A fast automatic extraction algorithm of elliptic object groups from remote sensing images

Authors:
Changren Zhu;Runsheng Wang
Affiliations:
ATR National Laboratory, National University of Defense Technology, Changsha, Hunan 410073, PR China;ATR National Laboratory, National University of Defense Technology, Changsha, Hunan 410073, PR China
Venue:
Pattern Recognition Letters - Special issue: Pattern recognition for remote sensing (PRRS 2002)
Year:
2004

Citing 5
Cited 1

Detecting partially occluded ellipses using the Hough transform

Image and Vision Computing - 4th Alvey Vision Meeting
A Data-Driven Intermediate Level Feature Extraction Algorithm

IEEE Transactions on Pattern Analysis and Machine Intelligence
Partitioned Hough transform for ellipsoid detection

Pattern Recognition
Extracting geometric primitives

CVGIP: Image Understanding
The dynamic generalized Hough transform: its relationship to the probabilistic Hough transforms and an application to the concurrent detection of circles and ellipses

CVGIP: Image Understanding

A simple least squares method for fitting of ellipses and circles depends on border points of a two-tone image and their 3-D extensions

Pattern Recognition Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

A fast automatic extraction of elliptic object group (EOG), which consists of multiple elliptic objects in an orderly and dense distribution, is very important for the understanding of some remote sensing images. Thus, based on the analysis of its characteristics in many remote sensing images, a fast automatic extraction algorithm of the EOG from remote sensing images based on a hierarchical processing idea is proposed in this paper. It consists of two steps: the detection of EOG candidate regions and the extraction of the EOGs. In the first step, EOG candidate regions are extracted in a small image according to their texture attributes of edge density and phase symmetry based on a hierarchical processing. In the second step, every single elliptic object is detected with an ellipse fitting and region analysis in every EOG candidate region in the original image, and then the ellipses that are both similar and neighboring are grouped into different EOGs. The algorithm has a good performance chiefly due to the following two aspects: On the one hand, it is fast if a hierarchical processing is adopted. On the other hand, in every single ellipse extraction process, an algorithm based on region fitting is developed. Unlike many other algorithms, which need multiple accumulations or multiple fitting and error analysis for every ellipse, it does the ellipse fitting and analysis only once for every ellipse. Moreover, because the object regions are obtained by the enclosure of the edge, it can overcome the drawbacks of over-segmentation and under-segmentation, which can be often caused by gray segmentation. The experimental results show the effectiveness and efficiency of the proposed algorithm.