Automatic detection of geospatial objects using multiple hierarchical segmentations
Date
2008-07Source Title
IEEE Transactions on Geoscience and Remote Sensing
Print ISSN
0196-2892
Publisher
Institute of Electrical and Electronics Engineers
Volume
46
Issue
7
Pages
2097 - 2111
Language
English
Type
ArticleItem Usage Stats
270
views
views
343
downloads
downloads
Abstract
The object-based analysis of remotely sensed imagery provides valuable spatial and structural information that is complementary to pixel-based spectral information in classification. In this paper, we present novel methods for automatic object detection in high-resolution images by combining spectral information with structural information exploited by using image segmentation. The proposed segmentation algorithm uses morphological operations applied to individual spectral bands using structuring elements in increasing sizes. These operations produce a set of connected components forming a hierarchy of segments for each band. A generic algorithm is designed to select meaningful segments that maximize a measure consisting of spectral homogeneity and neighborhood connectivity. Given the observation that different structures appear more clearly at different scales in different spectral bands, we describe a new algorithm for unsupervised grouping of candidate segments belonging to multiple hierarchical segmentations to find coherent sets of segments that correspond to actual objects. The segments are modeled by using their spectral and textural content, and the grouping problem is solved by using the probabilistic latent semantic analysis algorithm that builds object models by learning the object-conditional probability distributions. The automatic labeling of a segment is done by computing the similarity of its feature distribution to the distribution of the learned object models using the Kullback-Leibler divergence. The performances of the unsupervised segmentation and object detection algorithms are evaluated qualitatively and quantitatively using three different data sets with comparative experiments, and the results show that the proposed methods are able to automatically detect, group, and label segments belonging to the same object classes. © 2008 IEEE.
Keywords
Hierarchical segmentationImage segmentation
Object-based analysis
Unsupervised object detection
Boolean functions
Feature extraction
Image processing
Information theory
Learning algorithms
Mathematical morphology
Probability
Probability distributions
Risk assessment
Set theory
Applied (CO)
Automatic detection
Automatic labelling
Comparative experiments
Conditional probability distributions
Connected components
Data sets
Different scales
Feature distribution
Generic algorithms
Geo spatial objects
Grouping problems
High-resolution (HR) images
Individual (PSS 544-7)
Kullback leibler divergence (KLD)
Morphological operations
New algorithm
Novel methods
Object classes
Object detection
Object detection algorithms
Object modelling
Object-based
Probabilistic latent semantic analysis (PLSA)
Remotely sensed imagery (RSI)
Segmentation algorithms
Spectral banding
Spectral informations
Structural informations
Structuring element (SE)
Unsupervised segmentation
Object recognition
Algorithms
Image analysis
Information retrieval
Labels
Mathematical models
Remote sensing
Permalink
http://hdl.handle.net/11693/23067Published Version (Please cite this version)
http://dx.doi.org/10.1109/TGRS.2008.916644Collections
Related items
Showing items related by title, author, creator and subject.
-
Image sequence analysis for emerging interactive multimedia services-the European COST 211 framework
Alatan, A. A.; Onural, L.; Wollborn, M.; Mech, R.; Tuncel, E.; Sikora, T. (Institute of Electrical and Electronics Engineers, 1998-11)Flexibility and efficiency of coding, content extraction, and content-based search are key research topics in the field of interactive multimedia. Ongoing ISO MPEG-4 and MPEG-7 activities are targeting standardization to ... -
Automated detection of objects using multiple hierarchical segmentations
Akçay H. Gökhan; Aksoy, Selim (IEEE, 2007-07)We introduce an unsupervised method that combines both spectral and structural information for automatic object detection. First, a segmentation hierarchy is constructed by combining structural information extracted by ... -
Segmentation-based extraction of important objects from video for object-based indexing
Baştan, Muhammet; Güdükbay, Uğur; Ulusoy, Özgür (IEEE, 2008-06)We describe a method to automatically extract important video objects for object-based indexing. Most of the existing salient object detection approaches detect visually conspicuous structures in images, while our method ...