File size: 3,080 Bytes
2796071
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""Project animated point cloud onto image using camera parameters."""

import json

import numpy as np
from PIL import Image
from scipy.ndimage import gaussian_filter


def project_points(
    points: np.ndarray, camera: dict, img_size: int
) -> tuple[np.ndarray, np.ndarray]:
    """Project 3D points to pixel coordinates.

    Uses Blender camera convention:
      X_cam = points @ R^T + T, camera looks along -Z.

    Args:
        points: (N, 3) world-space points.
        camera: dict with R, T, focal_length_ndc, principal_point_ndc.
        img_size: output image resolution (square).

    Returns:
        (x_pixels, y_pixels): each (N,) array of pixel coordinates.
    """
    R = np.asarray(camera["R"], dtype=np.float64)
    T = np.asarray(camera["T"], dtype=np.float64)
    focal = camera["focal_length_ndc"]
    pp = camera["principal_point_ndc"]

    X_cam = points @ R.T + T
    depth = -X_cam[:, 2]
    Z = np.clip(depth, 1e-4, None)

    x_ndc = focal[0] * X_cam[:, 0] / Z + pp[0]
    y_ndc = focal[1] * X_cam[:, 1] / Z + pp[1]

    x_px = img_size / 2.0 * (1.0 + x_ndc)
    y_px = img_size / 2.0 * (1.0 - y_ndc)
    return x_px, y_px


def render_projection(
    image: Image.Image, x_px: np.ndarray, y_px: np.ndarray, img_size: int
) -> np.ndarray:
    """Overlay projected points as a heatmap on an image.

    Returns:
        (H, W, 3) float32 blended image in [0, 1].
    """
    valid = (x_px >= 0) & (x_px < img_size) & (y_px >= 0) & (y_px < img_size)
    mask = np.zeros((img_size, img_size), dtype=np.float32)
    mask[y_px[valid].astype(int), x_px[valid].astype(int)] = 1.0
    mask = np.clip(gaussian_filter(mask, sigma=1.5) * 5.0, 0, 1)

    img_np = np.array(image.convert("RGB")).astype(np.float32) / 255.0
    overlay = np.array([1.0, 0.3, 0.1])
    alpha = 0.4 * mask[..., None]
    blended = img_np * (1 - alpha) + overlay * alpha
    return np.clip(blended, 0, 1)


if __name__ == "__main__":
    import argparse

    parser = argparse.ArgumentParser(description="Project point cloud onto an image.")
    parser.add_argument("--image", required=True, help="Path to input image")
    parser.add_argument("--points", required=True, help="Path to surfaces.npy (T,V,6)")
    parser.add_argument(
        "-t",
        "--timestep",
        type=int,
        default=0,
        help="Keyframe index to project (default: 0)",
    )
    parser.add_argument("--camera", required=True, help="Path to camera.json")
    parser.add_argument("--output", required=True, help="Path to output image")
    args = parser.parse_args()

    # (T, V, 6) -> take keyframe t, xyz only
    surfaces = np.load(args.points)
    points = surfaces[args.timestep, :, :3].astype(np.float64)

    with open(args.camera) as f:
        camera = json.load(f)
    image = Image.open(args.image)
    img_size = image.size[0]

    x_px, y_px = project_points(points, camera, img_size)
    result = render_projection(image, x_px, y_px, img_size)

    out = Image.fromarray((result * 255).astype(np.uint8))
    out.save(args.output)
    print(f"Saved projection to {args.output}")