An Approach to Multiple Change Detection in VHR Optical Images Based on Iterative Clustering and Adaptive Thresholding

Yady Tatiana Solano-Correa; Francesca Bovolo; Lorenzo Bruzzone

doi:10.1109/LGRS.2019.2896385

An Approach to Multiple Change Detection in VHR Optical Images Based on Iterative Clustering and Adaptive Thresholding

Título traducido de la contribución: Un enfoque para la detección de cambios múltiples en imágenes ópticas VHR basado en agrupamiento iterativo y umbralización adaptativa

Yady Tatiana Solano-Correa, Francesca Bovolo, Lorenzo Bruzzone

Producción: Contribución a una revista › Artículo › revisión exhaustiva

17 Citas (Scopus)

Resumen

Uno de los enfoques más comunes para la detección de cambios no supervisada (CD) en imágenes multiespectrales es el análisis de vectores de cambio (CVA). El CVA calcula la imagen de diferencia multiespectral y explota su distribución estadística en coordenadas (hiper)esféricas mediante dos pasos: 1) magnitud y 2) umbralización de dirección. Los dos pasos requieren suposiciones sobre: 1) el modelo de distribuciones de clase y 2) el número de cambios. Sin embargo, ambas suposiciones rara vez se cumplen o son difíciles de formular, especialmente al considerar imágenes VHR. Por lo tanto, proponemos un enfoque para CD múltiple en imágenes ópticas VHR basado en agrupamiento iterativo y umbralización adaptativa en coordenadas (hiper)esféricas. El enfoque propuesto: 1) es libre de distribución; 2) no es supervisado; 3) identifica automáticamente el número de cambios; y 4) es robusto al ruido. Los resultados obtenidos en dos conjuntos de datos multitemporales de un solo sensor y multisensor, que incluyen imágenes de WorldView-2 y QuickBird, corroboran la efectividad del enfoque propuesto.

Título traducido de la contribución	Un enfoque para la detección de cambios múltiples en imágenes ópticas VHR basado en agrupamiento iterativo y umbralización adaptativa
Idioma original	Inglés
Número de artículo	8648509
Páginas (desde-hasta)	1334-1338
Número de páginas	5
Publicación	IEEE Geoscience and Remote Sensing Letters
Volumen	16
N.º	8
DOI	https://doi.org/10.1109/LGRS.2019.2896385
Estado	Publicada - ago. 2019
Publicado de forma externa	Sí

Palabras clave

Umbral adaptativo
detección de cambios (CD)
agrupamiento
multitemporal
imágenes de muy alta resolución

Acceder al documento

10.1109/LGRS.2019.2896385

Otros archivos y enlaces

Enlace a la publicación en Scopus

Citar esto

@article{d84feb5506614d9c8b1de09cd9f1bf9f,

title = "An Approach to Multiple Change Detection in VHR Optical Images Based on Iterative Clustering and Adaptive Thresholding",

abstract = "One of the most common approaches to unsupervised change detection (CD) in multispectral images is change vector analysis (CVA). CVA computes the multispectral difference image and exploits its statistical distribution in (hyper-) spherical coordinates by means of two steps: 1) magnitude and 2) direction thresholding. The two steps require assumptions on: 1) the model of class distributions and 2) the number of changes. However, both assumptions are seldom satisfied or difficult to formulate, especially when considering VHR images. Thus, we propose an approach to multiple CD in VHR optical images based on iterative clustering and adaptive thresholding in (hyper-) spherical coordinate. The proposed approach: 1) is distribution free; 2) is unsupervised; 3) automatically identifies the number of changes; and 4) is robust to noise. Results obtained on two multitemporal single-sensor and multisensor data sets, including images from WorldView-2 and QuickBird, corroborate the effectiveness of the proposed approach.",

keywords = "Adaptive thresholding, change detection (CD), clustering, multitemporal, very high-resolution images, Umbral adaptativo, detecci{\'o}n de cambios (CD), agrupamiento, multitemporal, im{\'a}genes de muy alta resoluci{\'o}n",

author = "Solano-Correa, \{Yady Tatiana\} and Francesca Bovolo and Lorenzo Bruzzone",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.",

year = "2019",

month = aug,

doi = "10.1109/LGRS.2019.2896385",

language = "English",

volume = "16",

pages = "1334--1338",

journal = "IEEE Geoscience and Remote Sensing Letters",

issn = "1545-598X",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "8",

}

TY - JOUR

T1 - An Approach to Multiple Change Detection in VHR Optical Images Based on Iterative Clustering and Adaptive Thresholding

AU - Solano-Correa, Yady Tatiana

AU - Bovolo, Francesca

AU - Bruzzone, Lorenzo

PY - 2019/8

Y1 - 2019/8

N2 - One of the most common approaches to unsupervised change detection (CD) in multispectral images is change vector analysis (CVA). CVA computes the multispectral difference image and exploits its statistical distribution in (hyper-) spherical coordinates by means of two steps: 1) magnitude and 2) direction thresholding. The two steps require assumptions on: 1) the model of class distributions and 2) the number of changes. However, both assumptions are seldom satisfied or difficult to formulate, especially when considering VHR images. Thus, we propose an approach to multiple CD in VHR optical images based on iterative clustering and adaptive thresholding in (hyper-) spherical coordinate. The proposed approach: 1) is distribution free; 2) is unsupervised; 3) automatically identifies the number of changes; and 4) is robust to noise. Results obtained on two multitemporal single-sensor and multisensor data sets, including images from WorldView-2 and QuickBird, corroborate the effectiveness of the proposed approach.

AB - One of the most common approaches to unsupervised change detection (CD) in multispectral images is change vector analysis (CVA). CVA computes the multispectral difference image and exploits its statistical distribution in (hyper-) spherical coordinates by means of two steps: 1) magnitude and 2) direction thresholding. The two steps require assumptions on: 1) the model of class distributions and 2) the number of changes. However, both assumptions are seldom satisfied or difficult to formulate, especially when considering VHR images. Thus, we propose an approach to multiple CD in VHR optical images based on iterative clustering and adaptive thresholding in (hyper-) spherical coordinate. The proposed approach: 1) is distribution free; 2) is unsupervised; 3) automatically identifies the number of changes; and 4) is robust to noise. Results obtained on two multitemporal single-sensor and multisensor data sets, including images from WorldView-2 and QuickBird, corroborate the effectiveness of the proposed approach.

KW - Adaptive thresholding

KW - change detection (CD)

KW - clustering

KW - multitemporal

KW - very high-resolution images

KW - Umbral adaptativo

KW - detección de cambios (CD)

KW - agrupamiento

KW - multitemporal

KW - imágenes de muy alta resolución

UR - https://www.scopus.com/pages/publications/85069529532

U2 - 10.1109/LGRS.2019.2896385

DO - 10.1109/LGRS.2019.2896385

M3 - Article

AN - SCOPUS:85069529532

SN - 1545-598X

VL - 16

SP - 1334

EP - 1338

JO - IEEE Geoscience and Remote Sensing Letters

JF - IEEE Geoscience and Remote Sensing Letters

IS - 8

M1 - 8648509

ER -

An Approach to Multiple Change Detection in VHR Optical Images Based on Iterative Clustering and Adaptive Thresholding

Resumen

Palabras clave

Acceder al documento

Otros archivos y enlaces

Huella

Citar esto