An improved differential optical flow method

Using optical flow method on visual navigation system could ensure MAVs fly like insects in complex environments. The algorithm that combines with local differential techniques and global techniques can yield the advantage of having both robust and dense optical flow fields, and the introduction of the adjacent frame difference, can greatly reduce background errors and improve the reliability of the results. Thus, the purpose of this paper is to propose an adjacent frame difference based local and global combination optical flow (Sub-LK-HS) method, which is more efficient, more robust and more accurate. Furthermore, in order to compare the performance of different optical flow algorithms accurately, the original optical flow evaluation parameters are extended. At last, experimental results based on synthetic and real optical flows demonstrate the effectiveness of the proposed method and it outperforms other general methods.