A Deep-Space Trail Forecast Method Based on the Enlargement of Field of View

During the procedure of target tracing, lots of stars will cross the field of view. These stars will disturb the procedure and, the precision of system will be degraded. At the same time, because the optical surveillance system is limited by environment, the targets are often covered by clouds. The above situations will seriously degrade the system´s robustness, and once the target is disappeared, the traditional system can´t track the target again. Traditional forecast methods merely use coder information, so the precision is low. In order to upgrade the precision, we combine the coder information and the missdistance to forecast the next frame target position. The method is based on field range enlargement. Proved by practical experiments, this method can effectively improve the robustness of the target tracking system.