A New Band Selection Strategy for Target Detection in Hyperspectral Images

Diani, Marco; Acito, N.; Greco, Mario; Corsini, Giovanni

doi:10.1007/978-3-540-85567-5_53

Cited by 12 publications

(9 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…E.g., Diani and others proposed a novel approach to data dimensionality reduction that is tailored to target detection applications. it extracts the subset of bands that optimizes the probability of detecting a target in a given background, for a fixed probability of false alarm [18]. Prabhakar and others through regression analysis revealed a significant linear relation between leafhopper severity and chlorophyll.…”

Section: A Overseas Research Status Of Band Selection For Hyperspectmentioning

confidence: 96%

Research Advance on Band Selection-Based Dimension Reduction of Hyperspectral Remote Sensing Images

Wang

Huang

Liu

et al. 2012

2012 2nd International Conference on Remote Sensing, Environment and Transportation Engineering

View full text Add to dashboard Cite

The typical characteristics of hyperspectral remote sensing data are combining image with spectrum, high spectral resolution, many bands and much redundant information. A hyperspectral image cube can be used to obtain millions spectrum curves. Aiming at a hyperspectral remote sensing image containing huge amounts of data, removing redundant information and reducing processing dimensions are the premise and foundation for hyperspectral remote sensing processing and applications. Hyperspectral data dimension reduction techniques mainly include the feature extraction and band selection. This paper has fully studied the theories and methods of dimension reduction for band selection, analyses their advantages, disadvantages and validity, and deeply discusses the current situation and tendency of the development of band selection based on dimension reduction of hyperspectral remote sensing image at last.

show abstract

Section: A Overseas Research Status Of Band Selection For Hyperspectmentioning

confidence: 96%

Research Advance on Band Selection-Based Dimension Reduction of Hyperspectral Remote Sensing Images

Wang

Huang

Liu

et al. 2012

2012 2nd International Conference on Remote Sensing, Environment and Transportation Engineering

View full text Add to dashboard Cite

show abstract

“…According to (13), one of the most important problems of CC cost function is that it innately tends to increase the data size and select more bands. Thus, CC reaches the maximum rate (number 1) when nominator and denominator of the fraction are equal to one another, occurring when the number of selected bands is equal to the number of original image's bands, i.e.…”

Section: The Cost Functionsmentioning

confidence: 99%

Selecting optimal bands for sub‐pixel target detection in hyperspectral images based on implanting synthetic targets

Hashjin

Boloorani

Khazai

et al. 2019

IET Image Processing

View full text Add to dashboard Cite

Target detection at sub-pixel abundances is in fact one of the challenging issues of hyperspectral image processing. Selection of optimal bands to improve sub-pixel target detection (STD) performance is one of the common solutions, applied by many researchers. Nevertheless, absence of sufficient training data is the main weakness of selecting optimal bands with regards to this approach. The present research introduces a new band selection method for STD in hyperspectral images, based on creating training data, in which the desired target spectrum is implanted randomly in a series of host pixels from the entire hyperspectral image. Afterward, via running an optimization algorithm twice, with the aim of minimizing the false alarm rate (FAR) in local Adaptive Coherence Estimator (ACE) target detection algorithm, the number of optimal bands and optimal spectral bands are selected. In this study, the performance of three optimization methods, including Genetic Algorithm (GA), Grey Wolf Optimization (GWO), and Particle Swarm Optimization (PSO), are compared. Experimental results on Hymap and Hyperion datasets show that, the proposed method obtains the minimum FAR compared to the rest of evaluated methods. Also, based on the results obtained, GWO outperforms GA and PSO optimization methods in the STD domain.

show abstract

“…They take the detection performance as the objective function that has to be maximized. Diani et al (2008) selected the subset of bands that maximizes the probability of detection for a fixed probability of false alarm, when a target with a known spectral signature must be detected in a given scenario. Serpico and Bruzzone (2000) proposed the steepest ascent (SA) search algorithm for feature selection in hyperspectral data.…”

Section: Feature Selectionmentioning

confidence: 99%

Using class-based feature selection for the classification of hyperspectral data

Maghsoudi

Zoej

Collins

2011

International Journal of Remote Sensing

View full text Add to dashboard Cite

A New Band Selection Strategy for Target Detection in Hyperspectral Images

Cited by 12 publications

References 14 publications

Research Advance on Band Selection-Based Dimension Reduction of Hyperspectral Remote Sensing Images

Research Advance on Band Selection-Based Dimension Reduction of Hyperspectral Remote Sensing Images

Selecting optimal bands for sub‐pixel target detection in hyperspectral images based on implanting synthetic targets

Using class-based feature selection for the classification of hyperspectral data

Contact Info

Product

Resources

About