Scholarlink

<p>To tackle the challenges of noise and insufficient contrast in enhancing low-light images, this paper introduces a low-light image enhancement algorithm, which is based on adaptive weight fusion denoising in HSV space. In HSV space, the saturation (S) channel is enhanced by bilateral filtering and contrast stretching to enhance the saturation and contrast of the image. The core V channel is then processed using three different enhancement strategies: logarithmic transformation (Log), adaptive gamma correction (AGC) and contrast limited adaptive histogram equalization (CLAHE), the low-exposure weight map is constructed adaptively based on the difference between the pixel values and the low-exposure level, and the improved multi-branch weight fusion is performed at different pyramid levels, and the image pyramid is applied for multi-scale reconstruction denoising to optimize the image information at different scales. Ultimately, the processed image is transformed back into the RGB color space for display purposes. The experimental findings demonstrate that this approach significantly enhances image brightness, contrast and noise reduction capabilities in low-light environments, resulting in superior visual quality.</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

An adaptive weight fusion low-light image enhancement based on HSV space

  • 期刊论文

  • 发表日期：

    2025年8月21日

查看全文
  • Kai Jiang,
  • Baoju Zhang,
  • Bo Zhang,
  • Cuiping Zhang,
  • Bohua Chu,
  • Yameng Zhang,
  • Mengqi Xue

摘要

To tackle the challenges of noise and insufficient contrast in enhancing low-light images, this paper introduces a low-light image enhancement algorithm, which is based on adaptive weight fusion denoising in HSV space. In HSV space, the saturation (S) channel is enhanced by bilateral filtering and contrast stretching to enhance the saturation and contrast of the image. The core V channel is then processed using three different enhancement strategies: logarithmic transformation (Log), adaptive gamma correction (AGC) and contrast limited adaptive histogram equalization (CLAHE), the low-exposure weight map is constructed adaptively based on the difference between the pixel values and the low-exposure level, and the improved multi-branch weight fusion is performed at different pyramid levels, and the image pyramid is applied for multi-scale reconstruction denoising to optimize the image information at different scales. Ultimately, the processed image is transformed back into the RGB color space for display purposes. The experimental findings demonstrate that this approach significantly enhances image brightness, contrast and noise reduction capabilities in low-light environments, resulting in superior visual quality.