Video De-Identification

Person identification based on biometric method has been broadly studied in the last two decades, while information appearing in different actions like bend has been recently acted to this end. Such that in most applications it is sufficient to recognize the performed activity, whereas the ID of persons performing performance is not an important aspect. Since the same human body representations, e.g., body silhouettes, can be employed for both tasks. it is need to create privacy preserving representa tions automatically . We have applied 2D Gaussian filtering to unclear the human body silhouettes that gives information about the person ID. It is done experimentally showed that how the use of filtering affects the person identificat ion and action recogn ition performances in different camera set-ups formed by an arbitrary number of cameras. In addition, the discriminative ability of different activities is examined to detect cases in which it is possible to apply Gaussian filter with a greatest variance.