Illumination and shadow compensation of hyperspectral images using a digital surface model and non-linear least squares estimation

Friman, Ola; Tolt, Gustav; Ahlberg, Jörgen

doi:10.1117/12.898084

Cited by 25 publications

(16 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Several parameters for thresholding and atmospheric conditions must be carefully defined in this approach. When a DSM is available, ray tracing is effective for detecting shadows [14], [23], [24]; however, it is necessary to identify the area containing exactly the same objects as those in the shadow area for de-shadowing, which may be challenging in the case of a forest. We adopt an unmixing-based approach for shadow correction [25], since it can be fully automated and can effectively improve the classification accuracy without requiring the blue-skewed skylight illumination of the shadow.…”

Section: A Shadow Correctionmentioning

confidence: 99%

Hyperspectral Tree Species Classification of Japanese Complex Mixed Forest With the Aid of Lidar Data

Matsuki

Yokoya

Iwasaki

2015

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

View full text Add to dashboard Cite

The classification of tree species in forests is an important task for forest maintenance and management. With the increase in the spatial resolution of remote sensing imagery, individual tree classification is the next target of research area for the forest inventory. In this work, we propose a methodology involving the combination of hyperspectral and LiDAR data for individual tree classification, which can be extended to areas of shadow caused by the illumination of tree crowns with sunlight. To remove the influence of shadows in hyperspectral data, an unmixing-based correction is applied as preprocessing. Spectral features of trees are obtained by principal component analysis of the hyperspectral data. The sizes and shapes of individual trees are derived from the LiDAR data after individual tree-crown delineation. Both spectral and tree-crown features are combined and input into a support vector machine classifier pixel by pixel. This procedure is applied to data taken over Tama Forest Science Garden in Tokyo, Japan, to classify it into 16 classes of tree species. It is found that both shadow correction and tree-crown information improve the classification performance, which is further improved by postprocessing based on tree-crown information derived from the LiDAR data. Regarding the classification results in the case of 10% training data, when using the random sampling of pixels to select training samples, a classification accuracy of 82% was obtained, while the use of reference polygons as a more practical means of sample selection reduced the accuracy to 71%. These values are, respectively, 21.5% and 9% higher than those that are obtained using hyperspectral data only.

show abstract

Section: A Shadow Correctionmentioning

confidence: 99%

Hyperspectral Tree Species Classification of Japanese Complex Mixed Forest With the Aid of Lidar Data

Matsuki

Yokoya

Iwasaki

2015

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

View full text Add to dashboard Cite

show abstract

“…Illumination effects in hyperspectral imagery can be suppressed in pre-processing by converting radiance or reflectance information to spectral angles and linearly correcting them based on shadow/non-shadow statistics [36]. A more intricate method to compensate for the effect of shadow in hyperspectral urban imagery using LiDAR information has been proposed in [37]. Here, a DSM and other LiDAR derived features such as sky view factor are used to model different irradiance components which are in turn entered in a non-linear spectral correction model.…”

Section: Introductionmentioning

confidence: 99%

Synergistic Use of LiDAR and APEX Hyperspectral Data for High-Resolution Urban Land Cover Mapping

Priem

Canters

2016

Remote Sensing

View full text Add to dashboard Cite

Land cover mapping of the urban environment by means of remote sensing remains a distinct challenge due to the strong spectral heterogeneity and geometric complexity of urban scenes. Airborne imaging spectroscopy and laser altimetry have each made remarkable contributions to urban mapping but synergistic use of these relatively recent data sources in an urban context is still largely underexplored. In this study a synergistic workflow is presented to cope with the strong diversity of materials in urban areas, as well as with the presence of shadow. A high-resolution APEX hyperspectral image and a discrete waveform LiDAR dataset covering the Eastern part of Brussels were made available for this research. Firstly, a novel shadow detection method based on LiDAR intensity-APEX brightness thresholding is proposed and compared to commonly used approaches for shadow detection. A combination of intensity-brightness thresholding with DSM model-based shadow detection is shown to be an efficient approach for shadow mask delineation. To deal with spectral similarity of different types of urban materials and spectral distortion induced by shadow cover, supervised classification of shaded and sunlit areas is combined with iterative LiDAR post-classification correction. Results indicate that height, slope and roughness features contribute to improved classification accuracies in descending order of importance. Results of this study illustrate the potential of synergistic application of hyperspectral imagery and LiDAR for urban land cover mapping.

show abstract

“…This is because atmospheric absorption features are not accounted for by the approximation and have a large impact on the incident light spectra. Finally, multimodal approaches [8,19] which use additional sensors such as Light Detection and Ranging (LiDAR) and Global Positioning Systems (GPS) can form geometry-based illumination models of the scene and compensate for the variations in lighting, however, the additional sensors required in these approaches are not always available.…”

Section: Illumination Invariance In Hyperspectral Classificationmentioning

confidence: 99%

Hyperspectral CNN Classification with Limited Training Samples

Windrim¹,

Ramakrishnan²,

Melkumyan³

et al. 2017

Procedings of the British Machine Vision Conference 2017

View full text Add to dashboard Cite

Hyperspectral imaging sensors are becoming increasingly popular in robotics applications such as agriculture and mining, and allow per-pixel thematic classification of materials in a scene based on their unique spectral signatures. Recently, convolutional neural networks have shown remarkable performance for classification tasks, but require substantial amounts of labelled training data. This data must sufficiently cover the variability expected to be encountered in the environment. For hyperspectral data, one of the main variations encountered outdoors is due to incident illumination, which can change in spectral shape and intensity depending on the scene geometry. For example, regions occluded from the sun have a lower intensity and their incident irradiance skewed towards shorter wavelengths.In this work, a data augmentation strategy based on relighting is used during training of a hyperspectral convolutional neural network. It allows training to occur in the outdoor environment given only a small labelled region, which does not need to sufficiently represent the geometric variability of the entire scene. This is important for applications where obtaining large amounts of training data is labourious, hazardous or difficult, such as labelling pixels within shadows. Radiometric normalisation approaches for preprocessing the hyperspectral data are analysed and it is shown that methods based on the raw pixel data are sufficient to be used as input for the classifier. This removes the need for external hardware such as calibration boards, which can restrict the application of hyperspectral sensors in robotics applications. Experiments to evaluate the classification system are carried out on two datasets captured from a field-based platform.

show abstract

Illumination and shadow compensation of hyperspectral images using a digital surface model and non-linear least squares estimation

Cited by 25 publications

References 0 publications

Hyperspectral Tree Species Classification of Japanese Complex Mixed Forest With the Aid of Lidar Data

Hyperspectral Tree Species Classification of Japanese Complex Mixed Forest With the Aid of Lidar Data

Synergistic Use of LiDAR and APEX Hyperspectral Data for High-Resolution Urban Land Cover Mapping

Hyperspectral CNN Classification with Limited Training Samples

Contact Info

Product

Resources

About