gao. A Spatiotemporal Model of Video Quality Assessment via 3DGradient Differencing.Information Science, Vol.478, pp.141-151, 2019.] proposes a video quality assessment method based on three-dimensional gradient difference, although these methods improve the no-reference image quality assessment accuracy, but there is still a gap between the results and the subjective image quality evaluation results of the human eye

Images