Extracting Curvilinear Features from Synthetic Aperture Radar Images of Arctic Ice: Algorithm Discovery Using the Genetic Programming Paradigm

• Main Authors: Jason Daida Jonathan, Jonathan D. Hommes, Steven J. Ross, John F. Vesecky
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Published: Press 1995
• Subjects: Arctic; ice pack; Sea ice
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.26.3574

This paper focuses on how a method for automated programming (i.e., genetic programming) applies in the computeraided discovery of algorithms that enhance and extract features from remotely sensed images. Highlighted as a case study is the use of this method in the problem of extracting pressure ridge features from ERS-1 SAR imagery; a problem for which there has been no known satisfactory solution. 1. INTRODUCTION Pressure ridges in arctic ice are a significant geophysical feature in sea-ice research. [1] Pressure ridges (and their corresponding keels, which are below-water features) help to transfer kinetic energy from meteorological systems and polar oceanic currents to the ice pack. In particular, pressure ridges and keels significantly increase sea-ice drag coefficients, which subsequently affect sea-ice movement and deformation. To observe meso-scale features such as pressure ridges, researchers have used satellite SAR (synthetic aperture radar) imagery. In such imagery, press.