BEMD and assisted knowledge based meteorological facsimile map enhancement

An enhancement algorithm for weather fax map with zero-mean Gaussian noise is proposed based on BEMD and self features of meteorological facsimile maps. First decompose the noised meteorological facsimile map to IMF₁ and a residue, then estimate the variance of the noise via BEMD and IMF₁ through the intrinsic relation between the variances of IMF₁ and the noise. According to the current time and location of the ship and the frequency loaded in the meteorologic electrograph, the type and the geographic information of the meteorological facsimile map is obtained. Then the noised map subtracts the geographic information after map matching. Last perform non-local mean filtering and binary value optimization on key words and marks to obtain the refined map. Experiments prove that the proposed algorithm has better performance for the enhancement of meteorological facsimile maps.