official/vision/ops/preprocess_ops_test.py - master - githubmirror/models

official/vision/ops/preprocess_ops_test.py
537 строк17 KB
A. Unique TensorFlower
No public description
09 фев 2026, 19:00
09 фев 2026, 19:00799b0af
# Copyright 2026 The TensorFlow Authors. All Rights Reserved.
#
# Licensed 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.

"""Tests for preprocess_ops.py."""

import io

from absl.testing import parameterized
import numpy as np
from PIL import Image
import tensorflow as tf, tf_keras

from official.vision.ops import preprocess_ops


def _encode_image(image_array, fmt):
  image = Image.fromarray(image_array)
  with io.BytesIO() as output:
    image.save(output, format=fmt)
    return output.getvalue()


class InputUtilsTest(parameterized.TestCase, tf.test.TestCase):

  @parameterized.parameters(
      ([1], 10, 1.0, 0.0),
      ([1, 2], 10, 1.0, 0.0),
      ([1, 2, 3], 10, 1.0, 0.0),
      ([11], 10, 1.0, 0.0),
      ([12, 2], 10, 1.0, 0.0),
      ([13, 2, 3], 10, 1.0, 0.0),
      ([1, 2], 10, 'test', 'pad'),
  )
  def test_pad_to_fixed_size(
      self, input_shape, output_size, original_val, pad_val
  ):
    # Copies input shape to padding shape.
    clip_shape = input_shape[:]
    clip_shape[0] = min(output_size, clip_shape[0])
    padding_shape = input_shape[:]
    padding_shape[0] = max(output_size - input_shape[0], 0)
    expected_outputs = np.concatenate(
        [np.full(clip_shape, original_val), np.full(padding_shape, pad_val)],
        axis=0,
    )

    data = tf.fill(input_shape, original_val)
    output_data = preprocess_ops.clip_or_pad_to_fixed_size(
        data, output_size, constant_values=pad_val
    )
    output_data = output_data.numpy()
    self.assertAllEqual(output_size, output_data.shape[0])
    self.assertAllEqual(expected_outputs, output_data)

  @parameterized.named_parameters(
      dict(
          testcase_name='no_jittering',
          input_size=(100, 200),
          desired_size=(20, 10),
          aug_scale_max=1.0,
          output_scales=(20 / 100, 10 / 200),
      ),
      dict(
          testcase_name='with_jittering',
          input_size=(100, 200),
          desired_size=(20, 10),
          aug_scale_max=2.0,
          output_scales=(20 / 100, 10 / 200),
      ),
      dict(
          testcase_name='no_jittering_with_4_channels',
          input_size=(100, 200),
          desired_size=(20, 10),
          aug_scale_max=1.0,
          output_scales=(20 / 100, 10 / 200),
          channels=4,
      ),
  )
  def test_resize_and_crop_image_not_keep_aspect_ratio(
      self, input_size, desired_size, aug_scale_max, output_scales, channels=3
  ):
    image = tf.convert_to_tensor(np.random.rand(*input_size, channels))

    resized_image, image_info = preprocess_ops.resize_and_crop_image(
        image,
        desired_size=desired_size,
        padded_size=desired_size,
        aug_scale_max=aug_scale_max,
        keep_aspect_ratio=False,
    )
    resized_image_shape = tf.shape(resized_image)

    self.assertAllEqual([*desired_size, channels], resized_image_shape.numpy())
    if aug_scale_max == 1:
      self.assertNDArrayNear(
          [input_size, desired_size, output_scales, [0.0, 0.0]],
          image_info.numpy(),
          1e-5,
      )

  @parameterized.parameters(
      (100, 200, 100, 200, 32, 1.0, 1.0, 128, 224),
      (100, 256, 128, 256, 32, 1.0, 1.0, 128, 256),
      (200, 512, 200, 128, 32, 0.25, 0.25, 224, 128),
  )
  def test_resize_and_crop_image_rectangular_case(
      self,
      input_height,
      input_width,
      desired_height,
      desired_width,
      stride,
      scale_y,
      scale_x,
      output_height,
      output_width,
  ):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))

    desired_size = (desired_height, desired_width)
    resized_image, image_info = preprocess_ops.resize_and_crop_image(
        image,
        desired_size=desired_size,
        padded_size=preprocess_ops.compute_padded_size(desired_size, stride),
    )
    resized_image_shape = tf.shape(resized_image)

    self.assertAllEqual(
        [output_height, output_width, 3], resized_image_shape.numpy()
    )
    self.assertNDArrayNear(
        [
            [input_height, input_width],
            [desired_height, desired_width],
            [scale_y, scale_x],
            [0.0, 0.0],
        ],
        image_info.numpy(),
        1e-5,
    )

  @parameterized.parameters(
      (100, 200, 220, 220, 32, 1.1, 1.1, 224, 224),
      (512, 512, 1024, 1024, 32, 2.0, 2.0, 1024, 1024),
  )
  def test_resize_and_crop_image_square_case(
      self,
      input_height,
      input_width,
      desired_height,
      desired_width,
      stride,
      scale_y,
      scale_x,
      output_height,
      output_width,
  ):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))

    desired_size = (desired_height, desired_width)
    resized_image, image_info = preprocess_ops.resize_and_crop_image(
        image,
        desired_size=desired_size,
        padded_size=preprocess_ops.compute_padded_size(desired_size, stride),
    )
    resized_image_shape = tf.shape(resized_image)

    self.assertAllEqual(
        [output_height, output_width, 3], resized_image_shape.numpy()
    )
    self.assertNDArrayNear(
        [
            [input_height, input_width],
            [desired_height, desired_width],
            [scale_y, scale_x],
            [0.0, 0.0],
        ],
        image_info.numpy(),
        1e-5,
    )

  @parameterized.parameters((1,), (2,))
  def test_resize_and_crop_image_tensor_desired_size(self, aug_scale_max):
    image = tf.convert_to_tensor(np.random.rand(100, 200, 3))

    desired_size = tf.convert_to_tensor((220, 220), dtype=tf.int32)
    resized_image, image_info = preprocess_ops.resize_and_crop_image(
        image,
        desired_size=desired_size,
        padded_size=preprocess_ops.compute_padded_size(desired_size, 32),
        aug_scale_max=aug_scale_max,
    )
    resized_image_shape = tf.shape(resized_image)

    self.assertAllEqual([224, 224, 3], resized_image_shape.numpy())
    self.assertAllEqual([[100, 200], [220, 220]], image_info[:2].numpy())
    if aug_scale_max == 1:  # No random jittering.
      self.assertNDArrayNear(
          [[1.1, 1.1], [0.0, 0.0]],
          image_info[2:].numpy(),
          1e-5,
      )

  @parameterized.parameters(
      (100, 200, 100, 300, 32, 1.0, 1.0, 100, 200, 128, 320),
      (200, 100, 100, 300, 32, 1.0, 1.0, 200, 100, 320, 128),
      (100, 200, 80, 100, 32, 0.5, 0.5, 50, 100, 96, 128),
      (200, 100, 80, 100, 32, 0.5, 0.5, 100, 50, 128, 96),
  )
  def test_resize_and_crop_image_v2(
      self,
      input_height,
      input_width,
      short_side,
      long_side,
      stride,
      scale_y,
      scale_x,
      desired_height,
      desired_width,
      output_height,
      output_width,
  ):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))
    image_shape = tf.shape(image)[0:2]

    desired_size = tf.where(
        tf.greater(image_shape[0], image_shape[1]),
        tf.constant([long_side, short_side], dtype=tf.int32),
        tf.constant([short_side, long_side], dtype=tf.int32),
    )
    resized_image, image_info = preprocess_ops.resize_and_crop_image_v2(
        image,
        short_side=short_side,
        long_side=long_side,
        padded_size=preprocess_ops.compute_padded_size(desired_size, stride),
    )
    resized_image_shape = tf.shape(resized_image)

    self.assertAllEqual(
        [output_height, output_width, 3], resized_image_shape.numpy()
    )
    self.assertNDArrayNear(
        [
            [input_height, input_width],
            [desired_height, desired_width],
            [scale_y, scale_x],
            [0.0, 0.0],
        ],
        image_info.numpy(),
        1e-5,
    )

  @parameterized.parameters(
      (400, 600),
      (600, 400),
  )
  def test_center_crop_image(self, input_height, input_width):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))
    cropped_image = preprocess_ops.center_crop_image(image)
    cropped_image_shape = tf.shape(cropped_image)
    self.assertAllEqual([350, 350, 3], cropped_image_shape.numpy())

  @parameterized.parameters(
      (400, 600),
      (600, 400),
  )
  def test_center_crop_image_v2(self, input_height, input_width):
    image_bytes = tf.constant(
        _encode_image(
            np.uint8(np.random.rand(input_height, input_width, 3) * 255),
            fmt='JPEG',
        ),
        dtype=tf.string,
    )
    cropped_image = preprocess_ops.center_crop_image_v2(
        image_bytes, tf.constant([input_height, input_width, 3], tf.int32)
    )
    cropped_image_shape = tf.shape(cropped_image)
    self.assertAllEqual([350, 350, 3], cropped_image_shape.numpy())

  @parameterized.parameters(
      (400, 600),
      (600, 400),
  )
  def test_random_crop_image(self, input_height, input_width):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))
    _ = preprocess_ops.random_crop_image(image)

  @parameterized.parameters(
      (400, 600),
      (600, 400),
  )
  def test_random_crop_image_v2(self, input_height, input_width):
    image_bytes = tf.constant(
        _encode_image(
            np.uint8(np.random.rand(input_height, input_width, 3) * 255),
            fmt='JPEG',
        ),
        dtype=tf.string,
    )
    _ = preprocess_ops.random_crop_image_v2(
        image_bytes, tf.constant([input_height, input_width, 3], tf.int32)
    )

  @parameterized.parameters((400, 600, 0), (400, 600, 0.4), (600, 400, 1.4))
  def testColorJitter(self, input_height, input_width, color_jitter):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))
    jittered_image = preprocess_ops.color_jitter(
        image, color_jitter, color_jitter, color_jitter
    )
    assert jittered_image.shape == image.shape

  @parameterized.parameters((400, 600, 0), (400, 600, 0.4), (600, 400, 1))
  def testSaturation(self, input_height, input_width, saturation):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))
    jittered_image = preprocess_ops._saturation(image, saturation)
    assert jittered_image.shape == image.shape

  @parameterized.parameters((640, 640, 20), (1280, 1280, 30))
  def test_random_crop(self, input_height, input_width, num_boxes):
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))
    boxes_height = np.random.randint(0, input_height, size=(num_boxes, 1))
    top = np.random.randint(0, high=(input_height - boxes_height))
    down = top + boxes_height
    boxes_width = np.random.randint(0, input_width, size=(num_boxes, 1))
    left = np.random.randint(0, high=(input_width - boxes_width))
    right = left + boxes_width
    boxes = tf.constant(
        np.concatenate([top, left, down, right], axis=-1), tf.float32
    )
    labels = tf.constant(
        np.random.randint(low=0, high=num_boxes, size=(num_boxes,)), tf.int64
    )
    _ = preprocess_ops.random_crop(image, boxes, labels)

  @parameterized.parameters(
      ((640, 640, 3), (1000, 1000), None, (1000, 1000, 3)),
      ((1280, 640, 3), 320, None, (640, 320, 3)),
      ((640, 1280, 3), 320, None, (320, 640, 3)),
      ((640, 640, 3), 320, 100, (100, 100, 3)),
  )
  def test_resize_image(self, input_shape, size, max_size, expected_shape):
    resized_img, image_info = preprocess_ops.resize_image(
        tf.zeros((input_shape)), size, max_size
    )
    self.assertAllEqual(tf.shape(resized_img), expected_shape)
    self.assertAllEqual(image_info[0], input_shape[:-1])
    self.assertAllEqual(image_info[1], expected_shape[:-1])
    self.assertAllEqual(
        image_info[2],
        np.array(expected_shape[:-1]) / np.array(input_shape[:-1]),
    )
    self.assertAllEqual(image_info[3], [0, 0])

  def test_resize_and_crop_masks(self):
    # shape: (2, 1, 4, 3)
    masks = tf.constant([
        [[
            [0, 1, 2],
            [3, 4, 5],
            [6, 7, 8],
            [9, 10, 11],
        ]],
        [[
            [12, 13, 14],
            [15, 16, 17],
            [18, 19, 20],
            [21, 22, 23],
        ]],
    ])
    output = preprocess_ops.resize_and_crop_masks(
        masks, image_scale=[2.0, 0.5], output_size=[2, 3], offset=[1, 0]
    )
    # shape: (2, 2, 3, 3)
    expected_output = tf.constant([
        [
            [
                [3, 4, 5],
                [9, 10, 11],
                [0, 0, 0],
            ],
            [
                [0, 0, 0],
                [0, 0, 0],
                [0, 0, 0],
            ],
        ],
        [
            [
                [15, 16, 17],
                [21, 22, 23],
                [0, 0, 0],
            ],
            [
                [0, 0, 0],
                [0, 0, 0],
                [0, 0, 0],
            ],
        ],
    ])
    self.assertAllEqual(expected_output, output)

  @parameterized.parameters(
      (100, 200, 1.0, 224, 224, 224, 224),
      (512, 512, 1.0, 1024, 1024, 1024, 1024),
  )
  def test_deit3_resize_center_crop(
      self,
      input_height,
      input_width,
      center_crop_fraction,
      desired_height,
      desired_width,
      output_height,
      output_width,
  ):
    # Make sure that with center_crop_ratio = 1; result has desired resolution.
    image = tf.convert_to_tensor(np.random.rand(input_height, input_width, 3))

    desired_size = (desired_height, desired_width)
    center_cropped = preprocess_ops.center_crop_image(
        image, center_crop_fraction=center_crop_fraction
    )
    resized_image = tf.image.resize(
        center_cropped, desired_size, method=tf.image.ResizeMethod.BICUBIC
    )
    resized_image_shape = tf.shape(resized_image)

    self.assertAllEqual(
        [output_height, output_width, 3], resized_image_shape.numpy()
    )

  @parameterized.product(
      prenormalize=[True, False],
      dtype=[tf.uint8, tf.float32, tf.float64, tf.float16],
  )
  def test_normalize_image(self, prenormalize, dtype):
    image = tf.constant([[[0, 200, 255]]], dtype=tf.uint8)
    image = tf.tile(image, [64, 64, 1])

    if dtype != tf.uint8 and prenormalize:
      image = image / 255
    image = tf.cast(image, dtype=dtype)

    if dtype == tf.uint8 or prenormalize:
      normalized_image = preprocess_ops.normalize_image(
          image, offset=[0.5, 0.5, 0.5], scale=[0.5, 0.5, 0.5]
      )
    else:
      normalized_image = preprocess_ops.normalize_image(
          image, offset=[127.0, 127.0, 127.0], scale=[127.0, 127.0, 127.0]
      )
    max_val = tf.reduce_max(normalized_image)
    # If we mistakely use scale=[0.5, 0.5, 0.5] for non-normalized float input,
    # the normalized image data will contain very large values (e.g. 500).
    tf.assert_greater(2.0, max_val)

  @parameterized.named_parameters(
      dict(
          testcase_name='numbers',
          min_quality=20,
          max_quality=80,
          prob_to_apply=1.0,
      ),
      dict(
          testcase_name='tensors',
          min_quality=tf.constant(20),
          max_quality=tf.constant(80),
          prob_to_apply=tf.constant(1.0),
      ),
  )
  def test_random_jpeg_quality(self, min_quality, max_quality, prob_to_apply):
    input_image = np.random.randint(0, 255, (640, 320, 3), dtype=np.uint8)
    aug_image = preprocess_ops.random_jpeg_quality(
        input_image,
        min_quality=min_quality,
        max_quality=max_quality,
        prob_to_apply=prob_to_apply,
    )
    self.assertShapeEqual(input_image, aug_image)
    self.assertDTypeEqual(aug_image, np.uint8)

  @parameterized.parameters(0.25, 0.5, 0.75, 1, 1.25, 1.5)
  def test_resize_and_crop_image_and_masks(self, scale):
    image = tf.convert_to_tensor(np.random.rand(1024, 2048, 3))
    label = tf.convert_to_tensor(np.ones((1, 1024, 2048, 1), dtype=np.int32))
    image, image_info = preprocess_ops.resize_and_crop_image(
        image, (256, 256), (256, 256), scale, scale, centered_crop=True
    )
    image_scale = image_info[2, :]
    offset = image_info[3, :]
    label = preprocess_ops.resize_and_crop_masks(
        label, image_scale, (256, 256), offset, centered_crop=True
    )
    self.assertEqual(image.shape[0:2], label.shape[1:3])
    image_arr = image.numpy()
    label_arr = np.squeeze(label.numpy())

    scaled_height = round(1024 * 256 * scale / 2048)
    scaled_width = round(2048 * 256 * scale / 2048)
    height_offset = max((256 - scaled_height) // 2, 0)
    width_offset = max((256 - scaled_width) // 2, 0)

    self.assertEqual(
        label_arr[
            height_offset : 256 - height_offset,
            width_offset : 256 - width_offset,
        ].mean(),
        1,
    )
    self.assertEqual(label_arr[0:height_offset, :].mean(), 0)
    self.assertEqual(image_arr[0:height_offset, :, :].mean(), 0)
    self.assertEqual(label_arr[256 - height_offset :, :].mean(), 0)
    self.assertEqual(image_arr[256 - height_offset :, :, :].mean(), 0)
    if width_offset > 0:
      self.assertEqual(label_arr[height_offset, 0:width_offset].mean(), 0)
      self.assertEqual(label_arr[height_offset, 256 - width_offset :].mean(), 0)
      self.assertEqual(image_arr[height_offset, 0:width_offset, :].mean(), 0)
      self.assertEqual(
          image_arr[height_offset, 256 - width_offset :, :].mean(), 0
      )


if __name__ == '__main__':
  tf.test.main()