Back to Search
Start Over
Automatic Detection of Geospatial Objects Using Multiple Hierarchical Segmentations
- Source :
- IEEE Transactions on Geoscience and Remote Sensing
- Publication Year :
- 2008
- Publisher :
- Institute of Electrical and Electronics Engineers (IEEE), 2008.
-
Abstract
- Cataloged from PDF version of article. The object-based analysis of remotely sensed imagery provides valuable spatial and structural information that is complementary to pixel-based spectral information in classi- fication. 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.
- Subjects :
- Object detection algorithms
Image Segmentation, Mathematical Morphology
Computer science
Image analysis
Spectral informations
Image texture
Generic algorithms
Computer vision
Segmentation
Spectral banding
Risk assessment
Image segmentation
Mathematical models
Conditional probability distributions
Applied (CO)
Unsupervised Object Detection
Geo spatial objects
Remote sensing
Kullback leibler divergence (KLD)
Feature (computer vision)
Feature extraction
Set theory
Data sets
Different scales
Algorithms
Probabilistic latent semantic analysis (PLSA)
Segmentation algorithms
Information theory
Object detection
Automatic labelling
Comparative experiments
New algorithm
Novel methods
Hierarchical Segmentation
Feature distribution
High-resolution (HR) images
Object classes
Learning algorithms
Image processing
Object-based
Information retrieval
Morphological operations
Boolean functions
Electrical and Electronic Engineering
Connected components
Unsupervised segmentation
Probability
Remotely sensed imagery (RSI)
Pixel
Probabilistic latent semantic analysis
Labels
Segmentation-based object categorization
business.industry
Structural informations
Pattern recognition
Object recognition
Individual (PSS 544-7)
Object-based Analysis
Structuring element (SE)
Probability distributions
Grouping problems
Object modelling
Mathematical morphology
Computer Science::Computer Vision and Pattern Recognition
Automatic detection
General Earth and Planetary Sciences
Artificial intelligence
business
Subjects
Details
- ISSN :
- 15580644 and 01962892
- Volume :
- 46
- Database :
- OpenAIRE
- Journal :
- IEEE Transactions on Geoscience and Remote Sensing
- Accession number :
- edsair.doi.dedup.....3694be3bf1d5053bf8ea5b5ba1a80063