StereoMSI

StereoMSI数据集是由澳大利亚联邦科学与工业研究组织DATA61创建，专为基于示例的单色和颜色引导的频谱图像超分辨率研究而设计。该数据集包含350对注册的颜色-频谱图像，首次在PIRM2018频谱图像超分辨率挑战中发布。数据集内容涵盖多种场景，从工业环境到办公室，从沙漠到雨林。创建过程中，特别注意了曝光时间和图像质量，使用不同的相机捕捉立体对。该数据集适用于开发频谱图像超分辨率方法，以及其他多种学习型应用，如疾病诊断和食品监测等。

The StereoMSI Dataset was created by DATA61, Commonwealth Scientific and Industrial Research Organisation (CSIRO) of Australia, and is specifically designed for example-based monochromatic and color-guided spectral image super-resolution research. This dataset contains 350 pairs of registered color-spectral images, and was first released at the PIRM2018 Spectral Image Super-Resolution Challenge. The dataset covers a diverse range of scenarios, spanning from industrial environments and offices to deserts and rainforests. During its development, special attention was paid to exposure time and image quality, with stereo pairs captured using different cameras. This dataset is suitable for developing spectral image super-resolution methods and various other learning-based applications such as disease diagnosis and food monitoring.