METHOD FOR ESTIMATING ROTATION AXIS AND MASS CENTER OF SPATIAL TARGET BASED ON BINOCULAR OPTICAL FLOWS

摘要

A method for estimating a rotation axis and a mass center of a spatial target based on binocular optical flows. The method includes: extracting feature points from binocular image sequences sequentially and respectively, and calculating binocular optical flows formed thereby; removing areas ineffective for reconstructing a three-dimensional movement trajectory from the binocular optical flows of the feature points, whereby obtaining effective area-constrained binocular optical flows, and reconstructing a three-dimensional movement trajectory of a spatial target; and removing areas with comparatively large errors in reconstructing three-dimensional motion vectors from the optical flows by multiple iterations, estimating a rotation axis according to a three-dimensional motion vector sequence of each of the feature points obtained thereby, obtaining a spatial equation of an estimated rotation axis by weighted average of estimated results of the feature points, and obtaining spatial coordinates of a mass center of the target according to two estimated rotation axes.

主权项

1. A method for estimating a rotation axis and a mass center of a spatial target based on binocular optical flows, the method comprising:
(1) extracting feature points from a stereo image pair respectively; (2) calculating binocular optical flows formed by feature points in image sequences sequentially; (3) removing areas ineffective for reconstructing a three-dimensional movement trajectory from the binocular optical flows of the feature points, thereby obtaining effective area-constrained binocular optical flows; (4) reconstructing a three-dimensional movement trajectory of each of the feature points according to the effective area-constrained binocular optical flows thereof obtained in step (3); (5) calculating a normal vector of a three-dimensional movement trajectory plane of each of the feature points, calculating a cosine distance between the normal vector of each of the feature points and a direction vector of each segment of the trajectory thereof sequentially, and removing a segment of the trajectory of each of the feature points with a maximum modulus of cosine distance; (6) recalculating a normal vector of a trajectory plane of each of the feature points according to the residual trajectory thereof, calculating a cosine distance between the two normal vectors of each of the feature points, determining whether the cosine distance between the two normal vectors of each of the feature points is greater than a preset cosine distance threshold, and returning to step (5) and performing calculation according to the residual trajectory of each of the feature points if yes; (7) estimating a trajectory plane of each of the feature points and center and radius of the trajectory thereof; (8) performing weighted average on normal vectors of the trajectory planes and on centers of the trajectories respectively according to the radius of the trajectory of each of the feature points, whereby obtaining a spatial equation of a final rotation axis; and (9) repeating steps (1)-(8) when the rotation axis of the spatial target changes, whereby obtaining a spatial equation of a second rotation axis, and combining the spatial equation of the second rotation axis with that of the former rotation axis, whereby calculating coordinate of mass center of the spatial target.