Although a lot of camera calibration methods have been proposed so far, most of them deal with the pinhole camera model, which neglects the defocus effects introduced by real lens systems. In this paper, we present a unified camera calibration method which determines both internal and external camera parameters from several blurred edges detected in a single image. First, we show the calibration procedure using our thin lens based camera model. Then, we describe how to determine the position and the width of blurred features in an image. Experiments have been done with real images and those results have shown the validity of our method.