Abstract: Fishing vessel radar is an important part of shipborne system, which has advantage of observation without being affected by visibility under complicated meteorological conditions at sea．In inversion process from radar signal to interpretable image signal, reconstruction of missing data between angle of radar irradiation is the most time-consuming part in inversion process．Based on characteristics of meteorological radar inversion, statistical mechanics model was introduced into reconstruction calculation of radar layer contour, and Maximum a Posteriori-Markov Random Field（MAP-MRF）was applied．MAP-MRF framework realized fast radar image retrieval process of shipborne meteorological radar．The proposed method reduced inversion time and improve radar refresh rate so as to timely and accurately discover marine severe convection disaster targets．