Fix _ImageDrawTV

[AntoineSimoulin]

The code to draw rotated bounding boxes were wrong for the case where we want to fill the bounding boxes. Indeed we need to plot a polygon since the box maybe be rotated. The following code should now work without error.

from test.common_utils import make_bounding_boxes
from torchvision import tv_tensors
from gallery.transforms.helpers import plot

canvas_size = (360, 360)
num_boxes = 2
bounding_boxes = make_bounding_boxes(
    canvas_size=canvas_size,
    format=tv_tensors.BoundingBoxFormat.XYXYXYXY,
    num_boxes=num_boxes,
    clamping_mode="hard",
)

img = torch.ones(3, canvas_size[0], canvas_size[1])
drawn_boxes = draw_bounding_boxes(
    img,
    v2.ConvertBoundingBoxFormat("xyxyxyxy")(bounding_boxes),
    colors="black",
    fill=True,
)

plot([drawn_boxes])

Tests

Please run tests with

pytest test/test_utils.py -v -k "test_draw_rotated_boxes_fill"
1 passed, 94 deselected in 0.65s

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/vision/9137

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit f410b84 with merge base 646341d ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[NicolasHug]

stamping

[github-actions]

Hey @AntoineSimoulin!

You merged this PR, but no labels were added.
The list of valid labels is available at https://github.com/pytorch/vision/blob/main/.github/process_commit.py