Degradation Type-Aware Image Restoration for Effective Object Detection in Adverse Weather

被引:0
|
作者
Huang, Xiaochen [1 ,2 ]
Wang, Xiaofeng [1 ]
Teng, Qizhi [1 ]
He, Xiaohai [1 ]
Chen, Honggang [1 ,3 ]
机构
[1] Sichuan Univ, Coll Elect & Informat Engn, Chengdu 610065, Peoples R China
[2] Tianjin Univ Technol, Key Lab Comp Vis & Syst, Minist Educ, Tianjin 300384, Peoples R China
[3] Yunnan Univ, Yunnan Key Lab Software Engn, Kunming 650600, Peoples R China
关键词
object detection in various weather scenes; restoration-assisted object detection; degradation type awareness; multi-task joint learning;
D O I
10.3390/s24196330
中图分类号
O65 [分析化学];
学科分类号
070302 ; 081704 ;
摘要
Despite significant advancements in CNN-based object detection technology, adverse weather conditions can disrupt imaging sensors' ability to capture clear images, thereby adversely impacting detection accuracy. Mainstream algorithms for adverse weather object detection enhance detection performance through image restoration methods. Nevertheless, the majority of these approaches are designed for a specific degradation scenario, making it difficult to adapt to diverse weather conditions. To cope with this issue, we put forward a degradation type-aware restoration-assisted object detection network, dubbed DTRDNet. It contains an object detection network with a shared feature encoder (SFE) and object detection decoder, a degradation discrimination image restoration decoder (DDIR), and a degradation category predictor (DCP). In the training phase, we jointly optimize the whole framework on a mixed weather dataset, including degraded images and clean images. Specifically, the degradation type information is incorporated in our DDIR to avoid the interaction between clean images and the restoration module. Furthermore, the DCP makes the SFE possess degradation category awareness ability, enhancing the detector's adaptability to diverse weather conditions and enabling it to furnish requisite environmental information as required. Both the DCP and the DDIR can be removed according to requirement in the inference stage to retain the real-time performance of the detection algorithm. Extensive experiments on clear, hazy, rainy, and snowy images demonstrate that our DTRDNet outperforms advanced object detection algorithms, achieving an average mAP of 79.38% across the four weather test sets.
引用
收藏
页数:23
相关论文
共 50 条
  • [1] Learning degradation-aware visual prompt for maritime image restoration under adverse weather conditions
    He, Xin
    Jia, Tong
    Li, Junjie
    FRONTIERS IN MARINE SCIENCE, 2024, 11
  • [2] Joint image restoration for object detection in snowy weather
    Wang, Jing
    Xu, Meimei
    Xue, Huazhu
    Huo, Zhanqiang
    Luo, Fen
    IET COMPUTER VISION, 2024, 18 (06) : 759 - 771
  • [3] MULTI-WEATHER DEGRADATION-AWARE TRANSFORMER FOR IMAGE RESTORATION
    Zhu, Ruoxi
    Wu, Minfeng
    Xiong, Xiankui
    Zhu, Xuanpeng
    Fan, Yibo
    2024 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING, ICASSP 2024, 2024, : 3765 - 3769
  • [4] Fog-Aware Adaptive YOLO for Object Detection in Adverse Weather
    Abbasi, Hasan
    Amini, Marzieh
    Yu, F. Richard
    2023 IEEE SENSORS APPLICATIONS SYMPOSIUM, SAS, 2023,
  • [5] Pay "Attention" to Adverse Weather: Weather-aware Attention-based Object Detection
    Chaturvedi, Saket S.
    Zhang, Lan
    Yuan, Xiaoyong
    2022 26TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR), 2022, : 4573 - 4579
  • [6] Localization-aware logit mimicking for object detection in adverse weather conditions
    Luo, Peiyun
    Nie, Jing
    Xie, Jin
    Cao, Jiale
    Zhang, Xiaohong
    IMAGE AND VISION COMPUTING, 2024, 146
  • [7] Image-Adaptive YOLO for Object Detection in Adverse Weather Conditions
    Liu, Wenyu
    Ren, Gaofeng
    Yu, Runsheng
    Guo, Shi
    Zhu, Jianke
    Zhang, Lei
    THIRTY-SIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FOURTH CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE / THE TWELVETH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2022, : 1792 - 1800
  • [8] MWFormer: Multi-Weather Image Restoration Using Degradation-Aware Transformers
    Zhu, Ruoxi
    Tu, Zhengzhong
    Liu, Jiaming
    Bovik, Alan C.
    Fan, Yibo
    IEEE TRANSACTIONS ON IMAGE PROCESSING, 2024, 33 : 6790 - 6805
  • [9] Multiple adverse weather image restoration: A review
    Xiao, Hanguang
    Liu, Shihong
    Zuo, Kun
    Xu, Haipeng
    Cai, Yuyang
    Liu, Tianqi
    Yang, Zhiying
    NEUROCOMPUTING, 2025, 618
  • [10] Prompt-guided and degradation prior supervised transformer for adverse weather image restoration
    Liu, Weihan
    Shao, Mingwen
    Meng, Lingzhuang
    Qiao, Yuanjian
    Bao, Zhiyuan
    APPLIED INTELLIGENCE, 2025, 55 (02)