03_sitaram_TGRS06

03_sitaram_TGRS06 - 1 Modeling and Detection of Geospatial...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 1 Modeling and Detection of Geospatial Objects Using Texture Motifs Sitaram Bhagavathy, Member, IEEE, and B. S. Manjunath, Fellow, IEEE Abstract We propose the use of texture motifs , or charac- teristic spatially recurrent patterns for modeling and detecting geospatial objects. A method is proposed for learning a texture motif model from object examples and detecting objects based on the learned model. The model is learned in a two-layered frameworkthe first learns the constituent texture elements of the motif and the second, the spatial distribution of the elements. In the experimental session, we first demonstrate the model training and selection methodology for different objects given a limited dataset of each. We then emphasize the utility of such models for detecting the presence or absence of geospatial objects in large aerial image datasets comprising tens of thousands of image tiles. Index Terms geospatial object, object detection, object model I. INTRODUCTION Aerial and satellite images of the earth (or geospatial images) are critical sources of information in diverse fields such as geography, cartography, meteorology, surveillance, city planning. These images contain visual information about various natural and man-made features on or above the surface of the earth. Manual annotation of geospatial images covering even a relatively small area of the earth is a tedious task. This has necessitated research into automated annotation of geospatial images. An important component of this research comprises object detection methods, which are model-driven methods that seek to identify probable locations of specified features of interest or objects in geospatial images. For ex- ample, detection of buildings and roads is a useful step in cartography. Detection of objects such as harbors, airports, golf courses, housing colonies, vineyards, and parking lots is useful for updating geographical databases such as the Alexandria Digital Library (ADL) Gazetteer [1] which index the loca- tions of several object types. Automated object detection is an important step toward an object-based representation of geospatial images. The detection of geospatial objects with simple geometric or shape models such as buildings [2], [3], [4], [5], [6], roads [7], [8], [9], and other small objects [10], [11] has been explored adequately in the literature. This is not the case for compound objects, such as harbors and golf courses, characterized by several parts and their spatial layout. For example, harbors contain boats and golf courses contain trees and grass, both with a distinct spatial arrangement (Fig. 1). The authors are with the Department of Electrical and Computer Engineer- ing, University of California, Santa Barbara, CA 93106. Sitaram Bhagavathy is currently with Thomson Corporate Research, Princeton, NJ 08540. Email: { sitaram, manj } @ece.ucsb.edu. This research was supported by the following grants: NSF-DLI #IIS-49817432 and NSF IIS #0329267.grants: NSF-DLI #IIS-49817432 and NSF IIS #0329267....
View Full Document

Page1 / 11

03_sitaram_TGRS06 - 1 Modeling and Detection of Geospatial...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online