Zhu Shengying

Deep space moving target detection based on technology of image matching

A detection method of moving targets in deep space is researched based on technology of image registration. According to the characteristics of deep space image information, a SUSAN feature point extraction algorithm is modified to inhibit isolated noises which are similar to sidereal feature points, and it effectively eliminates isolated noises. The geometric constraint among stars in the star image settles a feature points matching algorithm. Finally, matching of two images by affine transformation and differential option locates the moving target. Trials show that this method realizes moving targets detect and search in deep space where detector is under the condition with various disturbances.

Autonomous optical navigation approach aided by radio beacon for deep space spacecraft

According to task requests of deep space exploration approach phase such as autonomy, real-time and so on, we did a research into the problem of autonomous navigation at this approach phase and put forward an autonomous navigation project which made use of radio beacon and navigation camera to determine relative position and velocity of the spacecraft. The radio beacon could measure its distance to the spacecraft and the navigation camera could acquire the line-of-sight vector to the target. With Kalman filter we could get the spacecrafts state information. The observability degree of the navigation system was increased by importing a new observation. In order to verify the performance of this autonomous navigation system, Simulink tests based on the practical data of Deep Impact mission were carried on. The results show that the presented method can not only meet navigation requests of approach phase but also gain the navigation system observability degree.