def run_dataset(data: DataSetBase) -> None:
"""Link matches pair-wise matches into tracks."""
start = timer()
features, colors, segmentations, instances = tracking.load_features(
data, data.images())
features_end = timer()
matches = tracking.load_matches(data, data.images())
matches_end = timer()
tracks_manager = tracking.create_tracks_manager(
features,
colors,
segmentations,
instances,
matches,
data.config["min_track_length"],
)
tracks_end = timer()
data.save_tracks_manager(tracks_manager)
write_report(
data,
tracks_manager,
features_end - start,
matches_end - features_end,
tracks_end - matches_end,
)
def write_report(data: DataSetBase, preport, pairs, wall_time) -> None:
report = {
"wall_time": wall_time,
"num_pairs": len(pairs),
"pairs": pairs,
}
report.update(preport)
data.save_report(io.json_dumps(report), "matches.json")
def write_report(data: DataSetBase, wall_time: float):
image_reports = []
for image in data.images():
try:
txt = data.load_report("features/{}.json".format(image))
image_reports.append(io.json_loads(txt))
except IOError:
logger.warning("No feature report image {}".format(image))
report = {"wall_time": wall_time, "image_reports": image_reports}
data.save_report(io.json_dumps(report), "features.json")
def run_dataset(data: DataSetBase, input, output) -> None:
recs_base = data.load_reconstruction(input)
if len(recs_base) == 0:
return
rec_base = recs_base[0]
tracks_manager = data.load_tracks_manager()
rec_base.add_correspondences_from_tracks_manager(tracks_manager)
images = data.images()
remaining_images = set(images) - set(rec_base.shots)
gcp = data.load_ground_control_points()
report = {}
rec_report = {}
report["extend_reconstruction"] = [rec_report]
rec, rec_report["grow"] = reconstruction.grow_reconstruction(
data,
tracks_manager,
rec_base,
remaining_images,
gcp,
)
rec_report["num_remaining_images"] = len(remaining_images)
report["not_reconstructed_images"] = list(remaining_images)
data.save_reconstruction([rec], output)
data.save_report(io.json_dumps(report), "reconstruction.json")
def is_high_res_panorama(
data: DataSetBase, image_key: str, image_array: np.ndarray
) -> bool:
"""Detect if image is a panorama."""
exif = data.load_exif(image_key)
if exif:
camera = data.load_camera_models()[exif["camera"]]
w, h = int(exif["width"]), int(exif["height"])
exif_pano = pygeometry.Camera.is_panorama(camera.projection_type)
elif image_array is not None:
h, w = image_array.shape[:2]
exif_pano = False
else:
return False
return w == 2 * h or exif_pano
def _load_segmentation_mask(data: DataSetBase,
image: str) -> Optional[np.ndarray]:
"""Build a mask from segmentation ignore values.
The mask is non-zero only for pixels with segmentation
labels not in segmentation_ignore_values.
"""
ignore_values = data.segmentation_ignore_values(image)
if not ignore_values:
return None
segmentation = data.load_segmentation(image)
if segmentation is None:
return None
return mask_from_segmentation(segmentation, ignore_values)
def match_images(
data: DataSetBase,
config_override: Dict[str, Any],
ref_images: List[str],
cand_images: List[str],
) -> Tuple[Dict[Tuple[str, str], List[Tuple[int, int]]], Dict[str, Any]]:
"""Perform pair matchings between two sets of images.
It will do matching for each pair (i, j), i being in
ref_images and j in cand_images, taking assumption that
matching(i, j) == matching(j ,i). This does not hold for
non-symmetric matching options like WORDS. Data will be
stored in i matching only.
"""
# Get EXIFs data
all_images = list(set(ref_images + cand_images))
exifs = {im: data.load_exif(im) for im in all_images}
# Generate pairs for matching
pairs, preport = pairs_selection.match_candidates_from_metadata(
ref_images,
cand_images,
exifs,
data,
config_override,
)
# Match them !
return (
match_images_with_pairs(data, config_override, exifs, pairs),
preport,
)
def match_images_with_pairs(
data: DataSetBase,
config_override: Dict[str, Any],
exifs: Dict[str, Any],
pairs: List[Tuple[str, str]],
poses: Optional[Dict[str, pygeometry.Pose]] = None,
) -> Dict[Tuple[str, str], List[Tuple[int, int]]]:
"""Perform pair matchings given pairs."""
cameras = data.load_camera_models()
args = list(
match_arguments(pairs, data, config_override, cameras, exifs, poses))
# Perform all pair matchings in parallel
start = timer()
logger.info("Matching {} image pairs".format(len(pairs)))
processes = config_override.get("processes", data.config["processes"])
mem_per_process = 512
jobs_per_process = 2
processes = context.processes_that_fit_in_memory(processes,
mem_per_process)
logger.info("Computing pair matching with %d processes" % processes)
matches = context.parallel_map(match_unwrap_args, args, processes,
jobs_per_process)
logger.info("Matched {} pairs {} in {} seconds ({} seconds/pair).".format(
len(pairs),
log_projection_types(pairs, exifs, cameras),
timer() - start,
(timer() - start) / len(pairs) if pairs else 0,
))
# Index results per pair
resulting_pairs = {}
for im1, im2, m in matches:
resulting_pairs[im1, im2] = m
return resulting_pairs
def run_dataset(data: DataSetBase):
"""Compute features for all images."""
start = timer()
features_processing.run_features_processing(data, data.images(), False)
end = timer()
write_report(data, end - start)
def load_features(
dataset: DataSetBase, images: t.List[str]
) -> t.Tuple[
t.Dict[str, np.ndarray],
t.Dict[str, np.ndarray],
t.Dict[str, np.ndarray],
t.Dict[str, np.ndarray],
]:
logging.info("reading features")
features = {}
colors = {}
segmentations = {}
instances = {}
for im in images:
features_data = dataset.load_features(im)
if not features_data:
continue
features[im] = features_data.points[:, :3]
colors[im] = features_data.colors
semantic_data = features_data.semantic
if semantic_data:
segmentations[im] = semantic_data.segmentation
if semantic_data.has_instances():
instances[im] = semantic_data.instances
return features, colors, segmentations, instances
def undistort_reconstruction(
tracks_manager: Optional[pymap.TracksManager],
reconstruction: types.Reconstruction,
data: DataSetBase,
udata: UndistortedDataSet,
) -> Dict[pymap.Shot, List[pymap.Shot]]:
all_images = set(data.images())
image_format = data.config["undistorted_image_format"]
urec = types.Reconstruction()
urec.points = reconstruction.points
urec.reference = reconstruction.reference
rig_instance_count = itertools.count()
utracks_manager = pymap.TracksManager()
logger.debug("Undistorting the reconstruction")
undistorted_shots = {}
for shot in reconstruction.shots.values():
if shot.id not in all_images:
logger.warning(
f"Not undistorting {shot.id} as it is missing from the dataset's input images."
)
continue
if shot.camera.projection_type == "perspective":
urec.add_camera(perspective_camera_from_perspective(shot.camera))
subshots = [get_shot_with_different_camera(urec, shot, image_format)]
elif shot.camera.projection_type == "brown":
urec.add_camera(perspective_camera_from_brown(shot.camera))
subshots = [get_shot_with_different_camera(urec, shot, image_format)]
elif shot.camera.projection_type in ["fisheye", "fisheye_opencv"]:
urec.add_camera(perspective_camera_from_fisheye(shot.camera))
subshots = [get_shot_with_different_camera(urec, shot, image_format)]
elif pygeometry.Camera.is_panorama(shot.camera.projection_type):
subshot_width = int(data.config["depthmap_resolution"])
subshots = perspective_views_of_a_panorama(
shot, subshot_width, urec, image_format, rig_instance_count
)
else:
logger.warning(
f"Not undistorting {shot.id} with unknown camera type."
)
continue
for subshot in subshots:
if tracks_manager:
add_subshot_tracks(tracks_manager, utracks_manager, shot, subshot)
undistorted_shots[shot.id] = subshots
udata.save_undistorted_reconstruction([urec])
if tracks_manager:
udata.save_undistorted_tracks_manager(utracks_manager)
udata.save_undistorted_shot_ids(
{
shot_id: [ushot.id for ushot in ushots]
for shot_id, ushots in undistorted_shots.items()
}
)
return undistorted_shots
def _load_combined_mask(data: DataSetBase, image: str) -> Optional[np.ndarray]:
"""Combine binary mask with segmentation mask.
Return a mask that is non-zero only where the binary
mask and the segmentation mask are non-zero.
"""
mask = data.load_mask(image)
smask = _load_segmentation_mask(data, image)
return combine_masks(mask, smask)
def run_dataset(data: DataSetBase,
algorithm: reconstruction.ReconstructionAlgorithm) -> None:
"""Compute the SfM reconstruction."""
tracks_manager = data.load_tracks_manager()
if algorithm == reconstruction.ReconstructionAlgorithm.INCREMENTAL:
report, reconstructions = reconstruction.incremental_reconstruction(
data, tracks_manager)
elif algorithm == reconstruction.ReconstructionAlgorithm.TRIANGULATION:
report, reconstructions = reconstruction.triangulation_reconstruction(
data, tracks_manager)
else:
raise RuntimeError(
f"Unsupported algorithm for reconstruction {algorithm}")
data.save_reconstruction(reconstructions)
data.save_report(io.json_dumps(report), "reconstruction.json")
def run_dataset(data: DataSetBase) -> None:
"""Add delaunay meshes to the reconstruction."""
tracks_manager = data.load_tracks_manager()
reconstructions = data.load_reconstruction()
all_shot_ids = set(tracks_manager.get_shot_ids())
for r in reconstructions:
for shot in r.shots.values():
if shot.id in all_shot_ids:
vertices, faces = mesh.triangle_mesh(shot.id, r,
tracks_manager)
shot.mesh.vertices = vertices
shot.mesh.faces = faces
data.save_reconstruction(reconstructions,
filename="reconstruction.meshed.json",
minify=True)
def write_report(data: DataSetBase, tracks_manager, features_time,
matches_time, tracks_time) -> None:
view_graph = [
(k[0], k[1], v)
for k, v in tracks_manager.get_all_pairs_connectivity().items()
]
report = {
"wall_times": {
"load_features": features_time,
"load_matches": matches_time,
"compute_tracks": tracks_time,
},
"wall_time": features_time + matches_time + tracks_time,
"num_images": tracks_manager.num_shots(),
"num_tracks": tracks_manager.num_tracks(),
"view_graph": view_graph,
}
data.save_report(io.json_dumps(report), "tracks.json")
def run_dataset(data: DataSetBase) -> None:
"""Match features between image pairs."""
images = data.images()
start = timer()
pairs_matches, preport = matching.match_images(data, {}, images, images)
matching.save_matches(data, images, pairs_matches)
matching.clear_cache()
end = timer()
write_report(data, preport, list(pairs_matches.keys()), end - start)
def _not_on_blackvue_watermark(p1: np.ndarray, p2: np.ndarray, matches,
im1: str, im2: str,
data: DataSetBase) -> List[Tuple[int, int]]:
"""Filter Blackvue's watermark."""
meta1 = data.load_exif(im1)
meta2 = data.load_exif(im2)
if meta1["make"].lower() == "blackvue":
matches = [m for m in matches if _blackvue_valid_mask(p1[m[0]])]
if meta2["make"].lower() == "blackvue":
matches = [m for m in matches if _blackvue_valid_mask(p2[m[1]])]
return matches
def save_matches(
data: DataSetBase,
images_ref: List[str],
matched_pairs: Dict[Tuple[str, str], List[Tuple[int, int]]],
) -> None:
"""Given pairwise matches (image 1, image 2) - > matches,
save them such as only {image E images_ref} will store the matches.
"""
images_ref_set = set(images_ref)
matches_per_im1 = {im: {} for im in images_ref}
for (im1, im2), m in matched_pairs.items():
if im1 in images_ref_set:
matches_per_im1[im1][im2] = m
elif im2 in images_ref_set:
matches_per_im1[im2][im1] = m
else:
raise RuntimeError(
"Couldn't save matches for {}. No image found in images_ref.".
format((im1, im2)))
for im1, im1_matches in matches_per_im1.items():
data.save_matches(im1, im1_matches)
def load_matches(
dataset: DataSetBase, images: t.List[str]
) -> t.Dict[t.Tuple[str, str], t.List[t.Tuple[int, int]]]:
matches = {}
for im1 in images:
try:
im1_matches = dataset.load_matches(im1)
except IOError:
continue
for im2 in im1_matches:
if im2 in images:
matches[im1, im2] = im1_matches[im2]
return matches
def invent_reference_from_gps_and_gcp(
data: DataSetBase,
images: Optional[List[str]] = None) -> geo.TopocentricConverter:
lat, lon, alt = 0.0, 0.0, 0.0
wlat, wlon, walt = 0.0, 0.0, 0.0
if images is None:
images = data.images()
for image in images:
d = data.load_exif(image)
if "gps" in d and "latitude" in d["gps"] and "longitude" in d["gps"]:
w = 1.0 / max(0.01, d["gps"].get("dop", 15))
lat += w * d["gps"]["latitude"]
lon += w * d["gps"]["longitude"]
wlat += w
wlon += w
if "altitude" in d["gps"]:
alt += w * d["gps"]["altitude"]
walt += w
if not wlat and not wlon:
for gcp in data.load_ground_control_points():
lat += gcp.lla["latitude"]
lon += gcp.lla["longitude"]
wlat += 1
wlon += 1
if gcp.has_altitude:
alt += gcp.lla["altitude"]
walt += 1
if wlat:
lat /= wlat
if wlon:
lon /= wlon
if walt:
alt /= walt
return geo.TopocentricConverter(lat, lon, 0) # Set altitude manually.
def read_images(
queue: queue.Queue,
data: DataSetBase,
images: List[str],
counter: Counter,
expected: int,
force: bool,
) -> None:
full_queue_timeout = 120
for image in images:
logger.info(f"Reading data for image {image} (queue-size={queue.qsize()}")
image_array = data.load_image(image)
if data.config["features_bake_segmentation"]:
segmentation_array = data.load_segmentation(image)
instances_array = data.load_instances(image)
else:
segmentation_array, instances_array = None, None
args = image, image_array, segmentation_array, instances_array, data, force
queue.put(args, block=True, timeout=full_queue_timeout)
counter.increment()
if counter.value() == expected:
logger.info("Finished reading images")
queue.put(None)
def run_dataset(data: DataSetBase, input: str, output: str) -> None:
"""Reconstruct the from a prior reconstruction."""
tracks_manager = data.load_tracks_manager()
rec_prior = data.load_reconstruction(input)
if len(rec_prior) > 0:
report, rec = reconstruction.reconstruct_from_prior(
data, tracks_manager, rec_prior[0])
data.save_reconstruction([rec], output)
data.save_report(io.json_dumps(report), "reconstruction.json")
def _not_on_vermont_watermark(
p1: np.ndarray,
p2: np.ndarray,
matches: List[Tuple[int, int]],
im1: str,
im2: str,
data: DataSetBase,
) -> List[Tuple[int, int]]:
"""Filter Vermont images watermark."""
meta1 = data.load_exif(im1)
meta2 = data.load_exif(im2)
if meta1["make"] == "VTrans_Camera" and meta1["model"] == "VTrans_Camera":
matches = [m for m in matches if _vermont_valid_mask(p1[m[0]])]
if meta2["make"] == "VTrans_Camera" and meta2["model"] == "VTrans_Camera":
matches = [m for m in matches if _vermont_valid_mask(p2[m[1]])]
return matches
def load_features_mask(
data: DataSetBase,
image: str,
points: np.ndarray,
mask_image: Optional[np.ndarray] = None,
) -> np.ndarray:
"""Load a feature-wise mask.
This is a binary array true for features that lie inside the
combined mask.
The array is all true when there's no mask.
"""
if points is None or len(points) == 0:
return np.array([], dtype=bool)
if mask_image is None:
mask_image = _load_combined_mask(data, image)
if mask_image is None:
logger.debug(
"No segmentation for {}, no features masked.".format(image))
return np.ones((points.shape[0], ), dtype=bool)
exif = data.load_exif(image)
width = exif["width"]
height = exif["height"]
orientation = exif["orientation"]
new_height, new_width = mask_image.shape
ps = upright.opensfm_to_upright(
points[:, :2],
width,
height,
orientation,
new_width=new_width,
new_height=new_height,
).astype(int)
mask = mask_image[ps[:, 1], ps[:, 0]]
n_removed = np.sum(mask == 0)
logger.debug("Masking {} / {} ({:.2f}) features for {}".format(
n_removed, len(mask), n_removed / len(mask), image))
return np.array(mask, dtype=bool)
def _extract_exif(image, data: DataSetBase):
with data.open_image_file(image) as fp:
d = exif.extract_exif_from_file(
fp,
partial(data.image_size, image),
data.config["use_exif_size"],
name=image,
)
if data.config["unknown_camera_models_are_different"] and (
not d["model"] or d["model"] == "unknown"):
d["model"] = f"unknown_{image}"
if data.config.get("default_projection_type"):
d["projection_type"] = data.config.get("default_projection_type")
d["camera"] = exif.camera_id(d)
return d
def run_dataset(dataset: DataSetBase, input, output) -> None:
"""Bundle a reconstructions.
Args:
input: input reconstruction JSON in the dataset
output: input reconstruction JSON in the dataset
"""
reconstructions = dataset.load_reconstruction(input)
camera_priors = dataset.load_camera_models()
rig_cameras_priors = dataset.load_rig_cameras()
tracks_manager = dataset.load_tracks_manager()
# load the tracks manager and add its observations to the reconstruction
# go through all the points and add their shots
for reconstruction in reconstructions:
reconstruction.add_correspondences_from_tracks_manager(tracks_manager)
gcp = dataset.load_ground_control_points()
orec.bundle(reconstruction, camera_priors, rig_cameras_priors, gcp,
dataset.config)
dataset.save_reconstruction(reconstructions, output)
def reconstruction_from_metadata(data: DataSetBase, images: Iterable[str]) -> types.Reconstruction:
"""Initialize a reconstruction by using EXIF data for constructing shot poses and cameras."""
data.init_reference()
rig_assignments = rig.rig_assignments_per_image(data.load_rig_assignments())
reconstruction = types.Reconstruction()
reconstruction.reference = data.load_reference()
reconstruction.cameras = data.load_camera_models()
for image in images:
camera_id = data.load_exif(image)["camera"]
if image in rig_assignments:
rig_instance_id, rig_camera_id, _ = rig_assignments[image]
else:
rig_instance_id = image
rig_camera_id = camera_id
reconstruction.add_rig_camera(pymap.RigCamera(pygeometry.Pose(), rig_camera_id))
reconstruction.add_rig_instance(pymap.RigInstance(rig_instance_id))
shot = reconstruction.create_shot(
shot_id=image,
camera_id=camera_id,
rig_camera_id=rig_camera_id,
rig_instance_id=rig_instance_id,
)
shot.metadata = get_image_metadata(data, image)
if not shot.metadata.gps_position.has_value:
reconstruction.remove_shot(image)
continue
gps_pos = shot.metadata.gps_position.value
shot.pose.set_rotation_matrix(rotation_from_shot_metadata(shot))
shot.pose.set_origin(gps_pos)
shot.scale = 1.0
return reconstruction
def detect(
image: str,
image_array: np.ndarray,
segmentation_array: Optional[np.ndarray],
instances_array: Optional[np.ndarray],
data: DataSetBase,
force: bool = False,
) -> None:
log.setup()
need_words = (
data.config["matcher_type"] == "WORDS"
or data.config["matching_bow_neighbors"] > 0
)
has_words = not need_words or data.words_exist(image)
has_features = data.features_exist(image)
if not force and has_features and has_words:
logger.info(
"Skip recomputing {} features for image {}".format(
data.feature_type().upper(), image
)
)
return
logger.info(
"Extracting {} features for image {}".format(data.feature_type().upper(), image)
)
start = timer()
p_unmasked, f_unmasked, c_unmasked = features.extract_features(
image_array, data.config, is_high_res_panorama(data, image, image_array)
)
# Load segmentation and bake it in the data
if data.config["features_bake_segmentation"]:
exif = data.load_exif(image)
s_unsorted, i_unsorted = bake_segmentation(
image_array, p_unmasked, segmentation_array, instances_array, exif
)
p_unsorted = p_unmasked
f_unsorted = f_unmasked
c_unsorted = c_unmasked
# Load segmentation, make a mask from it mask and apply it
else:
s_unsorted, i_unsorted = None, None
fmask = masking.load_features_mask(data, image, p_unmasked)
p_unsorted = p_unmasked[fmask]
f_unsorted = f_unmasked[fmask]
c_unsorted = c_unmasked[fmask]
if len(p_unsorted) == 0:
logger.warning("No features found in image {}".format(image))
size = p_unsorted[:, 2]
order = np.argsort(size)
p_sorted = p_unsorted[order, :]
f_sorted = f_unsorted[order, :]
c_sorted = c_unsorted[order, :]
if s_unsorted is not None:
semantic_data = features.SemanticData(
s_unsorted[order],
i_unsorted[order] if i_unsorted is not None else None,
data.segmentation_labels(),
)
else:
semantic_data = None
features_data = features.FeaturesData(p_sorted, f_sorted, c_sorted, semantic_data)
data.save_features(image, features_data)
if need_words:
bows = bow.load_bows(data.config)
n_closest = data.config["bow_words_to_match"]
closest_words = bows.map_to_words(
f_sorted, n_closest, data.config["bow_matcher_type"]
)
data.save_words(image, closest_words)
end = timer()
report = {
"image": image,
"num_features": len(p_sorted),
"wall_time": end - start,
}
data.save_report(io.json_dumps(report), "features/{}.json".format(image))
def match_candidates_from_metadata(
images_ref: List[str],
images_cand: List[str],
exifs: Dict[str, Any],
data: DataSetBase,
config_override: Dict[str, Any],
) -> Tuple[List[Tuple[str, str]], Dict[str, Any]]:
"""Compute candidate matching pairs between between images_ref and images_cand
Returns a list of pairs (im1, im2) such that (im1 in images_ref) is true.
Returned pairs are unique given that (i, j) == (j, i).
"""
overriden_config = data.config.copy()
overriden_config.update(config_override)
max_distance = overriden_config["matching_gps_distance"]
gps_neighbors = overriden_config["matching_gps_neighbors"]
graph_rounds = overriden_config["matching_graph_rounds"]
time_neighbors = overriden_config["matching_time_neighbors"]
order_neighbors = overriden_config["matching_order_neighbors"]
bow_neighbors = overriden_config["matching_bow_neighbors"]
bow_gps_distance = overriden_config["matching_bow_gps_distance"]
bow_gps_neighbors = overriden_config["matching_bow_gps_neighbors"]
bow_other_cameras = overriden_config["matching_bow_other_cameras"]
vlad_neighbors = overriden_config["matching_vlad_neighbors"]
vlad_gps_distance = overriden_config["matching_vlad_gps_distance"]
vlad_gps_neighbors = overriden_config["matching_vlad_gps_neighbors"]
vlad_other_cameras = overriden_config["matching_vlad_other_cameras"]
data.init_reference()
reference = data.load_reference()
if not all(map(has_gps_info, exifs.values())):
if gps_neighbors != 0:
logger.warn(
"Not all images have GPS info. " "Disabling matching_gps_neighbors."
)
gps_neighbors = 0
max_distance = 0
graph_rounds = 0
images_ref.sort()
if (
max_distance
== gps_neighbors
== time_neighbors
== order_neighbors
== bow_neighbors
== vlad_neighbors
== graph_rounds
== 0
):
# All pair selection strategies deactivated so we match all pairs
d = set()
t = set()
g = set()
o = set()
b = set()
v = set()
pairs = {sorted_pair(i, j) for i in images_ref for j in images_cand if i != j}
else:
d = match_candidates_by_distance(
images_ref, images_cand, exifs, reference, gps_neighbors, max_distance
)
g = match_candidates_by_graph(
images_ref, images_cand, exifs, reference, graph_rounds
)
t = match_candidates_by_time(images_ref, images_cand, exifs, time_neighbors)
o = match_candidates_by_order(images_ref, images_cand, order_neighbors)
b = match_candidates_with_bow(
data,
images_ref,
images_cand,
exifs,
reference,
bow_neighbors,
bow_gps_distance,
bow_gps_neighbors,
bow_other_cameras,
)
v = match_candidates_with_vlad(
data,
images_ref,
images_cand,
exifs,
reference,
vlad_neighbors,
vlad_gps_distance,
vlad_gps_neighbors,
vlad_other_cameras,
{},
)
pairs = d | g | t | o | set(b) | set(v)
pairs = ordered_pairs(pairs, images_ref)
report = {
"num_pairs_distance": len(d),
"num_pairs_graph": len(g),
"num_pairs_time": len(t),
"num_pairs_order": len(o),
"num_pairs_bow": len(b),
"num_pairs_vlad": len(v),
}
return pairs, report
def average_image_size(data: DataSetBase) -> float:
average_size_mb = 0
for camera in data.load_camera_models().values():
average_size_mb += camera.width * camera.height * 4 / 1024 / 1024
return average_size_mb / max(1, len(data.load_camera_models()))