Merge remote-tracking branch 'upstream/master' (27e669bc) · Commits · aflab / astrogeology / Autocnet

autocnet/graph/network.py

+40 −43

Original line number	Diff line number	Diff line
		import operator
		from functools import reduce
		import operator as op
		import os

		import networkx as nx
		@@ -12,6 +13,7 @@ from autocnet.control.control import C
		from autocnet.fileio import io_json
		from autocnet.fileio.io_gdal import GeoDataset
		from autocnet.matcher import feature_extractor as fe # extract features from image
		from autocnet.matcher import outlier_detector as od

		class CandidateGraph(nx.Graph):
		"""
		@@ -206,7 +208,7 @@ class CandidateGraph(nx.Graph):
		destination_key = dest_group['destination_image'].values[0]
		try:
		edge = self[source_key][destination_key]
		except:
		except: # pragma: no cover
		edge = self[destination_key][source_key]

		if 'matches' in edge.keys():
		@@ -215,18 +217,17 @@ class CandidateGraph(nx.Graph):
		else:
		edge['matches'] = dest_group

		def compute_homography(self, source_key, destination_key, outlier_algorithm=cv2.RANSAC):
		def compute_homographies(self, outlier_algorithm=cv2.RANSAC, clean_keys=[]):
		"""

		For each edge in the (sub) graph, compute the homography
		Parameters
		----------
		source_key : str
		The identifier for the source node
		destination_key : str
		The identifier for the destination node

		outlier_algorithm : object
		An openCV outlier detections algorithm, e.g. cv2.RANSAC

		clean_keys : list
		of string keys to masking arrays
		(created by calling outlier detection)
		Returns
		-------
		transformation_matrix : ndarray
		@@ -234,43 +235,39 @@ class CandidateGraph(nx.Graph):

		mask : ndarray
		Boolean array of the outliers
		"""

		A tuple of the form (transformation matrix, bad entry mask)
		The returned tuple is empty if there is no edge between the source and destination nodes or
		if it exists, but has not been populated with a matches dataframe.
		for source, destination, attributes in self.edges_iter(data=True):
		matches = attributes['matches']

		"""
		if self.has_edge(source_key, destination_key):
		try:
		edge = self[source_key][destination_key]
		except:
		edge = self[destination_key][source_key]
		if 'matches' in edge.keys():
		source_keypoints = []
		destination_keypoints = []

		for i, row in edge['matches'].iterrows():
		source_idx = row['source_idx']
		src_keypoint = [self.node[source_key]['keypoints'][int(source_idx)].pt[0],
		self.node[source_key]['keypoints'][int(source_idx)].pt[1]]
		destination_idx = row['destination_idx']
		dest_keypoint = [self.node[destination_key]['keypoints'][int(destination_idx)].pt[0],
		self.node[destination_key]['keypoints'][int(destination_idx)].pt[1]]

		source_keypoints.append(src_keypoint)
		destination_keypoints.append(dest_keypoint)
		transformation_matrix, mask = cv2.findHomography(np.array(source_keypoints),
		np.array(destination_keypoints),
		outlier_algorithm,
		5.0)
		mask = mask.astype(bool)
		return transformation_matrix, mask
		else:
		return ('', '')
		else:
		return ('','')
		if clean_keys:
		mask = np.prod([attributes[i] for i in clean_keys], axis=0, dtype=np.bool)
		matches = matches[mask]

		full_mask = np.where(mask == True)

		s_coords = np.empty((len(matches), 2))
		d_coords = np.empty((len(matches), 2))

		for i, (idx, row) in enumerate(matches.iterrows()):
		s_idx = int(row['source_idx'])
		d_idx = int(row['destination_idx'])

		s_coords[i] = self.node[source]['keypoints'][s_idx].pt
		d_coords[i] = self.node[destination]['keypoints'][d_idx].pt

		transformation_matrix, ransac_mask = od.compute_homography(s_coords,
		d_coords)

		ransac_mask = ransac_mask.ravel()
		# Convert the truncated RANSAC mask back into a full length mask
		if clean_keys:
		mask[full_mask] = ransac_mask
		else:
		mask = ransac_mask

		attributes['homography'] = transformation_matrix
		attributes['ransac'] = mask

		def to_cnet(self, clean_keys=[]):
		"""
		@@ -291,7 +288,7 @@ class CandidateGraph(nx.Graph):

		# Merge all of the masks
		if clean_keys:
		mask = np.array(list(map(operator.mul, *[attributes[i] for i in clean_keys])))
		mask = np.prod([attributes[i] for i in clean_keys], axis=0, dtype=np.bool)
		matches = matches[mask]

		kp1 = self.node[source]['keypoints']

autocnet/matcher/outlier_detector.py

+36 −0

Original line number	Diff line number	Diff line
		import cv2
		import numpy as np


		@@ -112,3 +113,38 @@ def mirroring_test(matches):
		duplicates.astype(bool, copy=False)
		return duplicates


		def compute_homography(kp1, kp2, outlier_algorithm=cv2.RANSAC, reproj_threshold=5.0):
		"""
		Given two arrays of keypoints compute a homography

		Parameters
		----------
		kp1 : ndarray
		(n, 2) of coordinates from the source image

		kp2 : ndarray
		(n, 2) of coordinates from the destination image

		outlier_algorithm : object
		The openCV algorithm to use for outlier detection

		reproj_threshold : float
		The RANSAC reprojection threshold


		Returns
		-------
		transformation_matrix : ndarray
		The 3x3 transformation matrix

		mask : ndarray
		Boolean array of the outliers
		"""

		transformation_matrix, mask = cv2.findHomography(kp1,
		kp2,
		outlier_algorithm,
		reproj_threshold)
		mask = mask.astype(bool)
		return transformation_matrix, mask
		No newline at end of file

functional_tests/test_two_image.py

+9 −7

Original line number	Diff line number	Diff line
		@@ -46,9 +46,9 @@ class TestTwoImageMatching(unittest.TestCase):
		self.assertEqual(1, cg.number_of_edges())

		# Step: Extract image data and attribute nodes
		cg.extract_features(50)
		cg.extract_features(500)
		for node, attributes in cg.nodes_iter(data=True):
		self.assertIn(len(attributes['keypoints']), [49, 50, 51])
		self.assertIn(len(attributes['keypoints']), range(490, 511))

		# Step: Then apply a FLANN matcher
		fl = FlannMatcher()
		@@ -65,19 +65,21 @@ class TestTwoImageMatching(unittest.TestCase):
		matches = attributes['matches']
		# Perform the symmetry check
		symmetry_mask = od.mirroring_test(matches)
		self.assertIn(symmetry_mask.sum(), range(45, 50))
		self.assertIn(symmetry_mask.sum(), range(430, 461))
		attributes['symmetry'] = symmetry_mask

		# Perform the ratio test
		ratio_mask = od.distance_ratio(matches, ratio=0.95)
		self.assertIn(ratio_mask.sum(), range(30,45))
		self.assertIn(ratio_mask.sum(), range(400, 451))
		attributes['ratio'] = ratio_mask

		mask = np.array(ratio_mask * symmetry_mask)
		self.assertIn(len(matches.loc[mask]), range(4,10))
		# Step: And create a C object
		cnet = cg.to_cnet(clean_keys=['symmetry', 'ratio'])
		self.assertIn(len(matches.loc[mask]), range(75,101))

		cg.compute_homographies(clean_keys=['symmetry', 'ratio'])

		# Step: And create a C object
		cnet = cg.to_cnet(clean_keys=['symmetry', 'ratio', 'ransac'])
		# Step update the serial numbers
		nid_to_serial = {}
		for node, attributes in cg.nodes_iter(data=True):