Additional updates to MLE for performance (7e49566c) · Commits · aflab / astrogeology / Autocnet

autocnet/transformation/transformations.py

+15 −10

Original line number	Diff line number	Diff line
		@@ -11,6 +11,7 @@ from autocnet.matcher.outlier_detector import compute_fundamental_matrix
		from autocnet.utils.utils import make_homogeneous, normalize_vector, crossform
		from autocnet.camera import camera


		class TransformationMatrix(np.ndarray):
		"""
		Abstract Base Class representing a 3x3 transformation matrix.
		@@ -195,10 +196,6 @@ class FundamentalMatrix(TransformationMatrix):
		References
		----------
		.. [Hartley2003]

		Returns
		-------

		"""

		p = camera.idealized_camera()
		@@ -207,9 +204,17 @@ class FundamentalMatrix(TransformationMatrix):
		correspondences1 = self.x1[self.local_mask]
		correspondences2 = self.x2[self.local_mask]

		result = optimize.leastsq(camera.projection_error, p1[:], args=(p,
		if 'method' in kwargs.keys():
		method = kwargs.pop('method')
		else:
		method = 'trf'
		result = optimize.least_squares(camera.projection_error, p1.ravel(), args=(p,
		correspondences1.values,
		correspondences2.calues),
		correspondences2.values),
		method=method,
		xtol=1e-6,
		ftol=1e-6,
		gtol=1e-6,
		**kwargs)

		if result[-1] > 4:
		@@ -217,9 +222,9 @@ class FundamentalMatrix(TransformationMatrix):

		# Scipy solvers are 1D, reshape to the correct form
		pgs = result[0].reshape(3,4)
		m = pgs[:, 3]
		t = pgs[:, 3]
		M = pgs[:, 0:3]
		self[:] = crossform(m).dot(M)
		self[:] = crossform(t).dot(M)

		def refine(self, method=ps.esda.mapclassify.Fisher_Jenks, values=None, bin_id=0, **kwargs):
		"""