Spectrum Reconstruction of the Light-Field Multimodal Imager

Su, Lijuan; Liu, Yujian; Yuan, Yan

doi:10.1109/access.2018.2890458

Cited by 5 publications

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional imaging spectrometers have to scan either spatial dimension or spectral dimension 4 , which inevitably sacrifices the resolution of time dimension and requires high stability of scanning process. To mitigate the above problems, various snapshot hyperspectral imaging spectrometers have been proposed, such as, coded aperture snapshot spectral imaging (CASSI) 5 , computed tomography imaging spectrometry (CTIS) 6 , light field imaging system(LFIS) 7 , and image mapping spectrometry (IMS) 8 .…”

Section: Introductionmentioning

confidence: 99%

An inpainting and super resolution method for image mapping spectrometer

Shao¹,

Su²,

Liu³

et al. 2023

SPIE Future Sensing Technologies 2023

View full text Add to dashboard Cite

Image Mapping Spectrometry (IMS) is a compact snapshot hyperspectral imaging technology. However, the image mapper used in the IMS causes degradation of the reconstructed spectral datacube, such as, low spatial resolution, missing areas and stripe artifacts. In this paper, we propose an end-to-end deep learning method to jointly inpainting and super resolution the restored spectral images of the IMS. The method includes an image inpainting network, which is designed to correct the nonuniform intensity and missing data, and an image super resolution network, which aims to enhance the spatial resolution of images. In addition, a local nonuniformity correction method is proposed to preprocess the IMS images. Simulation and experimental results demonstrate the effectiveness of the proposed method.

show abstract

Section: Introductionmentioning

confidence: 99%

An inpainting and super resolution method for image mapping spectrometer

Shao¹,

Su²,

Liu³

et al. 2023

SPIE Future Sensing Technologies 2023

View full text Add to dashboard Cite

show abstract

“…Furthermore, research on dimensionality reduction for hyperspectral images has been conducted, which aids better analysis of the spectral information required in applications [11], [12]. Light-field imaging devices have been developed for traditional modeling of spectral and polarimetric radiance [13].…”

Section: Introductionmentioning

confidence: 99%

U-Net-Based Multispectral Image Generation From an RGB Image

Zeng

Diao

2021

IEEE Access

View full text Add to dashboard Cite

Multispectral images have lower spatial resolution than RGB images. It is difficult to obtain multispectral images with both high spatial resolution and high spectral resolution because of expensive capture setup and sophisticated acquisition processes. In this paper, we propose a deep neural network structure based on U-Net to convert ordinary RGB images into multispectral images with high spectral resolution. Our variant U-Net neural network structure not only preserves detailed features of RGB images, but also promotes the fusion of different feature scales, enhancing the quality of multispectral image generation. Apart from the training stage, our proposed method does not require low-resolution multispectral images, as do some earlier learning-based methods; multispectral images can be obtained using only the corresponding RGB high-resolution images. We also employ the Inception block to achieve richer image features and the feature loss function to optimize the non-local features. Our proposed algorithm achieves state-of-the-art visual effects and quantitative measurements such as RMSE and rRMSE on several different public datasets.

show abstract