Projects

VIMS Lab works on Sea Ice include high-resolution motion analysis of sea ice fields using optical flow, deep learning for detecting sea ice leads, and a web app for analyzing sea ice rasters. Additionally, we work on 3D reconstruction of sea ice with custom stereo algorithms, understanding non-rigid sea ice dynamics, and coastal ice dynamics through reflector track extraction. Also, some works focus on motion analysis of satellite imagery and 3D non-rigid motion analysis from cloud image sequences.

card-image

card-image

SEA ICE MOTION ESTIMATION DATASET FROM SIDEx

A dataset in NetCDF format containing our motion estimation for SIDEx

card-image

SEA ICE MOTION ANALYSIS

A method for high resolution and accurate motion estimates of the sea ice field

card-image

card-image

SEA ICE LEAD DETECTION

A deep learning approach for sea ice lead detection from visible and C-band images

card-image

card-image

SEA ICE IMAGE DATABASE

A web app for the analysis of sea ice rasters captured by the PSITRES camera system

card-image

card-image

PSITRES

A 3D camera system build by VIMS Lab to quantitatively measure sea ice

card-image

card-image

3D RECONSTRUCTION OF SEA ICE

A custom stereo algorithm for textureless data to reconstruct sea icescape

card-image

card-image

NON-RIGID SEA ICE DYNAMICS

A non-intrusive method to track sea ice and understand the sea ice mass balance

card-image

card-image

COASTAL ICE DYNAMICS

A method to automatically extract tracks of reflectors and derive velocity fields for moving ice

card-image

MOTION ANALYSIS OF SATELLITE IMAGERY

A method to extracture structure/motion in a sequences of cloud images

card-image

card-image

3D NONRIGID MOTION ANALYSIS

A method to extract motion and structure from 2D cloud image sequences without known correspondences