The invention relates to an
aviation sequence image position estimating method based on SURF (Speeded Up Robust Features). The
aviation sequence image position estimating method can adapt to rotation of an
aviation sequence image, dimension change and
noise interference, realize
accurate estimation of an
airplane. The aviation sequence image position estimating method comprises the following steps of: firstly, structuring SURF dimension space, locating an extreme point by using a quick Hessian matrix, calculating a 64-dimensional SURF describer of the aviation image; then finishing
point match on the basis of a Hessian matrix track; at last, eliminating outliers by using a random sample
consensus algorithm (RANSAC) method, realizing
accurate estimation of a position parameter. The invention aims at a severe
working environment of an aircraft and aims to improve precision of
visual navigation; compared with a traditional
visual navigation method, the method can effectively overcome the shortcoming of nonsensitive to dimension, rotation, light and other factors in the process of position
estimation by an unmanned plane, and obviously improve the matching precision of a real-time image and a
reference image; and therefore, the method has an important actual meaning for a
visual navigation project application of the unmanned plane.