File size: 1,711 Bytes
2f5e0ac | 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 | import os
import sys
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..', '..')))
import utils3d
import numpy as np
import imageio
def run():
image_uv, image_mesh = utils3d.numpy.utils.image_mesh(128, 128)
image_mesh = image_mesh.reshape(-1, 4)
depth = np.ones((128, 128), dtype=np.float32) * 2
depth[32:96, 32:96] = 1
depth = depth.reshape(-1)
intrinsics = utils3d.numpy.transforms.intrinsics_from_fov(1.0, 128, 128).astype(np.float32)
intrinsics = utils3d.numpy.transforms.normalize_intrinsics(intrinsics, 128, 128)
extrinsics = utils3d.numpy.transforms.extrinsics_look_at([0, 0, 1], [0, 0, 0], [0, 1, 0]).astype(np.float32)
pts = utils3d.numpy.transforms.unproject_cv(image_uv, depth, extrinsics, intrinsics)
pts = pts.reshape(-1, 3)
image_mesh = utils3d.numpy.mesh.triangulate(image_mesh, vertices=pts)
perspective = utils3d.numpy.transforms.perspective(1.0, 1.0, 0.1, 10)
view = utils3d.numpy.transforms.view_look_at([1, 0, 1], [0, 0, 0], [0, 1, 0])
mvp = np.matmul(perspective, view)
ctx = utils3d.numpy.rasterization.RastContext(
standalone=True,
backend='egl',
device_index=0,
)
uv = utils3d.numpy.rasterization.rasterize_triangle_faces(
ctx,
pts,
image_mesh,
image_uv,
width=128,
height=128,
mvp=mvp,
)[0]
uv = (np.concatenate([uv, np.zeros((128, 128, 1), dtype=np.float32)], axis=-1) * 255).astype(np.uint8)
imageio.imwrite(os.path.join(os.path.dirname(__file__), '..', '..', 'results_to_check', 'rasterize_uv.png'), uv)
if __name__ == '__main__':
run()
|