|本期目录/Table of Contents|

[1]王海晖1,2,卢培磊1,等.无参考视频平滑度的评价方法[J].武汉工程大学学报,2015,37(06):56-62.[doi:10. 3969/j. issn. 1674-2869. 2015. 06. 012]
 ,,et al.Evaluation method of no-reference video smoothness[J].Journal of Wuhan Institute of Technology,2015,37(06):56-62.[doi:10. 3969/j. issn. 1674-2869. 2015. 06. 012]
点击复制

无参考视频平滑度的评价方法(/HTML)
分享到:

《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]

卷:
37
期数:
2015年06期
页码:
56-62
栏目:
机电与信息工程
出版日期:
2015-06-30

文章信息/Info

Title:
Evaluation method of no-reference video smoothness
文章编号:
1674-2869(2015)06-0056-07
作者:
王海晖1卢培磊1吴云韬1 陈双玉1孙志宏1
1. 武汉工程大学计算机科学与工程学院,湖北 武汉430205; 2. 智能机器人湖北省重点实验室(武汉工程大学),湖北 武汉 430205
Author(s):
WANG Hai-hui1 LU Pei-lei1 WU Yun-tao1 CHEN Shuang-yu1 SUN Zhi-hong1
1. School of Computer Science and Technology, Wuhan Institute of Technology, Wuhan 430205, China;2. Hubei Key Laboratory of Intelligent Robot (Wuhan Institute of Technology), Wuhan 430205, China
关键词:
视频平滑度ORB特征仿射变换平滑度评价
Keywords:
video smoothness ORB features affine transformation smoothness evaluation.
分类号:
TP391
DOI:
10. 3969/j. issn. 1674-2869. 2015. 06. 012
文献标志码:
A
摘要:
为了客观地评价视频图像的质量,提出了一种基于无参考视频平滑度的评价方法. 该方法使用ORB(Oriented FAST and Rotated BRIEF)特征点来描述视频帧,并通过使用仿射变换和FLANN(Fast Library for Approximate Nearest Neighbors)算法,得出相邻两幅视频帧中特征序列的平移轨迹和旋转轨迹,然后对轨迹进行高斯滤波和计算其抖动次数,最后计算出整个视频序列相邻两帧图像的参数序列,并得出视频平滑度. 由于该算法无需借助任何参考图像,仅依靠待评价图像本身的信息就可以进行质量评价,因此使用更加简便. 与标准视频图像检测结果相比,采用该方法所得出的检测结果具有极高的吻合度,完全满足实用性要求.
Abstract:
To evaluate the quality of video image objectively, a no-reference evaluation method based on video smoothness was proposed. The feature points of Oriented FAST and Rotated BRIEF were used to describe the video frame, and the affine transformation and the Fast Library for Approximate Nearest Neighbors algorithm were also used to get characteristics of translational trajectory and rotational trajectory in sequence of two consecutive video frames. Then we operated Gauss filter to the trajectories, and calculated the jitter times. Finally, we calculated parameters sequence of the whole video sequence of two adjacent frames, and got the video smoothness. The method is more convenient in use because it only relies on the evaluated images themselves to evaluate without using any reference images. Comparing with the detection results of standard video image, the detection results obtained by this method have goodness of fit, and meet the practical requirements.

参考文献/References:

[1] WANG Z, LU L, BOVIK A C. Video quality assessment based on structural distortion measurement[J]. Signal Processing:Image Communication,2004,19(2):121-132.[2] ONG E P, YANG X, LIN W, et al.Perceptual quality and objective quality measurements of compressed videos [J]. Journal of Visual Communication and Image Representation, 2006, 17(4): 717-737.[3] SESHADRINATHAN K, BOVIK A C. Motion-tuned Spatio-temporal Quality Assessment of Natural Videos[J]. IEEE Transactions on Image Processing, 2010, 19(2): 335-350.[4] GUNAWAN I P, GHANBARI M. Reduced-reference video quality assessment using discriminative local harmonic strength with motion consideration [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2008, 18(1):71-83.[5] ZENG K, WANG Z. Terrporal motion smoothness measurement forreduced-refere nee video quality asses-sment[C]//IEEE International Conference on Acoustics, Speech, and Signal Processing,2010:1010-1013.[6] ROSIN P L.Measuring corner properties[J]. Computer Vision and Image Understanding,1999, 73(2):291-307.[7] MARTIN A F, ROBERT C BES. Random sample onsensus: A paradigm for model fitting with applications to image analysis and automated cartography[J].Communication of the ACM, 1981,24(6):381-395.[8] LEE K Y, CHUANG Y Y, CHEN B Y, et al. Video stabilization using robust feature trajectories[C]// In Proceedings of the IEEE International Conference on Computer Vision, 2009:1397–1404.[9] RAWATAND P, SINGHAI J. Efficient video stabilization technique for hand held mobile videos[J].International Journal of Signal Processing, Image Processing and Pattern Recognition, 2013, 6(3):17-32.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2015-03-25基金项目:湖北省高等学校优秀中青年团队计划项目(T201206)作者简介:王海晖(1969-),男,河北石家庄人,教授,博士.研究方向:数字图像处理尧机器视觉尧模式识别.
更新日期/Last Update: 2015-08-23