# Copyright 2022 ByteDance Ltd. and/or its affiliates.
#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
from typing import Any, Tuple, List
import sys
matx = sys.modules['matx']
from .. import ASYNC, GaussianBlurOp
from ._base import BaseInterfaceClass, BatchBaseClass
from ._common import _setup_size
[docs]class GaussianBlur(BaseInterfaceClass):
[docs] def __init__(self,
kernel_size: List[int],
sigma: List[float] = [0.1, 2.0],
device_id: int = -2,
sync: int = ASYNC) -> None:
super().__init__(device_id=device_id, sync=sync)
self._kernel_size: Tuple[int, int] = _setup_size(kernel_size)
for ks in self._kernel_size:
assert 0 < ks and ks % 2 == 1, "Kernel size value should be an odd and positive number."
if len(sigma) == 1:
assert sigma[0] > 0, "If sigma is a single number, it must be positive."
self._sigma: Tuple[float, float] = (sigma[0], sigma[0])
elif len(sigma) == 2:
assert 0. < sigma[0] <= sigma[
1], "sigma values should be positive and of the form (min, max)."
self._sigma: Tuple[float, float] = (sigma[0], sigma[1])
else:
assert False, "sigma should be a single number or a list/tuple with length 2."
[docs] def __call__(self, device: Any, device_str: str, sync: int) -> Any:
return GaussianBlurImpl(device, device_str, self._kernel_size, self._sigma, sync)
class GaussianBlurImpl(BatchBaseClass):
def __init__(self,
device: Any,
device_str: str,
kernel_size: Tuple[int, int],
sigma: Tuple[float, float] = (0.1, 2.0),
sync: int = ASYNC) -> None:
super().__init__()
self.kernel_size: Tuple[int, int] = kernel_size
self.sigma: Tuple[float, float] = sigma
self.device_str: str = device_str
self.sync: int = sync
self.op: GaussianBlurOp = GaussianBlurOp(device)
self.name: str = "GaussianBlurImpl"
def _process(self, imgs: List[matx.NDArray]) -> List[matx.NDArray]:
size: int = len(imgs)
ksizes: List[Tuple[int, int]] = [self.kernel_size] * size
sigmas: List[Tuple[float, float]] = [self.sigma] * size
return self.op(imgs, ksizes, sigmas, sync=self.sync)
def __repr__(self) -> str:
return "{} (kernel_size={},sigma={}, device={}, sync={})".format(
self.name, self.kernel_size, self.sigma, self.device_str, self.sync)