Fixed time prediction for moving object pose in three-dimensional space

Prediction for moving object pose can provide advance movement information for measurement and target tracking systems with real-time performance. To solve the problem of real-time prediction for moving object pose in three-dimensional space, a fixed time prediction method with sliding window is proposed in this paper. First, the recurrence relation of the object poses at different timestamps is expressed by recursive motion matrix. Then the object pose at the future timestamp is calculated. Using this method, the object pose information can be updated as time going on, and the model parameters can be adjusted automatically at the same time. In this way, the shortcomings of reduction of operational speed, resulting from the rapid increase in the amount of data, could be overcome. The simulation results show that the absolute prediction errors of both position and orientation are less than 1 x 10(-3), assuming that the measurement system could provide high-precision pose measurement. The experimental results show that the pose predicting accuracy of moving object using the proposed method can reach the same level with that of the actual measurement. In conclusion, the proposed method for moving object pose prediction in three-dimensional space is effective and accurate.