Omnidirectional cameras are becoming increasingly popular in computer vision and robotics. Before performing any task, camera calibration is a step involving metric scene measurement, required in nearly all robotics tasks. In recent years many different methods have been developed to calibrate central omnidirectional cameras. They are based on different camera models and often limited to a specific mirror shape. In this paper we present a design research, manufacture and calibration of an omnidirectional vision sensor. Based on unified approaches, a model of image formation is adopted for our sensor which is made in our School. We have continued the process by a calibration of the sensor, i.e extraction of the parameter allowing the 2D/3D and 3D/2D passages.
展开▼