Merge pull request #72 from glostis/s2p_sift_cfg

Remove use of s2p.cfg in s2p.sift

s2p/__init__.py

@@ -23,7 +23,6 @@
 import os.path
 import json
 import datetime
-import subprocess
 import multiprocessing
 
 import numpy as np
@@ -43,7 +42,6 @@
 from s2p import triangulation
 from s2p import fusion
 from s2p import visualisation
-from s2p import ply
 
 
 def pointing_correction(tile, i):
@@ -63,7 +61,11 @@ def pointing_correction(tile, i):
 
     # correct pointing error
     print('correcting pointing on tile {} {} pair {}...'.format(x, y, i))
-    A, m = pointing_accuracy.compute_correction(img1, img2, rpc1, rpc2, x, y, w, h)
+    method = 'relative' if cfg['relative_sift_match_thresh'] is True else 'absolute'
+    A, m = pointing_accuracy.compute_correction(
+        img1, img2, rpc1, rpc2, x, y, w, h, method,
+        cfg['sift_match_thresh'], cfg['max_pointing_error']
+    )
 
     if A is not None:  # A is the correction matrix
         np.savetxt(os.path.join(out_dir, 'pointing.txt'), A, fmt='%6.3f')

s2p/estimation.py

@@ -5,8 +5,6 @@
 
 import numpy as np
 
-from s2p import common
-
 
 def fundamental_matrix_cameras(P1, P2):
     """

s2p/initialization.py

@@ -7,19 +7,15 @@
 import sys
 import json
 import copy
-import shutil
 import rasterio
-import warnings
 import numpy as np
 import rpcm
-from pyproj.exceptions import CRSError
 
 from s2p import common
 from s2p import geographiclib
 from s2p import rpc_utils
 from s2p import masking
 from s2p import parallel
-from s2p import geographiclib
 from s2p.config import cfg
 
 # This function is here as a workaround to python bug #24313 When

s2p/masking.py

@@ -3,7 +3,6 @@
 # Copyright (C) 2015, Enric Meinhardt <[email protected]>
 # Copyright (C) 2015, Julien Michel <[email protected]>
 
-import os
 import subprocess
 import numpy as np
 import warnings

s2p/pointing_accuracy.py

@@ -6,7 +6,6 @@
 
 import os
 import numpy as np
-import rpcm
 
 from s2p import sift
 from s2p import rpc_utils
@@ -100,7 +99,8 @@ def local_translation(r1, r2, x, y, w, h, m):
     return A
 
 
-def compute_correction(img1, img2, rpc1, rpc2, x, y, w, h):
+def compute_correction(img1, img2, rpc1, rpc2, x, y, w, h,
+                       method, sift_thresh, epipolar_threshold):
     """
     Computes pointing correction matrix for specific ROI
 
@@ -113,13 +113,16 @@ def compute_correction(img1, img2, rpc1, rpc2, x, y, w, h):
             image. (x, y) is the top-left corner, and (w, h) are the dimensions
             of the rectangle. The ROI may be as big as you want. If bigger than
             1 Mpix, only five crops will be used to compute sift matches.
+        method, sift_thresh, epipolar_threshold: see docstring of
+            s2p.sift.keypoints_match()
 
     Returns:
         a 3x3 matrix representing the planar transformation to apply to img2 in
         order to correct the pointing error, and the list of sift matches used
         to compute this correction.
     """
-    m = sift.matches_on_rpc_roi(img1, img2, rpc1, rpc2, x, y, w, h)
+    m = sift.matches_on_rpc_roi(img1, img2, rpc1, rpc2, x, y, w, h,
+                                method, sift_thresh, epipolar_threshold)
 
     if m is not None:
         A = local_translation(rpc1, rpc2, x, y, w, h, m)

s2p/rectification.py

@@ -7,14 +7,12 @@
 import warnings
 
 import numpy as np
-import rpcm
 
 from s2p import rpc_utils
 from s2p import estimation
 from s2p import evaluation
 from s2p import common
 from s2p import visualisation
-from s2p import block_matching
 from s2p.config import cfg
 
 
@@ -297,7 +295,7 @@ def rectify_pair(im1, im2, rpc1, rpc2, x, y, w, h, out1, out2, A=None, sift_matc
             module.
         sift_matches (optional): Nx4 numpy array containing a list of sift
             matches, in the full image coordinates frame
-        method (default: 'rpc'): option to decide wether to use rpc of sift
+        method (default: 'rpc'): option to decide whether to use rpc of sift
             matches for the fundamental matrix estimation.
         {h,v}margin (optional): horizontal and vertical margins added on the
             sides of the rectified images

s2p/rpc_utils.py

@@ -3,8 +3,6 @@
 # Copyright (C) 2015, Enric Meinhardt <[email protected]>
 
 
-import json
-import datetime
 import warnings
 import rasterio
 import numpy as np

s2p/sift.py

@@ -12,10 +12,8 @@
 from numpy.ctypeslib import ndpointer
 import ransac
 
-from s2p import common
 from s2p import rpc_utils
 from s2p import estimation
-from s2p.config import cfg
 
 # Locate sift4ctypes library and raise an ImportError if it can not be
 # found This call will raise an exception if library can not be found,
@@ -155,7 +153,7 @@ def keypoints_match(k1, k2, method='relative', sift_thresh=0.6, F=None,
         k2 (array): numpy array of shape (m, 132), where each row represents a
             sift keypoint
         method (optional, default is 'relative'): flag ('relative' or
-            'absolute') indicating wether to use absolute distance or relative
+            'absolute') indicating whether to use absolute distance or relative
             distance
         sift_thresh (optional, default is 0.6): threshold for distance between SIFT
             descriptors. These descriptors are 128-vectors, whose coefficients
@@ -239,7 +237,8 @@ def keypoints_match_from_nparray(k1, k2, method, sift_threshold,
     return matches.reshape((nb_matches.value, 4))
 
 
-def matches_on_rpc_roi(im1, im2, rpc1, rpc2, x, y, w, h):
+def matches_on_rpc_roi(im1, im2, rpc1, rpc2, x, y, w, h,
+                       method, sift_thresh, epipolar_threshold):
     """
     Compute a list of SIFT matches between two images on a given roi.
 
@@ -252,6 +251,8 @@ def matches_on_rpc_roi(im1, im2, rpc1, rpc2, x, y, w, h):
         x, y, w, h: four integers defining the rectangular ROI in the first
             image. (x, y) is the top-left corner, and (w, h) are the dimensions
             of the rectangle.
+        method, sift_thresh, epipolar_threshold: see docstring of
+            s2p.sift.keypoints_match()
 
     Returns:
         matches: 2D numpy array containing a list of matches. Each line
@@ -264,16 +265,13 @@ def matches_on_rpc_roi(im1, im2, rpc1, rpc2, x, y, w, h):
     rpc_matches = rpc_utils.matches_from_rpc(rpc1, rpc2, x, y, w, h, 5)
     F = estimation.affine_fundamental_matrix(rpc_matches)
 
-    # sift matching method:
-    method = 'relative' if cfg['relative_sift_match_thresh'] is True else 'absolute'
-
     # if less than 10 matches, lower thresh_dog. An alternative would be ASIFT
     thresh_dog = 0.0133
-    for i in range(2):
+    for _ in range(2):
         p1 = image_keypoints(im1, x, y, w, h, thresh_dog=thresh_dog)
         p2 = image_keypoints(im2, x2, y2, w2, h2, thresh_dog=thresh_dog)
-        matches = keypoints_match(p1, p2, method, cfg['sift_match_thresh'], F,
-                                  epipolar_threshold=cfg['max_pointing_error'],
+        matches = keypoints_match(p1, p2, method, sift_thresh, F,
+                                  epipolar_threshold=epipolar_threshold,
                                   model='fundamental')
         if matches is not None and matches.ndim == 2 and matches.shape[0] > 10:
             break

s2p/visualisation.py

@@ -4,13 +4,9 @@
 
 import numpy as np
 import rasterio
-import rpcm
 
 from s2p import common
-from s2p import sift
-from s2p import estimation
 from s2p import rpc_utils
-import s2p.pointing_accuracy
 
 
 def plot_line(im, x1, y1, x2, y2, colour):

tests/sift_test.py

@@ -41,7 +41,10 @@ def test_matches_on_rpc_roi():
     img2 = data_path('input_triplet/img_02.tif')
     rpc1 = rpcm.rpc_from_geotiff(img1)
     rpc2 = rpcm.rpc_from_geotiff(img2)
-    computed = sift.matches_on_rpc_roi(img1, img2, rpc1, rpc2, 100, 100, 200, 200)
+    computed = sift.matches_on_rpc_roi(
+        img1, img2, rpc1, rpc2, 100, 100, 200, 200,
+        method='relative', sift_thresh=0.6, epipolar_threshold=10
+    )
     expected = np.loadtxt(data_path('expected_output/units/matches_on_rpc_roi.txt'))
     np.testing.assert_allclose(computed, expected, rtol=0.01, atol=0.1,
                                verbose=True)