Skip to content

courtvision.console

grab_frames_from_clips(frame_interval=6, max_num_frames=800, output_dir=Path('/Users/benjamindecharmoy/projects/courtvision/balldataset'))

Grabs frames from the clips in the dataset and saves every frame_intervalth frame to output_dir for a maximum of max_num_frames frames.

Parameters:

Name Type Description Default
frame_interval int

Save every frame_interval. Defaults to 6.

 6
max_num_frames int

Max number of frames to save. Defaults to 800.

 800
output_dir Path

Frames are saved to this directory. Defaults to Path( "/Users/benjamindecharmoy/projects/courtvision/balldataset" ).

 Path('/Users/benjamindecharmoy/projects/courtvision/balldataset')
Source code in courtvision/console.py 
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
def grab_frames_from_clips(
    frame_interval: int = 6,
    max_num_frames: int = 800,
    output_dir: Path = Path(
        "/Users/benjamindecharmoy/projects/courtvision/balldataset"
    ),
):
    """Grabs frames from the clips in the dataset and saves every `frame_interval`th
    frame to `output_dir` for a maximum of `max_num_frames` frames.

    Args:
        frame_interval (int, optional): Save every `frame_interval`. Defaults to 6.
        max_num_frames (int, optional): Max number of frames to save. Defaults to 800.
        output_dir (Path, optional): Frames are saved to this directory. Defaults to Path( "/Users/benjamindecharmoy/projects/courtvision/balldataset" ).
    """
    # TODO: Add these as cli args https://github.com/BenjaminDev/courtvision/issues/10
    ANNOTATION_PATH = Path(
        "/Users/benjamindecharmoy/projects/courtvision/datasets/clip_segmentations"
    )
    ANNOTATION_DATA_PATH = Path(
        "/Users/benjamindecharmoy/projects/courtvision/datasets/clip_segmentations/data"
    )
    ANNOTATION_DATA_PATH.mkdir(exist_ok=True, parents=True)

    court_mesh_path = Path(
        "/Users/benjamindecharmoy/projects/courtvision/blender/basic_image.glb"
    )
    output_dir.mkdir(exist_ok=True, parents=True)
    annotations_file = get_latest_file(ANNOTATION_PATH, "json")
    with open(annotations_file, "r") as f:
        dataset = PadelDataset(samples=json.load(f))
    print(f"Found {len(dataset.samples)} clips")
    for i, (frame, uid, match_id) in enumerate(
        frames_from_clip_segments(
            dataset,
            local_path=ANNOTATION_DATA_PATH,
            stream_type=StreamType.VIDEO,
        )
    ):

        if i % frame_interval == 0:
            print(f"Processing frame {i} of clip {uid}")
            cv2.imwrite(
                (output_dir / f"{uid}_{i:04}.png").as_posix(),
                frame["data"].permute(1, 2, 0).numpy()[:, :, ::-1],
            )
        if i >= max_num_frames:
            break