Road Detection on Radar Imagery.
Abstract
Detecting a road network on radar imagery is becoming a new area in remote sensing. This report is an attempt to build a facet-based road network detection procedure for radar imagery. This procedure includes the line detection part and the road detection and connection part. The first part analytically detects linear features using a facet valley-finding technique which looks for the first directional derivative zero-crossing of a bicubic facet model. The second part statistically screens the linear features on a component-by-component base and then optimally connects the screened linear features using a dynamic programming algorithm. Keywords: Linear features, Radar imagery, Remote sensing; and Road network detection.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1985
- Accession Number
- ADA164730
Entities
People
- Jungwhan J. Kim
- Robert M. Haralick