Figure 13

Prediction accuracy analysis. (a) Variation in error of prediction when the human is out of field of the view of the sensors. The frame number axis shows at which frame the camera was simulated to have lost track of the human, and the prediction window axis shows the number of frames for which the algorithm was asked to predict the position of the human after it had lost track, (b) mean (represented by the bold dots) and the standard deviation (shown by the extended lines) of the error in prediction as more and more frames pass without the human in view. It can be easily seen that the error of prediction of the human's position keeps increasing as the number of frames within which the object is absent increases. This dataset had a 5.9% repetition of data.