A New Approach for the Extraction of Aboveground Circular Structures From Near-Nadir VHR Satellite Imagery
- 11 July 2013
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Geoscience and Remote Sensing
- Vol. 52 (6), 3125-3140
- https://doi.org/10.1109/tgrs.2013.2270372
Abstract
In this paper, a new automated approach for the extraction of aboveground circular storage structures from near-nadir very high resolution satellite imagery is proposed. The approach focuses on the cast shadows of the circular structures and splits the boundaries of the shadow regions into curved segments using the chord-to-point distance accumulation technique. Thereafter, the curved segments are tested with newly developed constraints for being a part of a circular structure, and the ones that pass all of the constraints are considered as candidates. The reciprocal relations between the candidate segments are assessed by a developed mutual evidence test, and for the candidates that expose a relation, a robust circle fitting is applied. For the candidates having no such relations, an approach that further validates the circle evidence is developed. The approach consists in introducing regions-of-interest (ROIs) for each candidate segment and applying a circular Hough transform in each ROI, where the parameters of the transform are self-controlled. Experiments performed on 12 challenging Geoeye-1 test images selected from industrial areas reveal that the proposed approach accurately detects aboveground circular structures in complex industrial environments. Besides, the comparison of the results of the proposed approach with the results of two different circle detection approaches verifies the success and the robustness of the approach developed.Keywords
This publication has 52 references indexed in Scilit:
- Automatic Detection and Segmentation of Orchards Using Very High Resolution ImageryIEEE Transactions on Geoscience and Remote Sensing, 2012
- Circle detection on images using learning automataIET Computer Vision, 2012
- The circlet transform: A robust tool for detecting features with circular shapesComputers & Geosciences, 2011
- Circle detection on images using genetic algorithmsPattern Recognition Letters, 2006
- Automated detection and classification of lunar craters using multiple approachesAdvances in Space Research, 2006
- Connectivity-based multiple-circle fittingPattern Recognition, 2004
- A new algorithm for ball recognition using circle Hough transform and neural classifierPattern Recognition, 2003
- Circle detection for extracting eddy size and position from satellite imagery of the oceanIEEE Transactions on Geoscience and Remote Sensing, 1994
- Detection of circular structures on satellite imagesInternational Journal of Remote Sensing, 1992
- Detection and classification of circular structures on SPOT imagesIEEE Transactions on Geoscience and Remote Sensing, 1992