IMAGE PROCESSING METHOD AND SYSTEM

摘要

A method of comparing two object poses, wherein each object pose is expressed in terms of position, orientation and scale with respect to a common coordinate system, the method comprising:
calculating a distance between the two object poses, the distance being calculated using the distance function:;dsRt(X,Y)=ds2(X,Y)σs2+dr2(X,Y)σr2+dt2(X,Y)σt2.;where X is the object pose of one object and Y is the object pose of the other object,;ds(X,Y)=log(s(X)s(Y)),dr(X,Y)=R(X)-R(Y)F,dt(X,Y)=t(X)-t(Y)s(Y),;s(X) and s(Y) are scalar functions representing the scale of the object poses X and Y respectively, R(X) and R(Y) are matrices expressing the rotation of object poses X and Y respectively, t(X) and t(Y) are vectors expressing the translation of object poses X and Y respectively, and σs, σr and σt are weighting factors for ds, dr and dt respectively.

主权项

1. A method of determining an average of a plurality of object poses, the method comprising using:s(μsRt(χ))=exp(∑iwilogs(Xi)∑iwi),R(μsRt(χ))=sop(∑iwiR(Xi)∑iwi),t(μsRt(χ))=∑iwit(Xi)s2(Xi)/∑iwis2(Xi) to calculate the mean object pose, where χ={(X1,w1),(X2,w2), . . . (XN,wN)} is a set of poses with Xi is the ith pose, wi is a weight associated with the ith pose, s(Xi) are scalar functions representing the scale of the object poses Xi, R(Xi) are matrices expressing the rotation of object poses Xi, t(Xi) are vectors expressing the translation of object poses Xi, sop(X)=arg minYεSO(n,R))∥Y−X∥F is the standard orthogonal projection of matrix X onto SO(n,R) where SO(n,R) is the set of all n-by-n standard orthogonal real matrices and the function sop(X) returns a standard orthogonal matrix Y such that the Frobenius norm of Y-X is minimized.