Vis-MVSNet
Vis-MVSNet copied to clipboard
Regarding the issue of fusion
The fusion result I got, the number of points is 0, I checked the results of views and update during the process, and the value of the mask has always been false. Is there something wrong with my operation?
'97': {'image': tensor([[[[ 96., 96., 94., ..., 59., 59., 59.], [ 96., 97., 94., ..., 60., 59., 59.], [ 96., 96., 94., ..., 60., 60., 60.], ..., [147., 147., 147., ..., 95., 94., 93.], [148., 148., 149., ..., 95., 94., 93.], [148., 148., 148., ..., 94., 94., 93.]],
[[ 57., 57., 57., ..., 28., 28., 28.],
[ 57., 58., 57., ..., 29., 28., 28.],
[ 57., 57., 57., ..., 29., 29., 29.],
...,
[ 92., 92., 92., ..., 45., 44., 43.],
[ 95., 95., 96., ..., 45., 44., 43.],
[ 95., 95., 95., ..., 44., 44., 43.]],
[[ 28., 28., 28., ..., 7., 7., 7.],
[ 28., 29., 28., ..., 8., 7., 7.],
[ 28., 28., 28., ..., 8., 8., 8.],
...,
[ 38., 38., 38., ..., 10., 9., 8.],
[ 41., 41., 42., ..., 10., 9., 8.],
[ 41., 41., 41., ..., 9., 9., 8.]]]]), 'cam': tensor([[[[ 9.9522e-01, -7.8904e-03, 9.7318e-02, -2.7631e+00],
[ 3.7533e-02, 9.5106e-01, -3.0671e-01, -1.1213e+00],
[-9.0136e-02, 3.0890e-01, 9.4681e-01, 3.4685e+00],
[ 0.0000e+00, 0.0000e+00, 0.0000e+00, 1.0000e+00]],
[[ 6.1405e+02, 0.0000e+00, 3.2000e+02, 0.0000e+00],
[ 0.0000e+00, 6.1405e+02, 1.8000e+02, 0.0000e+00],
[ 0.0000e+00, 0.0000e+00, 1.0000e+00, 0.0000e+00],
[ 5.9801e+00, 0.0000e+00, 2.5600e+02, 1.0600e+01]]]]), 'depth': tensor([[[[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
...,
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[0., 0., 0., ..., 0., 0., 0.]]]]), 'prob': tensor([[[[0.0592, 0.0592, 0.0589, ..., 0.0549, 0.0550, 0.0550],
[0.0592, 0.0592, 0.0589, ..., 0.0549, 0.0550, 0.0550],
[0.0596, 0.0596, 0.0593, ..., 0.0551, 0.0552, 0.0552],
...,
[0.0437, 0.0437, 0.0435, ..., 0.0387, 0.0397, 0.0397],
[0.0444, 0.0444, 0.0440, ..., 0.0391, 0.0402, 0.0402],
[0.0444, 0.0444, 0.0440, ..., 0.0391, 0.0402, 0.0402]],
[[0.1532, 0.1539, 0.1553, ..., 0.1667, 0.1647, 0.1637],
[0.1567, 0.1574, 0.1587, ..., 0.1689, 0.1667, 0.1656],
[0.1638, 0.1643, 0.1653, ..., 0.1732, 0.1706, 0.1693],
...,
[0.1703, 0.1709, 0.1721, ..., 0.1851, 0.1829, 0.1818],
[0.1608, 0.1624, 0.1657, ..., 0.1824, 0.1800, 0.1787],
[0.1560, 0.1582, 0.1625, ..., 0.1811, 0.1785, 0.1772]],
[[0.0741, 0.0972, 0.0919, ..., 0.1583, 0.1776, 0.1896],
[0.0912, 0.1920, 0.2210, ..., 0.1435, 0.1575, 0.1612],
[0.0764, 0.0889, 0.0943, ..., 0.1130, 0.1219, 0.1303],
...,
[0.0960, 0.1201, 0.1221, ..., 0.0967, 0.1112, 0.0946],
[0.1199, 0.1260, 0.1367, ..., 0.0946, 0.1113, 0.1100],
[0.1217, 0.1174, 0.1163, ..., 0.0926, 0.1032, 0.1273]]]]), 'mask': tensor([[[[False, False, False, ..., False, False, False],
[False, False, False, ..., False, False, False],
[False, False, False, ..., False, False, False],
...,
[False, False, False, ..., False, False, False],
[False, False, False, ..., False, False, False],
[False, False, False, ..., False, False, False]]]])}}
'97': {'depth': tensor([[[[0., 0., 0., ..., 0., 0., 0.], [0., 0., 0., ..., 0., 0., 0.], [0., 0., 0., ..., 0., 0., 0.], ..., [0., 0., 0., ..., 0., 0., 0.], [0., 0., 0., ..., 0., 0., 0.], [0., 0., 0., ..., 0., 0., 0.]]]]), 'mask': tensor([[[[False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], ..., [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False]]]])}}
'97': {'points': tensor([], size=(0, 3)), 'colors': tensor([], size=(0, 3)), 'dirs': tensor([], size=(0, 3))}}
[Open3D WARNING] [KDTreeFlann::SetRawData] Failed due to no data. tensor([], size=(0, 3)) [Open3D WARNING] Write PLY failed: point cloud has 0 points.
The depth values here are all 0, I don’t think it should be like this
Hi, could you please find the first step after which the masks become all False?
I had the same problem