A Virtual Curtain for the Detection of Humans and Access Control - Advanced Concepts for Intelligent Vision Systems - page 102

Fig. 4. Illustration of the silhouette reconstruction process

3 Features Extraction and Classification

The closest field to our application in the literature is that of human gait recog-
nition where relevant features are extracted from the time series of the binary
silhouettes of a moving object. Classification is then performed on the basis of
these features (often referred to as signature ) to recover the identity of the walk-
ing human in front of the camera. In our application, we can use these features
extraction and classification techniques to identify the time series of silhouettes
that correspond to a single walking human from the others.
A good introductory reading about gait recognition can be found in [3,4,5].
An extensive review of the existing techniques is presented in [6].