Camera model for reciprocal-wedge transform

摘要

The Reciprocal-Wedge Transform (RWT) was presented for space-variant sensing in active vision. The RWT facilitates an anisotropic variable resolution. A projective model for the transform and a potential hardware RWT camera were also proposed. A prominent problem of the previous rudimentary camera model is the requirement of focusing on a deep image plane along the optical axis. In this paper, a new hardware camera model is proposed which will conveniently realize the RWT. The new model overcomes the focus problem by using a lens focusing the paraxial image onto an orthogonally placed RWT plane. Unlike the log-polar sensor, variable sampling is not a requirement of the sensor circuit. Hence, ordinary sensor array of rectangular tessellation and uniform grid size can be used which is much easier to fabricate. Simulation is conducted using an optical software. Our experimental results show that the RWT images generated from the new camera model are well-focused and free from distortion.