BriefGPT.xyz
Jun, 2018
使用多尺度卷积神经网络从单个RGB图像实现准确的光谱超分辨率
Accurate Spectral Super-resolution from Single RGB Image Using Multi-scale CNN
HTML
PDF
Yiqi Yan, Lei Zhang, Wei Wei, Yanning Zhang
TL;DR
本文提出了一种基于深度卷积神经网络的多尺度图像转换方法,可以从RGB观测值中产生高分辨率的高光谱图像。在对中间特征图进行对称下采样和上采样的级联模式下,可以共同编码局部和非局部图像信息以获得更准确的光谱重建精度。
Abstract
Different from traditional hyperspectral super-resolution approaches that focus on improving the spatial resolution,
spectral super-resolution
aims at producing a high-resolution
hyperspectral image
from the RGB
→