Sea Ice Image Processing Research Based On GOCI Satellite

The GOCI (Geostationary Ocean Color Imager) satellite with high spatial resolution and high temporal resolution can satisfy the needs of monitoring drift of sea ice to provide real-time and high frequency data. The real-time and accurate monitoring information can be obtained to support the economic activities related with sea. The SIFT (Scale-invariant feature transform) algorithm is improved to realize two optimization schemes of the automatic selection of threshold and manual elimination of edge points to make use of the GOCI satellite image information effectively. The improved SIFT algorithm is utilized to extract the feature points of images of sea ice and match images. The feasibility and practicability of the improved algorithm is verified through comparison between the improved algorithm and conventional algorithm.