Scholarlink

<p>The drivers’ vision is affected due to the presence of several air pollutants such as fog, haze, and smog. Several image dehazing techniques have been designed in last decade to restore visibility in diverse applications. The image quality deteriorates due to attenuation of light with increase in depth. It suffers from various adverse effects such as edge distortion, color distortion, poor contrast, and saturation. Some researchers also designed real-time video dehazing architectures but are less effective. This paper provides a thorough review and discusses current advancements in the field of visibility restoration utilizing dehazing technologies. A novel single image dehazing algorithm called ClarityGAN is designed using FRIDA, FRIDA2, RESIDE, and HUDRS. It utilizes a combination of several loss functions to optimize generator and discriminator. It outperforms several existing prior-, CNN-, and GAN-based dehazing algorithms in terms of different performance metrics. It improves <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\:\mathfrak{s}\mathfrak{s}\mathfrak{i}\mathfrak{m}\)</EquationSource> </InlineEquation>, <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\:\mathfrak{p}\mathfrak{s}\mathfrak{n}\mathfrak{r}\)</EquationSource> </InlineEquation>, <InlineEquation ID="IEq3"> <EquationSource Format="TEX">\(\:\mathfrak{p}\mathfrak{i}\mathfrak{q}\mathfrak{e}\)</EquationSource> </InlineEquation>, and <InlineEquation ID="IEq4"> <EquationSource Format="TEX">\(\:\mathfrak{b}\mathfrak{r}\mathfrak{i}\mathfrak{s}\mathfrak{q}\mathfrak{u}\mathfrak{e}\)</EquationSource> </InlineEquation> by 2.17%, 5.23%, 12.22%, and 4.07%, respectively. Also, a novel dehazing framework has been proposed for real-time dehazing to eliminate the risk of accidents in late winter nights due to foggy weather. It is designed using Camlink 4k, GoPro Hero camera, and bluerigger cable. This paper presents the designing of both static and dynamic real-time dehazing framework, along with the existing challenges and significance. It has less computation cost than few existing real-time frameworks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A novel dehazing framework for road accidents prevention

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Sunil Kumar Singla,
  • Vijay Kumar,
  • Akshay Juneja

摘要

The drivers’ vision is affected due to the presence of several air pollutants such as fog, haze, and smog. Several image dehazing techniques have been designed in last decade to restore visibility in diverse applications. The image quality deteriorates due to attenuation of light with increase in depth. It suffers from various adverse effects such as edge distortion, color distortion, poor contrast, and saturation. Some researchers also designed real-time video dehazing architectures but are less effective. This paper provides a thorough review and discusses current advancements in the field of visibility restoration utilizing dehazing technologies. A novel single image dehazing algorithm called ClarityGAN is designed using FRIDA, FRIDA2, RESIDE, and HUDRS. It utilizes a combination of several loss functions to optimize generator and discriminator. It outperforms several existing prior-, CNN-, and GAN-based dehazing algorithms in terms of different performance metrics. It improves \(\:\mathfrak{s}\mathfrak{s}\mathfrak{i}\mathfrak{m}\) , \(\:\mathfrak{p}\mathfrak{s}\mathfrak{n}\mathfrak{r}\) , \(\:\mathfrak{p}\mathfrak{i}\mathfrak{q}\mathfrak{e}\) , and \(\:\mathfrak{b}\mathfrak{r}\mathfrak{i}\mathfrak{s}\mathfrak{q}\mathfrak{u}\mathfrak{e}\) by 2.17%, 5.23%, 12.22%, and 4.07%, respectively. Also, a novel dehazing framework has been proposed for real-time dehazing to eliminate the risk of accidents in late winter nights due to foggy weather. It is designed using Camlink 4k, GoPro Hero camera, and bluerigger cable. This paper presents the designing of both static and dynamic real-time dehazing framework, along with the existing challenges and significance. It has less computation cost than few existing real-time frameworks.