摘要
目前光学成像技术已经在深海探测中发挥重要的作用,但仍缺少深海视频质量主观评价研究,尤其是缺少公开的深海视频质量评价数据集。为此,构建了一个公开的、带有主观质量标签的深海视频质量评价数据集,该数据集包括5类代表性的真实深海场景视频。为了实现数据增广,使用基于深度学习和基于融合的水下图像增强方法进行视频质量增强,使用高斯模糊和高斯噪声进行视频质量退化;采用单激励绝对等级主观质量评价方法对深海视频进行视频质量评价,主观评价实验人数为20,得到总数量为142的深海视频质量评价数据集。在该数据集上验证了8种图像/视频质量客观评价模型的性能,结果显示当前视频质量客观评价模型用于深海视频质量评价还需提升性能。数据集公开在http://ieeedataport.org/documents/deepseavideoqualitydataset,有助于深海视频质量客观评价和增强技术的优化和改进。
Currently,optical imaging technology has played an important role in deepsea exploration.However,there is still a lack of research on subjective deepsea video quality assessment,especially the lack of public deepsea video quality assessment datasets.We construct a public deepsea video quality assessment dataset with subjective quality labels,which includes five types of representative real deepsea scene videos.The original deepsea video sequences are augmented by two deepsea video quality enhancement methods that are based on deep learning and fusion respectively,and two video quality degradation methods including Gaussian blurring and Gaussian noise.Subjective video quality assessment is conducted with 20 participants and the absolute category rating method is used for rating.Finally,we obtain a deepsea video quality assessment dataset with 142 samples.The performance of 8 objective image/video quality assessment models is verified on this dataset.The results show that the current objective video quality assessment models need to be improved for the application in deepsea video quality assessment.The deepsea video quality assessment dataset is publicly available at http://ieeedataport.org/documents/deepseavideoqualitydataset.It could help optimize and improve the objective deepsea video quality assessment models and underwater image/video enhancement technology.
作者
宋巍
刘晓晨
黄冬梅
孙科林
张兵
Song Wei;Liu Xiaochen;Huang Dongmei;Sun Kelin;Zhang Bing(College of Information,Shanghai Ocean University,Shanghai 201306,China;Shanghai University of Electric Power,Shanghai 201306,China;Institute of DeepSea Science and Engineering,Chinese Academy of Sciences,Sanya 572000,Hainan,China)
出处
《激光与光电子学进展》
CSCD
北大核心
2022年第17期409-418,共10页
Laser & Optoelectronics Progress
基金
国家自然科学基金(61972240,61702323)
上海市科委部分地方高校能力建设项目(20050501900)。
关键词
视觉
深海视频数据集
视频质量主观评价
数据增广
视频质量客观评价模型
vision
deepsea video dataset
subjective video quality assessment
data augmentation
objective video quality assessment model