#### DMCA

## Feature Extraction of Hyperspectral Data for under Spilled Blood Visualization Using Particle Swarm Optimization

### Citations

3761 | Particle swarm optimization
- Poli, Kennedy, et al.
- 2007
(Show Context)
Citation Context ...pping criterion is same fitness value repeated over a long period). B. Binary PSO The basic PSO algorithm starts with a population of random particles, from where the name “particle swarm” is derived =-=[5]-=-. Each particle in PSO is associated with a velocity. Particles’ velocities are adjusted according to the historical behaviour of each particle and its neighbours while they fly through the search spa... |

1791 |
Texture Features for Image Classification
- Haralick, Shanmugam, et al.
- 1973
(Show Context)
Citation Context ...defined as n ∑ i= 1 d( x , x ) = ( x − x ) , (7) 1 2 1 where the points are measured on n axes or features. The next two metrics are texture features calculated from the grey level coocurrence matrix =-=[9]-=-, which represents the relative frequency of occurrence of two pixels with grey levels i and j in an image, in the neighborhood of two pixels separated by a distance d in a given direction θ. Given th... |

484 |
Swarm Intelligence.
- Kennedy
- 2006
(Show Context)
Citation Context ...assification accuracy [2]. Particle swarm optimization (PSO) is a very promising evolutionary computation technique that had been developed recently for solving nonlinear optimization problems 1 of 4 =-=[3]-=-. PSO’s main attractiveness is its simplicity and velocity allied with robustness. PSO has similar capabilities as genetic algorithms but has the advantage of simpler implementation and reduced bookke... |

466 |
Remote sensing digital image analysis. An introduction. Third revised and enlarged edition
- Richards
- 1999
(Show Context)
Citation Context ...sity value I(λ ) estimated via histogram for each bin k, and m is the number of bins used. 2) Euclidean Distance: The Euclidean distance measures the resemblance of pairs of points in a feature space =-=[8]-=-. The distance d between two points x1 and x2 can be defined as n ∑ i= 1 d( x , x ) = ( x − x ) , (7) 1 2 1 where the points are measured on n axes or features. The next two metrics are texture featur... |

115 |
Computer Processing of Remotely-Sensed Images: An introduction,
- Mather
- 2004
(Show Context)
Citation Context ... or classification. In the remote sensing field, several approaches of feature extraction of hyperspectral data have been developed but mainly aimed at improving or preserving classification accuracy =-=[2]-=-. Particle swarm optimization (PSO) is a very promising evolutionary computation technique that had been developed recently for solving nonlinear optimization problems 1 of 4 [3]. PSO’s main attractiv... |

33 |
Feature selection for structure-activity correlation using binary particle swarms
- Agrafiotis, Cedeo
(Show Context)
Citation Context ...t optimization problems, or problems that can be converted to optimization problems. PSO has been successfully applied to feature selection for quantitative structureactivity relationship correlation =-=[4]-=-. Nonetheless, the problem to define the target function to be optimized is highly dependant on the application at hand. In this paper, we present a new approach for feature extraction of hyperspectra... |

24 |
The Image Processing Handbook. Third Edition
- Russ
- 1998
(Show Context)
Citation Context ...tive measure of the extracted band’s information content. During the search procedure, the criteria are maximized. 1) Information Entropy: The entropy estimates the amount of information in the image =-=[7]-=-. The information entropy measure S, of an image λ is defined as m ∑ k = 1 S(λ ) = − p( I ) ln p( I ) , (6) k where p is the probability distribution function of the image’s intensity value I(λ ) esti... |

3 | Towards a Surgical Tool Using Hyperspectral Imagery as Visual Aid
- ST, Uto, et al.
- 2004
(Show Context)
Citation Context ...cessfully demonstrated the possibility of exploring a spectral interval of less blood optical absorbance in the nearinfrared region of the spectrum to generate visualizations under the layer of blood =-=[1]-=-. The final goal is to obtain a visualization of blood-covered areas that cannot be seen in the visible region. Hyperspectral imaging systems are able to acquire several hundreds of spectral informati... |