Low-Rank Decomposition and Total Variation Regularization of Hyperspectral Video Sequences

Xu, Yang; Wu, Zebin; Chanussot, Jocelyn; Mura, Mauro Dalla; Bertozzi, Andrea L.; Wei, Zhihui

doi:10.1109/tgrs.2017.2766094

Cited by 39 publications

(13 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A correct detection is defined as the Intersection over Union (IoU) against the groundtruth is greater than a threshold. Since the vehicles in satellite videos are small, we set the threshold as 0.3 4 . In addition to the evaluation of detection performance, we also consider the rank of the estimated background as an indicator of the background modeling performance.…”

Section: Methodsmentioning

confidence: 99%

Error Bounded Foreground and Background Modeling for Moving Object Detection in Satellite Videos

Zhang

Jia

2020

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Detecting moving objects from ground-based videos is commonly achieved by using background subtraction techniques. Low-rank matrix decomposition inspires a set of state-ofthe-art approaches for this task. It is integrated with structured sparsity regularization to achieve background subtraction in the developed method of Low-rank and Structured Sparse Decomposition (LSD). However, when this method is applied to satellite videos where spatial resolution is poor and targets contrast to the background is low, its performance is limited as the data no longer fits adequately either the foreground structure or the background model. In this paper, we handle these unexplained data explicitly and address the moving target detection from space as one of the pioneer studies. We propose a new technique by extending the decomposition formulation with bounded errors, named Extended Low-rank and Structured Sparse Decomposition (E-LSD). This formulation integrates lowrank background, structured sparse foreground as well as their residuals in a matrix decomposition problem. To solve this optimization problem is challenging. We provide an effective solution by introducing an alternative treatment and adopting the direct extension of Alternating Direction Method of Multipliers (ADMM). The proposed E-LSD was validated on two satellite videos, and experimental results demonstrate the improvement in background modeling with boosted moving object detection precision over state-of-the-art methods.

show abstract

Section: Methodsmentioning

confidence: 99%

Error Bounded Foreground and Background Modeling for Moving Object Detection in Satellite Videos

Zhang

Jia

2020

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Furthermore, the planet signal is sparse in spatial and spectral dimensions, while the remaining field of view is low dimensional in the spectral direction. Low-rank and sparse unmixing can thus be foreseen (e.g., Tsinos et al 2017;Xu et al 2018), similar to the reasoning behind the decomposition of angular differential imaging data from Gomez Gonzalez et al (2016). The sparsity constrain can be pushed even further.…”

Section: Summary and Future Workmentioning

confidence: 99%

Spectral unmixing for exoplanet direct detection in hyperspectral data

Rameau

Chanussot

Carlotti

et al. 2021

A&A

View full text Add to dashboard Cite

Context. The direct detection of faint exoplanets with high-contrast instruments can be boosted by combining it with high spectral resolution. For integral field spectrographs yielding hyperspectral data, this means that the majority of the field of view consists of diffracted starlight spectra and a spatially localized planet. Observation analysis usually relies on classic cross-correlation with theoretical spectra, maximized at the position and with the properties of the planet. In a purely blind-search context, this supervised strategy can be biased with model mismatch and/or be computationally inefficient. Aims. Using an approach that is inspired by the analysis of hyperspectral data within the remote-sensing community, we aim to propose an alternative to cross-correlation that is fully data-driven, which decomposes the data into a set of individual spectra and their corresponding spatial distributions. This strategy is called spectral unmixing. Methods. We used an orthogonal subspace projection to identify the most distinct spectra in the field of view. Their spatial distribution maps were then obtained by inverting the data. These spectra were then used to break the original hyperspectral images into their corresponding spatial distribution maps via non-negative least squares. A matched filter with the instrument point-spread function (or visual inspection) was then used to detect the planet on one of the maps. The performance of our method was evaluated and compared with a cross-correlation using simulated hyperspectral data with medium resolution from the ELT/HARMONI integral field spectrograph. Results. We show that spectral unmixing effectively leads to a planet detection solely based on spectral dissimilarities at significantly reduced computational cost. The extracted spectrum holds significant signatures of the planet while being not perfectly separated from residual starlight. The sensitivity of the supervised cross-correlation is three to four times higher than with unsupervised spectral unmixing, the gap is biased toward the former because the injected and correlated spectrum match perfectly. The algorithm was furthermore vetted on real data obtained with VLT/SINFONI of the β Pictoris system. This led to the detection of β Pictoris b with a signal-to-noise ratio of 28.5. Conclusions. Spectral unmixing is a viable alternative strategy to a cross-correlation to search for and characterize exoplanets in hyperspectral data in a purely data-driven approach. The advent of large data from the forthcoming IFS on board JWST and future ELTs motivates further algorithm development along this path.

show abstract

“…Based on this advantage, HSIs have been employed in many applications, including scene classification, change detection, and anomaly detection [2], [5]- [8]. Among these applications, hyperspectral anomaly detection has attracted lots of attention [9] due to its significance in precision agriculture [10], mineral discovery [11], and military defense [12]- [14].…”

Section: Introductionmentioning

confidence: 99%

Hyperspectral Anomaly Detection With Morphological Random Walker

et al. 2021

View full text Add to dashboard Cite

This work proposes a new morphological random walker (MRW) method for hyperspectral anomaly detection. The proposed method introduces a morphology-based objective function into a random walker (RW) algorithm, sufficiently exploiting spatial morphological property and spatial similarity of HSIs for detection. The MRW method comprises two major stages. Firstly, we employ the extended morphological profiles (EMPs) and different operations to extract the spatial morphological property of HSIs. Second, according to the morphological property, we construct a morphology-based objective function. This function is incorporated into the RW-based optimization model, encoding the spatial similarity of HSIs in a weighted graph. Two factors determine the class of test pixels, including the spatial morphological information learned by EMPs, and the spatial correlation among adjoining pixels modeled by the weighted graph. Since the two factors are well considered in the MRW method, the proposed method illustrates outstanding detection performances for several widely used real HSIs.INDEX TERMS Hyperspectral images, anomaly detection, random walker, extended morphological profiles, morphological property.

show abstract

Low-Rank Decomposition and Total Variation Regularization of Hyperspectral Video Sequences

Cited by 39 publications

References 49 publications

Error Bounded Foreground and Background Modeling for Moving Object Detection in Satellite Videos

Error Bounded Foreground and Background Modeling for Moving Object Detection in Satellite Videos

Spectral unmixing for exoplanet direct detection in hyperspectral data

Hyperspectral Anomaly Detection With Morphological Random Walker

Contact Info

Product

Resources

About