可解释优化启发的展开网络用于低光图像增强
相关实验视频
Updated: Jun 23, 2026

11:23
Lensless Fluorescent Microscopy on a Chip
Published on: August 17, 2011
17.6K
概括
本研究介绍了URetinex-Net++,一种适应性深度学习网络,用于低光图像增强 (LLIE). 它通过有效分解图像来提高图像质量,同时保留细节和抑制噪音.
科学领域:
- 计算机视觉 计算机视觉
- 图像处理 图像处理
- 深度学习 (Deep Learning) 是一种深度学习.
背景情况:
- 基于Retinex的方法对低光图像增强 (LLIE) 有效,但由于手工制作的先验和传统优化,缺乏适应性和效率.
- 现有的方法在适应性层分解方面扎,以获得最佳的LLIE.
研究的目的:
- 提出一个基于Retinex的自适应深部展开网络 (URetinex-Net++),以实现高效和有效的低光图像增强.
- 通过解决颜色缺陷和提高整体性能来改进以前的URetinex-Net.
主要方法:
- 开发了URetinex-Net++,一个深度展开的网络,使用隐式先验将低光图像分解为反射率和照明层.
- 引入了三个基于学习的模块,用于数据依赖的初始化,高效的展开优化和组件调整.
- 包含一个跨阶段的融合块来减轻颜色缺陷.
主要成果:
- 拟议的展开优化模块适应地适应隐含的先验,使噪声抑制和细节保存成为可能.
- 与最先进的方法相比,URetinex-Net++在视觉质量和定量指标上都表现出卓越的性能.
- 增强的网络实现了性能提升,参数最小,计算成本低.
结论:
- URetinex-Net++通过利用自适应深度展开提供了一种有效和高效的解决方案,用于在低光下增强图像.
- 该方法在具有挑战性的低光条件下显著提高了图像质量和细节保存.
- 广泛的实验证实了URetinex-Net++在现有LLIE技术上的优势.
相关概念视频
Light Acquisition
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
Reducing Line Loss
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss in...
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss in...

