Robust video stabilization based on motion intention for low-cost micro aerial vehicles
Document typeConference report
Rights accessRestricted access - publisher's policy
Currently, different hand-held devices as domestic cameras, smart-phones, tablets, or on-board cameras for robots are becoming popular for video capturing. A main concern with these gadgets is undesired movement between consecutive frames. Video stabilization is a technique with increasing impact for solving this problem. In this paper, a proposal is introduced for robust video stabilization, in particular for on-board cameras in micro aerial vehicles. It is based on a combination of the RANSAC (RANdom SAmple Consensus) algorithm and gray level differences as cost function for local motion parameter estimation, as well as a low-pass filter for global motion smoothing. Experimentation will illustrate about of the robustness proposed solution.
CitationAguilar, W.; Angulo, C. Robust video stabilization based on motion intention for low-cost micro aerial vehicles. A: International Multi-Conference on Systems, Signals and Devices. "Proceedings 11th International Multi-Conference on Systems, Signals and Devices (SSD 2014)". Castelldefels: 2014, p. 1-6.