def test_pretrained_outdoor(self, device, dtype, data):
loftr = LoFTR('outdoor').to(device, dtype)
data_dev = utils.dict_to(data, device, dtype)
with torch.no_grad():
out = loftr(data_dev)
assert_close(out['keypoints0'], data_dev["loftr_outdoor_tentatives0"])
assert_close(out['keypoints1'], data_dev["loftr_outdoor_tentatives1"])
def test_real_sift_preextract(self, device, dtype, data):
torch.random.manual_seed(0)
# This is not unit test, but that is quite good integration test
feat = SIFTFeature(2000)
matcher = LocalFeatureMatcher(feat, DescriptorMatcher('snn',
0.8)).to(device)
ransac = RANSAC('homography', 1.0, 2048, 10).to(device, dtype)
data_dev = utils.dict_to(data, device, dtype)
pts_src = data_dev['pts0']
pts_dst = data_dev['pts1']
lafs, _, descs = feat(data_dev["image0"])
data_dev["lafs0"] = lafs
data_dev["descriptors0"] = descs
lafs2, _, descs2 = feat(data_dev["image1"])
data_dev["lafs1"] = lafs2
data_dev["descriptors1"] = descs2
with torch.no_grad():
out = matcher(data_dev)
homography, inliers = ransac(out['keypoints0'], out['keypoints1'])
assert inliers.sum().item() > 50 # we have enough inliers
# Reprojection error of 5px is OK
assert_close(transform_points(homography[None], pts_src[None]),
pts_dst[None],
rtol=5e-2,
atol=5)
def test_nomatch(self, device, dtype, data):
matcher = LocalFeatureMatcher(GFTTAffNetHardNet(100),
DescriptorMatcher('snn', 0.8)).to(
device, dtype)
data_dev = utils.dict_to(data, device, dtype)
with torch.no_grad():
out = matcher({
"image0": data_dev["image0"],
"image1": 0 * data_dev["image0"]
})
assert len(out['keypoints0']) == 0
def test_real_clean(self, device, dtype, data):
# generate input data
torch.random.manual_seed(0)
data_dev = utils.dict_to(data, device, dtype)
homography_gt = torch.inverse(data_dev['H_gt'])
homography_gt = homography_gt / homography_gt[2, 2]
pts_src = data_dev['pts0']
pts_dst = data_dev['pts1']
ransac = RANSAC('homography', inl_th=0.5,
max_iter=20).to(device=device, dtype=dtype)
# compute transform from source to target
dst_homo_src, _ = ransac(pts_src, pts_dst)
assert_close(transform_points(dst_homo_src[None], pts_src[None]),
pts_dst[None],
rtol=1e-3,
atol=1e-3)
def test_real_clean(self, device, dtype, data):
torch.random.manual_seed(0)
# generate input data
data_dev = utils.dict_to(data, device, dtype)
pts_src = data_dev['pts0']
pts_dst = data_dev['pts1']
# compute transform from source to target
ransac = RANSAC('fundamental',
inl_th=0.5,
max_iter=20,
max_lo_iters=10).to(device=device, dtype=dtype)
fundamental_matrix, _ = ransac(pts_src, pts_dst)
gross_errors = (sampson_epipolar_distance(pts_src[None],
pts_dst[None],
fundamental_matrix[None],
squared=False) > 1.0)
assert gross_errors.sum().item() == 0
def test_real_keynet(self, device, dtype, data):
torch.random.manual_seed(0)
# This is not unit test, but that is quite good integration test
matcher = LocalFeatureMatcher(KeyNetHardNet(500),
DescriptorMatcher('snn', 0.9)).to(
device, dtype)
ransac = RANSAC('homography', 1.0, 2048, 10).to(device, dtype)
data_dev = utils.dict_to(data, device, dtype)
pts_src = data_dev['pts0']
pts_dst = data_dev['pts1']
with torch.no_grad():
out = matcher(data_dev)
homography, inliers = ransac(out['keypoints0'], out['keypoints1'])
assert inliers.sum().item() > 50 # we have enough inliers
# Reprojection error of 5px is OK
assert_close(transform_points(homography[None], pts_src[None]),
pts_dst[None],
rtol=5e-2,
atol=5)
def test_real_dirty(self, device, dtype, data):
torch.random.manual_seed(0)
# generate input data
data_dev = utils.dict_to(data, device, dtype)
pts_src = data_dev['pts0']
pts_dst = data_dev['pts1']
kp1 = data_dev['loftr_indoor_tentatives0']
kp2 = data_dev['loftr_indoor_tentatives1']
ransac = RANSAC('fundamental',
inl_th=1.0,
max_iter=20,
max_lo_iters=10).to(device=device, dtype=dtype)
# compute transform from source to target
fundamental_matrix, _ = ransac(kp1, kp2)
gross_errors = (sampson_epipolar_distance(pts_src[None],
pts_dst[None],
fundamental_matrix[None],
squared=False) > 10.0)
assert gross_errors.sum().item() < 2
def test_real_dirty(self, device, dtype, data):
# generate input data
torch.random.manual_seed(0)
data_dev = utils.dict_to(data, device, dtype)
homography_gt = torch.inverse(data_dev['H_gt'])
homography_gt = homography_gt / homography_gt[2, 2]
pts_src = data_dev['pts0']
pts_dst = data_dev['pts1']
kp1 = data_dev['loftr_outdoor_tentatives0']
kp2 = data_dev['loftr_outdoor_tentatives1']
ransac = RANSAC('homography', inl_th=3.0, max_iter=30,
max_lo_iters=10).to(device=device, dtype=dtype)
# compute transform from source to target
dst_homo_src, _ = ransac(kp1, kp2)
# Reprojection error of 5px is OK
assert_close(transform_points(dst_homo_src[None], pts_src[None]),
pts_dst[None],
rtol=5,
atol=0.15)
def test_registration_real(self, device, dtype, data):
data_dev = utils.dict_to(data, device, dtype)
IR = ImageRegistrator('homography',
num_iterations=1200,
lr=2e-2,
pyramid_levels=5).to(device, dtype)
model = IR.register(data_dev['image0'], data_dev['image1'])
homography_gt = torch.inverse(data_dev['H_gt'])
homography_gt = homography_gt / homography_gt[2, 2]
h0, w0 = data['image0'].shape[2], data['image0'].shape[3]
h1, w1 = data['image1'].shape[2], data['image1'].shape[3]
model_denormalized = denormalize_homography(model, (h0, w0), (h1, w1))
model_denormalized = model_denormalized / model_denormalized[0, 2, 2]
bbox = torch.tensor([[[0, 0], [w0, 0], [w0, h0], [0, h0]]],
device=device,
dtype=dtype)
bbox_in_2_gt = transform_points(homography_gt[None], bbox)
bbox_in_2_gt_est = transform_points(model_denormalized, bbox)
# The tolerance is huge, because the error is in pixels
# and transformation is quite significant, so
# 15 px reprojection error is not super huge
assert_close(bbox_in_2_gt, bbox_in_2_gt_est, atol=15, rtol=0.1)
