Autonomous single-pass endmember approximation using lattice auto-associative memories

Ritter, Gerhard X.; Urcid, Gonzalo; Schmalz, Mark S.

doi:10.1016/j.neucom.2008.06.025

Cited by 41 publications

(28 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For comparison with the SPICE results in Figure 9, normalized distributions of abundance values across each endmember found by PCE were computed using Equation 15. The distributions found are shown in Figure 10.…”

Section: A Piece-wise Convex Endmember Detection Results On the Avirmentioning

confidence: 99%

“…By restricting the endmembers to be data points from the scene, these methods cannot find endmembers when pure pixels cannot be found in the image. Many methods have also been developed based on NonNegative Matrix Factorization [8], [9], [10], [11], Independent Components Analysis [12], [13] and others [14], [15]. These methods search for a single set of endmembers and, therefore, a single convex region to describe a hyperspectral scene.…”

mentioning

confidence: 99%

See 1 more Smart Citation

PCE: Piecewise Convex Endmember Detection

Zare

Gader

2010

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Abstract-A new hyperspectral endmember detection method that represents endmembers as distributions, autonomously partitions the input data set into several convex regions, and simultaneously determines endmember distributions and proportion values for each convex region is presented. Spectral unmixing methods that treat endmembers as distributions or hyperspectral images as piece-wise convex data sets have not been previously developed.Piece-wise Convex Endmember detection, PCE, can be viewed in two parts, the first, the Endmember Distributions detection (ED) algorithm, estimates a distribution for each endmember rather than estimating a single spectrum. By using endmember distributions, PCE can incorporate an endmember's inherent spectral variation and the variation due to changing environmental conditions. ED uses a new sparsity-promoting polynomial prior while estimating abundance values. The second part of PCE partitions the input hyperspectral data set into convex regions and estimates endmember distributions and proportions for each of these regions. The number of convex regions is determined autonomously using the Dirichlet process. PCE is effective at handling highly-mixed hyperspectral images where all of the pixels in the scene contain mixtures of multiple endmembers. Furthermore, each convex region found by PCE conforms to the Convex Geometry Model for hyperspectral imagery. This model requires that the proportions associated with a pixel be nonnegative and sum-to-one.Algorithm results on hyperspectral data indicate that PCE produces endmembers that represent the true ground truth classes of the input data set. The algorithm can also effectively represent endmembers as distributions, thus, incorporating an endmember's spectral variability.

show abstract

Section: A Piece-wise Convex Endmember Detection Results On the Avirmentioning

confidence: 99%

mentioning

confidence: 99%

PCE: Piecewise Convex Endmember Detection

Zare

Gader

2010

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The first group of methods assume that pure pixels exist in the hyperspectral scene and thus categorized under the name of pure pixel (PP) based methods. The most notable methods in this group include pixel purity index (PPI) algorithm [16,17], N-Finder [18], the iterative error analysis (IEA) algorithm [19], the vertex component analysis (VCA) algorithm [20], the simplex growing algorithm (SGA) [21], the sequential maximum angle convex cone (SMACC) algorithm [22], the alternating volume maximization (AVMAX) [23], the successive volume maximization (SVMAX) [23], the collaborative convex framework [24], and finally Lattice Associative Memories (LAM) [25,26,27]. The second group of geometrical based approaches area minimum volume (MV) based methods.…”

Section: Unmixingmentioning

confidence: 99%

Anomaly detection with sparse unmixing and Gaussian mixture modeling of hyperspectral images

Erdinç

Aksoy

2015

2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS)

View full text Add to dashboard Cite

One of the main applications of hyperspectral image analysis is anomaly detection where the problem of interest is the detection of small rare objects that stand out from their surroundings. A common approach to anomaly detection is to first model the background scene and then to use a detector that quantifies the difference of a particular pixel from this background. However, identifying the dominant background components and modeling them is a challenging task. We propose an anomaly detection framework that uses Gaussian mixture models for characterizing the scene background in hyperspectral images. First, the full spectrum is divided into several contiguous band groups for dimensionality reduction as well as for exploiting the peculiarities of different parts of the spectrum. Then, sparse spectral unmixing is performed for each band group for identifying significant endmembers in the scene. Three methods for identifying the dominant background groups such as thresholding, hierarchical clustering and biclustering are used in the endmember abundance space to retrieve the sets of pixel groups that represent dominant background components. Next, these pixel groups are used for initializing individual Gaussian mixture models that are estimated separately for each spectral band group. The proposed method enables automatic identification of the number of mixture components and effective initialization of the estimation procedure for the mixture model. Finally, the Gaussian mixture models for all groups are statistically fused for obtaining the final anomaly map for the scene. Comparative experiments showed that the proposed methods performed better than two other density-based anomaly detectors, especially for small false positive rates, on an airborne hyperspectral data set. Anomali tespiti, hiperspektral görüntü analizindeki ana uygulamalardan biridir. Anomali tespitindeki problem, görüntüde seyrek olarak bulunan,çevresine göre farklılık gösteren küçük nesnelerin tespitidir. Yaygın yaklaşımlardan biri görüntü arka planının modellenmesi ve sınanan pikselin bu modele olan farklılıgına göre sınıflandırılmasıdır. Ancak, karmaşık arka planların modellenmesi kolay degildir. Bu kapsamda, Gauss karışım modeli tabanlı bir anomali tespiti yöntemï onerilmektedir.Öncelikle görüntü belirli sayıda spektral gruplara ayrılmaktadır. Ardından seyrek spektral ayrıştırma bütün spektral grouplarda uygulanmaktadır. Sonrasında, elde edilen son eleman bolluk degerleri kullanılarak, eşikleme tabanlı, sıra düzenli tabanlı, veçift yönlüöbekleme tabanlı olmaküzereüç farklı yöntem ile baskın arka plan grupları tespit edilmektedir. Ardından, belirlenen arka plan gruplarını temsil eden pikseller başlangıç degerlerinin hesaplanmasında kullanılmaküzere, her bir spektral grup için bir Gauss karışım modelï ogrenilmektedir.Önerilen yöntem karışım modeli için bileşen sayısının otomatik belirlenmesine ve kestirim sürecinin etkili başlatılmasına olanak saglamaktadir. Son olarak,ögrenilen Gauss karışım modelleri istatistiksel olarak birleştirilerek sonu...

show abstract

“…These ideas have been applied in diverse areas, such as pattern recognition (Ritter et al, 1998), associative memories in image processing (Ritter et al, 2003;Ritter & Gader, 2006;), computational intelligence (Graña, 2008), industrial applications modeling and knowledge representation (Kaburlasos & Ritter, 2007), and hyperspectral image segmentation (Graña et al, 2009;Ritter et al, 2009;Ritter & Urcid, 2010;.…”

Section: Lattice Algebra Operationsmentioning

confidence: 99%