Finding Curves in SAR CCD Images
Technical paper, Sep-Oct 2010
MASSACHUSETTS INST OF TECH LEXINGTON LINCOLN LAB
Pagination or Media Count:
This paper introduces a pattern recognition and computer vision approach to mitigating false alarms in synthetic aperture radar SAR coherence change detection CCD images. In this paper, we perform an automatic detection of roads in SAR CCD images. The approach is based on a curve tracing algorithm originally proposed by Steger with modifications to better suit the goal of curve detection in SAR CCD images 1. In our technique. the traditional Stegers method is used to detect curve points, and cubic splines arc used to approximate the original curve. To detect roads more accurately, preprocessing and outlier removal techniques are performed along with the curve detection.
- Numerical Mathematics
- Active and Passive Radar Detection and Equipment
- Radiofrequency Wave Propagation