Spherical FAST Corner Detector

In this paper, we propose a high speed corner detector in discrete spherical image with full view. Usually, the fisheye camera is used to obtain a wide view image and a full view image, but these wide images obtained by fisheye camera are not rotation invariant, that is, the distortion of scenes on the image varies with their place of image. On the other hand, the content of a spherical image is invariant with the camera rotation. In this paper, we map a wide view image to discrete spherical image for building a rotation invariant spherical image, and modify the conventional FAST (Features from Accelerated Segment Test) corner detector so that it can be applied directly to detect corner features in discrete spherical image. The effectiveness of the proposed method is shown in the experimental results.