The aim of this work is to present first results for 3D velocity field
measurements obtained by Optical Flow PIV coupled with an original technique
of camera calibration . The calibration method search the parameters of an
optimal pinhole camera model using only a single image of a dot grid target,
the known camera focal length and sensor pixel size. After the target marks
have been located, an optimal projective transform is calculated by a least
square method between the target plane coordinates and the image plane
coordinates. The camera model is recovered via a canonical decomposition of
the linear part of the projective transform and an iterative search for the
location where the optical axis intersects the image plane. The interaction
between a flow and a cavity is a test configuration useful to study the
unsteady flows and the development of the 3D instabilities. It also represents
a configuration well adapted to studies relating to transport (aerodynamics
of vehicles) and environment (air renewal in a cavity). Flow issued from a
boundary layer and a cavity interaction was characterised by PIV based on
Optical Flow measurements performed inside and out of the cavity for two
perpendicular planes which showed that the flow is three dimensional.
Therefore we choose this configuration to perform 3D velocity measurements
by stereoscopic PIV.

The results presented in this paper were obtained for low velocities, close
to the laminar turbulent transition, for a given geometry of cavity
characterised by a shape factor H/L=0.5, where H is the height of the
cavity and L its length. The Reynolds number based on the height of the
cavity and the mean velocity is about 4000.