Hits:

Indexed by:期刊论文

Date of Publication:2010-07-01

Journal:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING

Included Journals:SCIE、EI

Volume:48

Issue:7

Page Number:2771-2780

ISSN No.:0196-2892

Key Words:Distortion; feature extraction; image analysis; photography; sea ice

Abstract:Digital cameras are widely used on research vessels to ascertain ice conditions in ice-infested seas. However, geometric distortion of images cannot be avoided if the camera is tilted relative to the sea surface. In this study, the geometric distortion induced by oblique photography was quantitatively estimated based on photogrammetric theory to enable accurate measurement of sea ice features. One accurate method of calculating the ice concentration and floe size was developed, and a corresponding simplified method was also tested. Proper selection of camera system parameters (camera height, tilt, and focal length) cannot only avoid possible errors induced by geometric distortion but also make full use of the image information. Additionally, oblique sea ice images obtained during the 2006 Winter Weddell Outflow Study were investigated using the methods developed here. The results revealed that the geometric distortion of oblique images could be accurately corrected by analyzing each image pixel and that the simplified method could be applied to ensure the precision and efficiency of calculating ice features.