Author:
Ye Zhongchang,Ye Xin,Zhao Zhonghua
Abstract
Intelligent video surveillance (IVS) technology is widely used in various security systems. However, quality degradation in surveillance images (SIs) may affect its performance on vision-based tasks, leading to the difficulties in the IVS system extracting valid information from SIs. In this paper, we propose a hybrid no-reference image quality assessment (NR IQA) model for SIs that can help to identify undesired distortions and provide useful guidelines for IVS technology. Specifically, we first extract two main types of quality-aware features: the low-level visual features related to various distortions, and the high-level semantic information, which is extracted by a state-of-the-art (SOTA) vision transformer backbone. Then, we fuse these two kinds of features into the final quality-aware feature vector, which is mapped into the quality index through the feature regression module. Our experimental results on two surveillance content quality databases demonstrate that the proposed model achieves the best performance compared to the SOTA on NR IQA metrics.
Reference64 articles.
1. Gonzalez-Cepeda, J., Ramajo, A., and Armingol, J.M. (2022). Intelligent Video Surveillance Systems for Vehicle Identification Based on Multinet Architecture. Information, 13.
2. Intelligent video surveillance: A review through deep learning techniques for crowd analysis;Sreenu;J. Big Data,2019
3. Muller-Schneiders, S., Jager, T., Loos, H.S., and Niem, W. (2005, January 15–16). Performance evaluation of a real time video surveillance system. Proceedings of the 2005 IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance, Beijing, China.
4. Aqqa, M., Mantini, P., and Shah, S.K. (2019, January 25–27). Understanding How Video Quality Affects Object Detection Algorithms. Proceedings of the VISIGRAPP (5: VISAPP), Prague, Czech Republic.
5. Intelligent video surveillance;Held;Computer,2012
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献