Scholarlink

Image detail enhancement techniques aim to improve the visual quality of the original image. Traditional methods decompose the image into smooth and detail layers, subsequently amplifying the detail signals to achieve image enhancement. The principal innovation of this paper is a novel strategy for detail layer extraction and fusion, namely TaylorNet, which combines a Taylor series decomposition of the signal with a weighted residual technique. Through the implementation of ablation experiments, we not only optimize the experimental parameters but also rigorously validate the effectiveness and accuracy of the proposed method. Extensive experimental evaluations demonstrate that the proposed technique achieves superior enhancement results on datasets.

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

TaylorNet: Single Image Detail Enhancement Based on Taylor Series Decomposition

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • He Jiang,
  • Zhou Zheng,
  • Mang Sun,
  • Hao Gu,
  • Fudi Yi,
  • Wen Cui,
  • Chang Liu,
  • Baihui Deng,
  • Bingyuan Su,
  • Qiwei Dong,
  • Dongdong Guo

摘要

Image detail enhancement techniques aim to improve the visual quality of the original image. Traditional methods decompose the image into smooth and detail layers, subsequently amplifying the detail signals to achieve image enhancement. The principal innovation of this paper is a novel strategy for detail layer extraction and fusion, namely TaylorNet, which combines a Taylor series decomposition of the signal with a weighted residual technique. Through the implementation of ablation experiments, we not only optimize the experimental parameters but also rigorously validate the effectiveness and accuracy of the proposed method. Extensive experimental evaluations demonstrate that the proposed technique achieves superior enhancement results on datasets.