[1]朱文佳,张 婷,程茹秋,等.考虑背景失真的无参考视频质量评价方法[J].南京师大学报(自然科学版),2025,48(03):102-111.[doi:10.3969/j.issn.1001-4616.2025.03.012]
　Zhu Wenjia,Zhang Ting,Cheng Ruqiu,et al.No-Reference Video Quality Assessment Method Based on Background Distortions[J].Journal of Nanjing Normal University(Natural Science Edition),2025,48(03):102-111.[doi:10.3969/j.issn.1001-4616.2025.03.012]
点击复制

考虑背景失真的无参考视频质量评价方法()

《南京师大学报（自然科学版）》[ISSN:1001-4616/CN:32-1239/N]

卷:
48
期数:
2025年03期
页码:
102-111
栏目:
计算机科学与技术
出版日期:
2025-06-20

文章信息/Info

Title:
No-Reference Video Quality Assessment Method Based on Background Distortions
文章编号:
1001-4616(2025)03-0102-10
作者:
朱文佳1; 2; 张 婷2; 程茹秋2; 余 烨2
(1.安徽百诚慧通科技股份有限公司,安徽 合肥 230001)
(2.合肥工业大学计算机与信息学院,安徽 合肥 230000)
Author(s):
Zhu Wenjia1; 2; Zhang Ting2; Cheng Ruqiu2; Yu Ye2
(1.Anhui Baichenghuitong Technology Co.,Ltd.,Hefei 230001,China)
(2.School of Computer and Information,Hefei University of Technology,Hefei 230000,China)
关键词:
视频质量评价; 无参考; 背景失真; 通道挖掘机制; 时序建模
Keywords:
video quality assessment(VQA); no-reference; background distortion; channel mining mechanism; time series modeling
分类号:
TP391.41
DOI:
10.3969/j.issn.1001-4616.2025.03.012
文献标志码:
A
摘要:
现实场景下拍摄的视频由于存在各种未知失真类型、缺少参考视频,对此类视频的质量评价是一个十分具有挑战性的任务. 近年来,研究人员将人类视觉系统的先验知识融合在质量评价任务中. 在此基础上,提出一种考虑背景失真的无参考视频质量评价方法. 该方法在考虑视频内容的同时,显著增强了对视频背景中信息丢失问题的敏感度,在特征提取阶段充分考虑背景特征的提取; 随后,通过引入结合门控机制的通道挖掘技术,高效整合高低维特征,使特征通道更加精准地聚焦于背景失真细节; 最终,利用时序建模模块构建特征的时间维度模型,并通过线性回归方法生成视频质量的客观量化评分. 使用SROCC(spearman rank order correlation coefficient)、PLCC(pearson linear correlation coefficient)和RMSE(root mean squared error)等评价指标在公开数据集KoNViD-1k、LIVE-Qualcomm和CVD2014开展实验,结果表明该方法不仅与人类主观感知具有高度相关性,且预测误差较小,有效提升了视频质量评估的准确性和可靠性,能够更贴近地模拟人类对视频质量的直观评价.
Abstract:
Due to the existence of various unknown distortion types and the lack of reference videos,it is a very challenging task to evaluate the quality of such videos. In recent years,researchers have integrated the transcendental knowledge of human visual system into the task of quality evaluation. On this basis,a non-reference video quality assessment(VQA)method considering background distortion is proposed. This method not only considers the video content,but also significantly enhances the sensitivity to the information loss in the video background,and fully considers the extraction of background features in the feature extraction stage. Then,by introducing the channel mining technology combined with gating mechanism,the high and low dimensional features are efficiently integrated,so that the feature channels can focus on the details of background distortion more accurately. Finally,the time dimension model of features is constructed by using the time series modeling module,and the objective quantitative score of video quality is generated by linear regression method. The evaluation indexes such as SROC(Spearman rank order correlation coefficient),PLCC(Pearson linear correlation coefficient)and RMSE(root mean squared error)are used to carried out experiments on KoNViD-1k,LIVE-Qualcomm and CVD2014 datasets. The results show that this method not only has a high correlation with human subjective perception,but also has a small prediction error,which effectively improves the accuracy and reliability of VQA and can more closely simulate human visual evaluation of video quality.

参考文献/References:

[1]KOU T,LIU X,SUN W,et al. Stablevqa:a deep no-reference quality assessment model for video stability[C]//Proceedings of the 31st ACM International Conference on Multimedia. Ottawa,Canada,2023:1066-1076.
[2]SUN W,MIN X,LU W,et al. A deep learning based no-reference quality assessment model for ugc videos[C]//Proceedings of the 30th ACM International Conference on Multimedia. Lisbon,Portugal,2022:856-865.
[3]ZHANG Z,SUN W,ZHU Y,et al. Evaluating point cloud from moving camera videos:A no-reference metric[J]. IEEE transactions on multimedia,2023:1-13.
[4]LI D,JIANG T,JIANG M. Quality assessment of in-the-wild videos[C]//Proceedings of the 27th ACM International Conference on Multimedia. Nice,France:Association for Computing Machinery,2019:2351-2359.
[5]LI D,JIANG T,JIANG M. Unified quality assessment of in-the-wild videos with mixed datasets training[J]. International journal of computer vision,2021,129(4):1238-1257.
[6]VARGA D,SZIRáNYI T. No-reference video quality assessment via pretrained CNN and LSTM networks[J]. Signal,image and video processing,2019,13(8):1569-1576.
[7]鄢杰斌,方玉明,刘学林. 图像质量评价研究综述——从失真的角度[J]. 中国图像图形学报,2022,27(5):1430-1466.
[8]BABU R V,SURESH S,PERKIS A. No-reference JPEG-image quality assessment using GAP-RBF[J]. Signal processing,2007,87(6):1493-1503.
[9]王怡影,李琛璐,柴豆豆 等. 基于亮度掩盖的图像质量评价模型[J]. 新乡学院学报,2021,38(3):38-41.
[10]RAMPRASAATH R S,MICHAEL C,ABHISHEK D,et al. Grad-CAM:Visual explanations from deep networks via gradient-based localization[C]//Proceedings of 2017 IEEE International Conference on Computer Vision(ICCV). Venice,Italy:IEEE,2017:618-626.
[11]程茹秋,余烨,石岱宗 等.图像与视频质量评价综述[J]. 中国图像图形学报,2022,27(5):1410-1429.
[12]AGARLA M,CELONA L,SCHETTINI R. No-reference quality assessment of in-capture distorted videos[J]. Journal of imaging,2020,6(8):74.
[13]AGARLA M,CELONA L,SCHETTINI R. An efficient method for no-reference video quality assessment[J]. Journal of imaging,2021,7(3):55.
[14]JIANG J,WANG X,LI B,et al. Multi-dimensional feature fusion network for no-reference quality assessment of in-the-wild videos[J]. Sensors,2021,21(16):5322.
[15]姚军财,申静,黄陈蓉. 基于多层 BP 神经网络的无参考视频质量客观评价[J]. 自动化学报,2022,48(2):594-607.
[16]李钟华,陈禅,王雪津,等. 基于跨网络特征补偿的声呐图像质量评价方法[J]. 北京航空航天大学学报,网络首发2024-10-16. DOI:10.13700/j.bh.1001-5965.2024.0440.
[17]YU Y,CHEN F,JU J,et al. LMT-GP:Combined latent mean-teacher and gaussian process for semi-supervised low-light image enhancement[C]//Proceeding of the 2024 European Conference on Computer Vision(ECCV). Milan,Italy,2024:261-279.
[18]LIU W T,DUANMU Z F,WANG Z. End-to-end blind quality assessment of compressed videos using deep neural networks[C]//Proceedings of the 26th ACMet international conference on Multimedia. Seoul South Korea:ACM,2018:546-554.
[19]王春峰,苏荔,张维刚,等. 基于 3D 卷积神经网络的无参考视频质量评价[J]. 软件学报,2017,27(S2):103-112.
[20]余烨,傅云翔,杨昌东,等. 基于FR-ResNet的车辆型号精细识别研究[J]. 自动化学报,2021,47(5):1125-1136.
[21]YU Y,LIU H,FU Y,et al. Embedding pose information for multiview vehicle model recognition[J]. IEEE transactions on circuits and systems for video technology(TCSVT),2022,32(8):5467-5480.
[22]HOSU V,HAHN F,JENADELEH M,et al. The konstanz natural video database(KoNViD-1k)[C]//Proceedings of 2017 Ninth International Conference on Quality of Multimedia Experience(QoMEX). Erfurt:IEEE,2017:1-6.
[23]GHADIYARAM D,PAN J,BOVIK A C,et al. In-capture mobile video distortions:a study of subjective behavior and objective algorithms[J]. IEEE transactions on circuits and systems for video technology,2018,28(9):2061-2077.
[24]NUUTINEN M,VIRTANEN T,VAAHTERANOKSA M,et al. CVD2014—A database for evaluating no-reference video quality assessment algorithms[J]. IEEE transactions on image processing,2016,25(7):3073-3086.
[25]SAAD M A,BOVIK A C,CHARRIER C. Blind prediction of natural video quality[J]. IEEE transactions on image processing,2014,23(3):1352-1365.
[26]MITTAL A,SAAD M A,BOVIK A C. 2016. A completely blind video integrity oracle[J]. IEEE transactions on image processing,2016,25(1):289-300.
[27]MITTAL A,MOORTHY A K,BOVIK A C. No-reference image quality assessment in the spatial domain[J]. IEEE transactions on image processing,2012,21(12):4695-4708.
[28]XU J T,YE P,LIU Y,DOERMANN D. No-reference video quality assessment via feature learning[C]//Proceedings of 2014 IEEE International Conference on Image Processing(ICIP). Paris,France:IEEE,2014:491-495.
[29]MITTAL A,SOUNDARARAJAN R,BOVIK A C. Making a “completely blind” image quality analyzer[J]. IEEE signal processing letters,2013,20(3):209-212.

备注/Memo

备注/Memo:
收稿日期:2024-07-22.
基金项目:安徽省自然科学基金资助项目(2308085MF216)、国家自然科学基金资助项目(62372153).
通讯作者:朱文佳,硕士,工程师,研究方向:计算机视觉与智能交通; 余烨,博士,副教授,研究方向:计算机视觉、低光图像增强及视频质量评价等. E-mail:yuye@hfut.edu.cn
更新日期/Last Update: 2025-06-20