Spaces:
Build error
Build error
File size: 4,371 Bytes
08d80be 76c1d81 08d80be 76c1d81 08d80be 76c1d81 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 |
import os
import re
from pathlib import Path
from typing import Optional
import numpy as np
import rerun as rr # pip install rerun-sdk
from services.utils.read_write_model import read_model
# From https://github.com/rerun-io/rerun/tree/main/examples/python/structure_from_motion
def read_and_log_sparse_reconstruction(
exp_name: str,
dataset_path: Path,
max_image_number: Optional[int] = 15,
filter_output: bool = True,
filter_min_visible: int = 50,
filter_max_visible: int = 500
) -> str:
try:
rr.init(exp_name)
rec = rr.memory_recording()
cameras, images, points3D = read_model(dataset_path / "sparse", ext=".bin")
print(f"Loaded {len(cameras)} cameras, {len(images)} images, and {len(points3D)} points3D")
if filter_output:
# Filter out noisy points
points3D = {id: point for id, point in points3D.items() if point.rgb.any() and len(point.image_ids) > 4}
rr.log_view_coordinates("/", up="-Y", timeless=True)
print(f"Number of image frames: {len(images)}")
if max_image_number is not None:
# Sample the image sequence to reduce output size
image_ids = sorted(images.keys())
image_ids = np.random.permutation(image_ids)[:max_image_number]
images = {id: images[id] for id in image_ids}
print(f"Number of image frames: {len(images)}")
# Iterate through images (video frames) logging data related to each frame.
for image in sorted(images.values(), key=lambda im: im.name): # type: ignore[no-any-return]
image_file = dataset_path / "images" / image.name
if not os.path.exists(image_file):
continue
# COLMAP sets image ids that don't match the original video frame
idx_match = re.search(r"\d+", image.name)
assert idx_match is not None
frame_idx = int(idx_match.group(0))
quat_xyzw = image.qvec[[1, 2, 3, 0]] # COLMAP uses wxyz quaternions
camera = cameras[image.camera_id]
np.array([1.0, 1.0])
visible = [id != -1 and points3D.get(id) is not None for id in image.point3D_ids]
print(f"Frame {frame_idx} has {np.sum(visible)} visible points")
# Randomly sample points to reduce output size
if filter_output and np.sum(visible) > filter_max_visible:
visible_indices = np.random.permutation(np.where(visible)[0])[:filter_max_visible]
visible = np.zeros_like(visible, dtype=bool)
visible[visible_indices] = True
print(f"Frame {frame_idx} has {np.sum(visible)} visible points after sampling")
visible_ids = image.point3D_ids[visible]
if filter_output and len(visible_ids) < filter_min_visible:
continue
visible_xyzs = [points3D[id] for id in visible_ids]
visible_xys = image.xys[visible]
rr.set_time_sequence("frame", frame_idx)
points = [point.xyz for point in visible_xyzs]
point_colors = [point.rgb for point in visible_xyzs]
point_errors = [point.error for point in visible_xyzs]
rr.log_scalar("plot/avg_reproj_err", np.mean(point_errors), color=[240, 45, 58])
rr.log_points("points", points, colors=point_colors, ext={"error": point_errors})
# COLMAP's camera transform is "camera from world"
rr.log_transform3d(
"camera", rr.TranslationRotationScale3D(image.tvec, rr.Quaternion(xyzw=quat_xyzw)), from_parent=True
)
rr.log_view_coordinates("camera", xyz="RDF") # X=Right, Y=Down, Z=Forward
# Log camera intrinsics
assert camera.model == "PINHOLE"
rr.log_pinhole(
"camera/image",
width=camera.width,
height=camera.height,
focal_length_px=camera.params[:2],
principal_point_px=camera.params[2:],
)
rr.log_image_file("camera/image", img_path=dataset_path / "images" / image.name)
rr.log_points("camera/image/keypoints", visible_xys, colors=[34, 138, 167])
return rec.as_html()
except Exception as e:
print(e) |