从RGB图像中重建高光谱图像,通过多尺度光谱空间序列学习
Wenjing Chen1,2, Lang Liu2, Rong Gao1,2
1Hubei Provincial Key Laboratory of Green Intelligent Computing Power Network, Hubei University of Technology, Wuhan 430068, China.
Entropy (Basel, Switzerland)
|September 27, 2025
概括
本研究介绍了MSS-Mamba用于从RGB图像中进行超光谱图像重建,通过高效的远程依赖模型增强光谱超分辨率 (SSR). 该方法通过整合多个尺度的光谱空间信息来实现高保真度的结果.
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:
- 变压器模型具有先进的高光谱图像重建 (光谱超分辨率,SSR).
- 现有的变压器方法在平衡计算效率和远程特征提取方面面临挑战.
- 马巴为远程依赖提供线性复杂性,并在视觉任务中显示出前景.
研究的目的:
- 提出MSS-Mamba,一种新的多尺度光谱空间序列学习方法,用于从RGB图像中重建超光谱图像.
- 通过改进特征提取和多层次信息处理来增强Mamba在SSR方面的能力.
主要方法:
- 引入了连续光谱空间扫描 (CS3) 机制,以增强Mamba的跨维特征提取.
- 开发了一种序列代币化策略,使用一个多尺度信息融合 (MIF) 模块来解决层次化的多尺度学习限制.
- MIF模块使用双分支架构进行单独的全球和本地处理,通过自适应路由器实现动态功能融合.
主要成果:
- 拟议的MSS-Mamba方法有效地从RGB输入中重建高光谱图像.
- 在ARAD_1k,CAVE和grss_dfc_2018数据集上的实验结果证明了该方法的卓越性能.
- 这种方法成功地生成了具有全球上下文和局部细节的特征地图,用于高保真重建.
结论:
- 使用Mamba的MSS-Mamba提供了一种高效和有效的解决方案,用于使用Mamba的光谱超分辨率.
- 新的CS3机制和MIF模块显著提高了高光谱图像重建质量.
- 这项工作通过利用高效的序列建模来完成复杂的重建任务,推动了超光谱成像领域的发展.
相关概念视频
Upsampling
583
Managing signal sampling rates is essential in digital signal processing to maintain signal integrity. A decimated signal, characterized by a reduced frequency range due to its lower sampling rate, can be upsampled by inserting zeros between each sample. This upsampling process expands the original spectrum and introduces repeated spectral replicas at intervals dictated by the new Nyquist frequency. To refine this zero-inserted sequence, it is passed through a lowpass filter with a cutoff...
583
Downsampling
609
When considering a sampled sequence with zero values between sampling instants, one can replace it by taking every N-th value of the sequence. At these integer multiples of N, the original and sampled sequences coincide. This process, known as decimation, involves extracting every N-th sample from a sequence, thereby creating a more efficient sequence.
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
609


