We present a novel technique for effectively calibrating a binocular stereo rig using the information from both scenes and classical calibration objects. The calibration provided by the classical methods is only valid for the space near the position of the calibration object. Our technique tries to make the best use of the rigidity of the geometry between two cameras. The idea is to first estimate precisely the epipolar geometry which is valid for a wide range in space from all available matches, extracted from both the environment and the calibration objects. This allows us to conduct an accurate projective reconstruction. Using the a priori knowledge of the calibration object, we are finally able to calibrate the stereo rig in a Euclidean space. The proposed technique has been tested with a number of real images, and significant improvement has been observed.