Gibbs random field models: a toolbox for spatial information extraction
Computers & GeosciencesPublished 1 May 2000
Michael Schröder, Marc Walessa, Hubert Rehrauer, K. Seidel, Mihai Datcu
Citations20
SJR quartileQ1
SJR score1.04
SNIP1.49
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
Gibbs random field models in the form of a powerful toolbox for spatial information extraction from remote sensing images are presented via parametrised energy functions that characterise local interactions between neighbouring pixels.
Abstract
In this paper, we present Gibbs random field models in the form of a powerful toolbox for spatial information extraction from remote sensing images. These models are defined via parametrised energy functions that characterise local interactions between neighbouring pixels. After shortly revisiting the information theoretical concept and defining a family of Gibbs models, we give a tour through examples of different kinds of spatial information extraction. These examples range from parameter estimation and analysis, via selection of the model that best describes the image data, up to the segmentation of the whole image into regions with uniform properties of the model. Finally, the concept of across-image segmentation of spatial information leads to an application for content-based queries from remote sensing image archives.
Keywords
Computer ScienceEngineering
IEEE Transactions on Pattern Analysis and Machine IntelligenceStochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
17,980 Citations1984Stuart Geman, Donald Geman
The analogy between images and statistical mechanics systems is made and the analogous operation under the posterior distribution yields the maximum a posteriori (MAP) estimate of the image given the degraded observations, creating a highly parallel ``relaxation'' algorithm for MAP estimation.
Journal of the Royal Statistical Society Series B (Statistical Methodology)Spatial Interaction and the Statistical Analysis of Lattice Systems
6,179 Citations1974Julian Besag
IEEE Signal Processing MagazineThe expectation-maximization algorithm
3,138 Citations1996Todd K. Moon
The EM (expectation-maximization) algorithm is ideally suited to problems of parameter estimation, in that it produces maximum-likelihood (ML) estimates of parameters when there is a many-to-one mapping from an underlying distribution to the distribution governing the observation.
IEEE Transactions on Pattern Analysis and Machine IntelligenceMarkov Random Field Texture Models
1,388 Citations1983George R. Cross, Anil K. Jain
The power of the binomial model to produce blurry, sharp, line-like, and blob-like textures is demonstrated and the synthetic microtextures closely resembled their real counterparts, while the regular and inhomogeneous textures did not.
IEEE Transactions on Image ProcessingA multiscale random field model for Bayesian image segmentation
599 Citations1994Charles A. Bouman, M. V. Shapiro
Simulations on synthetic images indicate that the new algorithm performs better and requires much less computation than MAP estimation using simulated annealing, and is found to improve classification accuracy when applied to the segmentation of multispectral remotely sensed images with ground truth data.
IEEE Transactions on Pattern Analysis and Machine IntelligenceUnsupervised texture segmentation using Markov random field models
380 Citations1991B.S. Manjunath, Rama Chellappa
The results of the adaptive segmentation algorithm of Lakshamanan and Derin are compared with a simple nearest-neighbor classification scheme to show that if enough information is available, simple techniques could be used as alternatives to computationally expensive schemes.
IEEE Transactions on Image ProcessingMultiresolution Gauss-Markov random field models for texture segmentation
233 Citations1997S. Krishnamachari, Rama Chellappa
The experiments with synthetic, Brodatz texture, and real satellite images show that the multiresolution technique results in a better segmentation and requires lesser computation than the single resolution algorithm.
IEEE Transactions on Signal ProcessingCompound Gauss-Markov random fields for image estimation
182 Citations1991Fure-Ching Jeng, John W. Woods
The authors consider both the maximum a posteriori probability (MAP) estimate and the minimum mean-squared error (MMSE) estimate for image estimation and image restoration for images modeled as compound Gauss-Markov random fields.
IEEE Transactions on Geoscience and Remote SensingSpatial information retrieval from remote-sensing images. I. Information theoretical perspective
119 Citations1998Mihai Datcu, K. Seidel +1 more
The authors introduce a pragmatic approach to extract structural information from RS images by selecting from a library of a priori models those which best explain the structures within an image.
IEEE Transactions on Geoscience and Remote SensingSpatial information retrieval from remote-sensing images. II. Gibbs-Markov random fields
107 Citations1998Michael Schröder, Hubert Rehrauer +2 more
The authors present Gibbs-Markov random field models as a powerful and robust descriptor of spatial information in typical remote-sensing image data as well as examples for both synthetic aperture radar (SAR) and optical data.
IEEE Transactions on Geoscience and Remote SensingA new clustering algorithm applicable to multispectral and polarimetric SAR images
61 Citations1993Y. Wong, Edward C. Posner
A scale-space clustering algorithm, more powerful than existing ones, may be useful for remote sensing for land use because it can cluster data in any multidimensional space and its insensitive to variability in cluster densities, sizes and ellipsoidal shapes.
Query by image content from remote sensing archives
24 Citations1998K. Seidel, Michael Schröder +3 more
The authors present the structure of a new intelligent remote sensing image archive providing query by image content using a family of robust signal models which are not selected according to a certain application but instead are able to describe the information in a condensed way.
New architectures for remote sensing image archives
16 Citations2002K. Seidel, R. Mastropietro +1 more
The authors propose a unified treatment of the archivation of very diverse image products from the point of view of content-based picture query and retrieval systems.
Multiscale image segmentation with a dynamic label tree
8 Citations1998H. Rehreuer, K. Seidel +1 more
The authors' model allows the dependencies to switch, so that the hierarchical structure itself is optimised, and achieves very smooth segmentations, even at coarse scale, and can be fast computed.