{"title":"HazeTrendNet：通过雾浓度趋势指导的单图像去雾","authors":"Chen Wang, Yuanyuan Fan","doi":"10.1049/ell2.70396","DOIUrl":null,"url":null,"abstract":"<p>Single-image dehazing is vital for restoring clear visuals from haze-affected images in applications like surveillance and autonomous driving. Most existing models struggle with local haze variations and detail preservation in dense haze. This study introduces HazeTrendNet, a lightweight dehazing framework, incorporating haze-concentration-trend guidance via transmittance estimation, dynamic convolution kernel selection and haze-aware attention. Experiments show state-of-the-art performance: PSNR/SSIM of 42.29 dB/0.997 on RESIDE-Indoor, 17.68 dB/0.655 on Dense-Haze and 22.14 dB/0.812 on NH-HAZE, outperforming EENet, SANet and FocalNet. With 6.17 M parameters and 56.46 GFLOPs, it suits edge deployment.</p>","PeriodicalId":11556,"journal":{"name":"Electronics Letters","volume":"61 1","pages":""},"PeriodicalIF":0.8000,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/ell2.70396","citationCount":"0","resultStr":"{\"title\":\"HazeTrendNet: Single-Image Dehazing via Haze-Concentration-Trend Guidance\",\"authors\":\"Chen Wang, Yuanyuan Fan\",\"doi\":\"10.1049/ell2.70396\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Single-image dehazing is vital for restoring clear visuals from haze-affected images in applications like surveillance and autonomous driving. Most existing models struggle with local haze variations and detail preservation in dense haze. This study introduces HazeTrendNet, a lightweight dehazing framework, incorporating haze-concentration-trend guidance via transmittance estimation, dynamic convolution kernel selection and haze-aware attention. Experiments show state-of-the-art performance: PSNR/SSIM of 42.29 dB/0.997 on RESIDE-Indoor, 17.68 dB/0.655 on Dense-Haze and 22.14 dB/0.812 on NH-HAZE, outperforming EENet, SANet and FocalNet. With 6.17 M parameters and 56.46 GFLOPs, it suits edge deployment.</p>\",\"PeriodicalId\":11556,\"journal\":{\"name\":\"Electronics Letters\",\"volume\":\"61 1\",\"pages\":\"\"},\"PeriodicalIF\":0.8000,\"publicationDate\":\"2025-08-23\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/ell2.70396\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Electronics Letters\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/ell2.70396\",\"RegionNum\":4,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Electronics Letters","FirstCategoryId":"5","ListUrlMain":"https://ietresearch.onlinelibrary.wiley.com/doi/10.1049/ell2.70396","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
HazeTrendNet: Single-Image Dehazing via Haze-Concentration-Trend Guidance
Single-image dehazing is vital for restoring clear visuals from haze-affected images in applications like surveillance and autonomous driving. Most existing models struggle with local haze variations and detail preservation in dense haze. This study introduces HazeTrendNet, a lightweight dehazing framework, incorporating haze-concentration-trend guidance via transmittance estimation, dynamic convolution kernel selection and haze-aware attention. Experiments show state-of-the-art performance: PSNR/SSIM of 42.29 dB/0.997 on RESIDE-Indoor, 17.68 dB/0.655 on Dense-Haze and 22.14 dB/0.812 on NH-HAZE, outperforming EENet, SANet and FocalNet. With 6.17 M parameters and 56.46 GFLOPs, it suits edge deployment.
期刊介绍：
Electronics Letters is an internationally renowned peer-reviewed rapid-communication journal that publishes short original research papers every two weeks. Its broad and interdisciplinary scope covers the latest developments in all electronic engineering related fields including communication, biomedical, optical and device technologies. Electronics Letters also provides further insight into some of the latest developments through special features and interviews.
Scope
As a journal at the forefront of its field, Electronics Letters publishes papers covering all themes of electronic and electrical engineering. The major themes of the journal are listed below.
Antennas and Propagation
Biomedical and Bioinspired Technologies, Signal Processing and Applications
Control Engineering
Electromagnetism: Theory, Materials and Devices
Electronic Circuits and Systems
Image, Video and Vision Processing and Applications
Information, Computing and Communications
Instrumentation and Measurement
Microwave Technology
Optical Communications
Photonics and Opto-Electronics
Power Electronics, Energy and Sustainability
Radar, Sonar and Navigation
Semiconductor Technology
Signal Processing
MIMO
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
