wassname/scikit-image
1
0
Fork 0
mirror of https://github.com/wassname/scikit-image.git synced 2026-09-11 12:43:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

Implementing descriptor_orb; for loop in Cython

Browse Source
This commit is contained in:
Ankit Agrawal
2013-11-29 20:42:47 +01:00
committed by Johannes Schönberger
parent e8eb2de6e9
commit 07fdfdb5fe
4 changed files with 69 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
skimage/feature/orb.py
+30 -9
View File
@@ -7,6 +7,7 @@ from skimage.feature import (corner_fast, corner_orientations, corner_peaks,
corner_harris)
from skimage.transform import pyramid_gaussian
from .orb_cy import _orb_loop
def keypoints_orb(image, n_keypoints=200, fast_n=9, fast_threshold=0.20,
harris_k=0.05, downscale_factor=np.sqrt(2), n_scales=5):
@@ -42,7 +43,7 @@ def keypoints_orb(image, n_keypoints=200, fast_n=9, fast_threshold=0.20,
Downscale factor for the image pyramid.
n_scales : int
Number of scales from the bottom of the image pyramid to extract
the features from.
the features from.
Returns
-------
@@ -59,7 +60,7 @@ def keypoints_orb(image, n_keypoints=200, fast_n=9, fast_threshold=0.20,
"ORB : An efficient alternative to SIFT and SURF"
http://www.vision.cs.chubu.ac.jp/CV-R/pdf/Rublee_iccv2011.pdf
"""
"""
image = np.squeeze(image)
if image.ndim != 2:
raise ValueError("Only 2-D gray-scale images supported.")
@@ -99,7 +100,8 @@ def keypoints_orb(image, n_keypoints=200, fast_n=9, fast_threshold=0.20,
return keypoints[best_indices], orientations[best_indices], scales[best_indices]
def descriptor_orb(image, keypoints, keypoints_angle):
def descriptor_orb(image, keypoints, keypoints_orientations,
keypoints_scales, downscale_factor=np.sqrt(2), n_scales=5):
image = np.squeeze(image)
if image.ndim != 2:
@@ -107,14 +109,33 @@ def descriptor_orb(image, keypoints, keypoints_angle):
image = img_as_float(image)
mask = _mask_border_keypoints(image, keypoints, 13)
keypoints = keypoints[mask]
keypoints_angle = keypoints_angle[mask]
pyramid = list(pyramid_gaussian(image, n_scales - 1, downscale_factor))
pos1 = binary_tests[:, :2]
pos2 = binary_tests[:, 2:]
pos0 = binary_tests[:, :2].astype(np.int32)
pos1 = binary_tests[:, 2:].astype(np.int32)
descriptors = np.zeros((keypoints.shape[0], 256), dtype=bool)
pos0 = np.ascontiguousarray(pos0)
pos1 = np.ascontiguousarray(pos1)
descriptors = np.empty((0, 256), dtype=np.bool)
for k in range(n_scales):
curr_image = np.ascontiguousarray(pyramid[k])
curr_scale_mask = keypoints_scales == k
curr_scale_kpts = keypoints[curr_scale_mask]
curr_scale_kpts_orientation = keypoints_orientations[curr_scale_mask]
border_mask = _mask_border_keypoints(curr_image, curr_scale_kpts, 13)
curr_scale_kpts = curr_scale_kpts[border_mask]
curr_scale_kpts_orientation = curr_scale_kpts_orientation[border_mask]
curr_scale_descriptors = np.zeros((curr_scale_kpts.shape[0], 256), dtype=np.bool, order='C')
curr_scale_kpts = np.ascontiguousarray(curr_scale_kpts)
curr_scale_kpts_orientation = np.ascontiguousarray(curr_scale_kpts_orientation)
_orb_loop(curr_image, curr_scale_descriptors.view(np.uint8), curr_scale_kpts, curr_scale_kpts_orientation, pos0, pos1)
descriptors = np.vstack((descriptors, curr_scale_descriptors))
for i in range(keypoints.shape[0]):
angle = keypoints_angle[i]