In this project, the eye-out-of-hand hand-eye calibration method based on svd algorithm is used. Is there any SVD-based eye-in-hand hand-eye calibration method? If there is, could you please tell me where the corresponding code is, or relevant materials are OK, and available codes not based on svd are OK, please reply!

[yizhipipixia]

Estimate rigid transform with SVD (from Nghia Ho)

def get_rigid_transform(A, B):
assert len(A) == len(B)
N = A.shape[0]; # Total points
centroid_A = np.mean(A, axis=0)
centroid_B = np.mean(B, axis=0)
AA = A - np.tile(centroid_A, (N, 1)) # Centre the points
BB = B - np.tile(centroid_B, (N, 1))
H = np.dot(np.transpose(AA), BB) # Dot is matrix multiplication for array
U, S, Vt = np.linalg.svd(H)
R = np.dot(Vt.T, U.T)
if np.linalg.det(R) < 0: # Special reflection case
Vt[2,:] *= -1
R = np.dot(Vt.T, U.T)
t = np.dot(-R, centroid_A.T) + centroid_B.T
return R, t