Shadow removal from VNIR hyperspectral remote sensing imagery with endmember signature analysis

Highlights:Graphical/Tabular Abstract  Shadow detection from LiDAR and hyperspectral data  Conversion from radiance domain to reflectance domain based on the physical radiance model  Analysis of the reflection and radiance values in shadow areasTargets hidden in shadow areas are very difficult to detect from hyperspectral images. One way to mitigate this problem is to detect the shadow areas using light detection and ranging (LiDAR) sensors, and to correct for the hyperspectral data in these regions. For this purpose, this paper makes a detailed analysis of real data in shadow regions and compares it the data from open areas where there is no shadow present. Figure A. Shadow analysis using hyperspectral and LiDAR dataPurpose:In this study, the radiance and reflectance values of the shadow regions in hyperspectral data, and their relation to the sky-view factors in a physical model are investigated. With this work, our goal is to understand the physical phenomenon in hyperspectral images that occurs in shadow areas. Theory and Methods:First, shadow regions are found using LiDAR data and their radiance and reflectance values are plotted. Then, sky-view factors are obtained from LiDAR data. Third, using the Modtran software, sky radiance and sun irradiance values are obtained and the radiance values in the shadow region are examined. Lastly, the parameters obtained from the Modtran software as well as the data collected from shadow areas are used in the physical radiance model; and the reflectance values in shadowy areas are analyzed. Results:Path radiance, sky radiance and radiance from adjacent objects were analyzed from three datasets with shadow locations and from open areas where there is no shadow. The physical radiance model as well as the analysis of the real data confirmed that the photons reflected from the objects surronding the shadow region are more dominant than the sky-view factors. Conclusion:Upon investigating the reflectance and radiance values in comparison to the physical radiance model, it was found that the sky-view factors are not as dominant as the photons reflected from the objects surronding the shadow region.

show abstract

“…Omruuzun vd. [15] ve Bernstein vd. [16] ise gölgelik ve güneşlik alanların spektral imzalarını incelemişlerdir.…”

Section: Gi̇ri̇ş ( Introduction )unclassified

Gölgelik Alanlarda Komşu Nesnelerin Işımaya Olan Etkisinin Gerçek Veriler ve Fiziksel Işıma Modeli Üzerinden İncelenmesi

Yüksel

Oduncu

2018

Gazi Üniversitesi Mühendislik-Mimarlık Fakültesi Dergisi

View full text Add to dashboard Cite

show abstract

Hyperspectral image shadow compensation via cycle-consistent adversarial networks

2021

View full text Add to dashboard Cite

Shadow-Aware Nonlinear Spectral Unmixing for Hyperspectral Imagery

Scheunders

Cerra

Müller

2022

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

In hyperspectral imagery, differences in ground surface structures cause a large variation in the optical scattering in sunlit and (partly) shadowed pixels. The complexity of the scene demands a general spectral mixture model that can adapt to the different scenarios of the ground surface. In this article, we propose a physics-based spectral mixture model, i.e., the extended shadow multilinear mixing (ESMLM) model that accounts for typical ground scenarios in the presence of shadows and nonlinear optical effects, by considering multiple illumination sources. Specifically, the diffuse solar illumination alters as the wavelength changes, requiring a wavelength-dependent modeling of shadows. Moreover, we allow different types of nonlinear interactions for different illumination conditions. The proposed model is described in a graph-based representation, which sums up all possible radiation paths initiated by the illumination sources. Physical assumptions are made to simplify the proposed model, resulting in material abundances and four physically interpretable parameters. Additionally, shadow-removed images can be reconstructed. The proposed model is compared with other state-of-the-art models using one synthetic dataset and two real datasets. Experimental results show that the ESMLM model performs robustly in various illumination conditions. In addition, the physically interpretable parameters contain valuable information on the scene structures and assist in performing shadow removal that outperforms other state-of-the-art works.

show abstract

Shadow removal from VNIR hyperspectral remote sensing imagery with endmember signature analysis

Cited by 5 publications

References 17 publications

Gölgelik Alanlarda Komşu Nesnelerin Işımaya Olan Etkisinin Gerçek Veriler ve Fiziksel Işıma Modeli Üzerinden İncelenmesi

Gölgelik Alanlarda Komşu Nesnelerin Işımaya Olan Etkisinin Gerçek Veriler ve Fiziksel Işıma Modeli Üzerinden İncelenmesi

Hyperspectral image shadow compensation via cycle-consistent adversarial networks

Shadow-Aware Nonlinear Spectral Unmixing for Hyperspectral Imagery

Contact Info

Product

Resources

About