return bytes(s, 'utf-8')
cwd = '/home/krematas/code/scanner/examples/apps/image_sets'
db = Database()
encoded_image = db.sources.Files()
frame = db.ops.ImageDecoder(img=encoded_image)
encoded_mask = db.sources.Files()
frame2 = db.ops.ImageDecoder(img=encoded_mask)
db.register_op('ProcessImage', [('image', ColumnType.Video), ('mask', ColumnType.Video)], [('resized', ColumnType.Video)])
db.register_python_kernel('ProcessImage', DeviceType.CPU, cwd + '/dummy_kernel.py')
resized = db.ops.ProcessImage(image=frame, mask=frame2, w=60, h=60)
output_op = db.sinks.FrameColumn(columns={'frame': resized})
job = Job(
op_args={
encoded_image: {'paths': ['/home/krematas/Mountpoints/grail/data/barcelona/players/images/00114_00010.jpg',
'/home/krematas/Mountpoints/grail/data/barcelona/players/images/00114_00008.jpg']},
encoded_mask: {'paths': ['/home/krematas/Mountpoints/grail/data/barcelona/players/masks/00114_00010.png',
'/home/krematas/Mountpoints/grail/data/barcelona/players/masks/00114_00008.png']},
output_op: 'example_resized',
})
[out_table] = db.run(output_op, [job], force=True)
out_table.column('frame').save_mp4('haha')
('mask', mask_path)],
force=True)
print(db.summarize())
print('Failures:', failed)
cam_data = np.load(
'/home/krematas/Mountpoints/grail/data/barcelona/calib/00114.npy').item()
db.register_op('Calibrate', [('frame', ColumnType.Video),
('mask', ColumnType.Video)],
[('resized', ColumnType.Video)])
# Custom Python kernels for ops reside in a separate file, here calibrate_kernel.py.
cwd = '/home/krematas/code/scanner/examples/apps/soccer_calibration'
db.register_python_kernel('Calibrate', DeviceType.CPU,
cwd + '/calibrate_kernel.py')
frame = db.sources.FrameColumn()
mask = db.sources.FrameColumn()
# Then we use our op just like in the other examples.
resized = db.ops.Calibrate(frame=frame,
mask=mask,
w=3840,
h=2160,
A=cam_data['A'],
R=cam_data['R'],
T=cam_data['T'])
output_op = db.sinks.FrameColumn(columns={'frame': resized})
job = Job(
