Feature subset selection using separability index matrix

Han, Jong-Seob; Lee, Sang Wan; Bien, Zeungnam

doi:10.1016/j.ins.2012.09.042

Cited by 19 publications

(7 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method has been used in many FS approaches 4,18,20,41 and can be applied to large datasets. This method is much faster than BB.…”

Section: Sequential Forward Selectionmentioning

confidence: 99%

“…18,19 Such methods search in the original hyperspectral space consisting of hundreds of narrow channels and select the channels that provide the optimal value of the given metric. If the target application is classification, this metric is usually a separability metric, which is calculated based on the known classes in the scene.…”

Section: Spectral Region Splitting With the Class Separability Metricmentioning

confidence: 99%

“…19,20,24 Ideally, the best bands for an image classification and class separability are those that provide larger interclass distances and smaller intraclass diameters in the selected feature space. 18,21,24 For example, in one-dimensional feature space, the interclass distance can be obtained simply by measuring the Euclidean distance between the class means, and the class variances can provide the intraclass diameter. Separability measures such as Mahalanobis (Mh) distance and JeffreysMatusita (JM) 23 account for both the inter-and the intraclass distance implicitly, i.e., a larger Mh distance gives a higher ratio of inter-to intraclass distances.…”

Section: Introductionmentioning

confidence: 99%

“…These separability measures can be plugged in to a DR method to obtain a channel set that maximizes class discrimination. 19 There are several supervised channel selection methods, 5,[18][19][20]25 mostly aiming at maximizing class separability. In this paper, we apply a method called the spectral region splitting (SRS) to identify spectral regions that maximize class separability.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Spectral region identification versus individual channel selection in supervised dimensionality reduction of hyperspectral image data

Aria

Menenti

Gorte

2017

J. Appl. Remote Sens

View full text Add to dashboard Cite

, "Spectral region identification versus individual channel selection in supervised dimensionality reduction of hyperspectral image data," J. Appl. Remote Sens. 11(4), 046010 (2017), doi: 10.1117/1.JRS.11.046010. Abstract. Hyperspectral images may be applied to classify objects in a scene. The redundancy in hyperspectral data implies that fewer spectral features might be sufficient for discriminating the objects captured in a scene. The availability of labeled classes of several areas in a scene paves the way for a supervised dimensionality reduction, i.e., using a discrimination measure between the classes in a scene to select spectral features. We show that averaging adjacent spectral channels and using wider spectral regions yield a better class separability than the selection of individual channels from the original hyperspectral dataset. We used a method named spectral region splitting (SRS), which creates a new feature space by averaging neighboring channels. In addition to the common benefits of channel selection methods, the algorithm constructs wider spectral regions when it is useful. Using different class separability measures over various datasets resulted in a better discrimination between the classes than the best-selected channels using the same measure. The reason is that the wider spectral regions led to a reduction in intraclass distances and an improvement in class discrimination. The overall classification accuracy of two hyperspectral scenes gave an increase of about twopercent when using the spectral regions determined by applying SRS. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

“…This method has been used in many FS approaches 4,18,20,41 and can be applied to large datasets. This method is much faster than BB.…”

Section: Sequential Forward Selectionmentioning

confidence: 99%

Section: Spectral Region Splitting With the Class Separability Metricmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Spectral region identification versus individual channel selection in supervised dimensionality reduction of hyperspectral image data

Aria

Menenti

Gorte

2017

J. Appl. Remote Sens

View full text Add to dashboard Cite

show abstract

“…It does require an evaluator to measure the intrinsic characteristics of each feature. Han et al [14] presented a new criterion function that identifies features pertinent to the classification task at a very low computational cost. Chen [3] selected salient features using compactness and separability.…”

Section: Related Workmentioning

confidence: 99%