New directions in motion-prediction-based systems

Motion prediction has gotten much attention in the past few years. It has endless uses either in real life or in virtual life (gaming). The current dynamic motion prediction models, such as the Markov model, depend only on the object motions’ history. They ignore the object’s direction and speed, which play a key role in dynamic motion prediction. This paper introduces two new probability density functions (PDFs) for elliptic and ellipsoidal regions. Then, it introduces two new motion prediction models based on the new proposed PDFs. This paper proposes new motion detection models that depend not only on the motion’s history but also on moving objects’ speed and direction (objects’ velocity). Moreover, it integrates the proposed models with the Markov model to attain a more precise prediction. Besides, it introduces several applications for the proposed models.