Multi-hypothesis Projection-Based Shift Estimation for Sweeping Panorama Reconstruction

Abstract

Global alignment is an important step in many imaging applications for hand-held cameras. We propose a fast algorithm that can handle large global translations in either x-or y-direction from a pan-tilt camera. The algorithm estimates the translations in x- and y-direction separately using 1D correlation of the absolute gradient projections along the x- and y-axis. Synthetic experiments show that the proposed multiple shift hypotheses approach is robust to translations up to 90% of the image width, whereas other projection-based alignment methods can handle up to 25% only. The proposed approach can also handle larger rotations than other methods. The robustness of the alignment to non-purely translational image motion and moving objects in the scene is demonstrated by a sweeping panorama application on live images from a Canon camera with minimal user interaction.