零基础入门语义分割-地表建筑物识别 Task2 数据扩增-学习笔记

# 首先读取原始图片
img = cv2.imread(train_mask['name'].iloc[0])
mask = rle_decode(train_mask['mask'].iloc[0])
plt.figure(figsize=(16, 8))
plt.subplot(1, 2, 1)
plt.imshow(img)
plt.subplot(1, 2, 2)
plt.imshow(mask)

# 垂直翻转
plt.figure(figsize=(16, 8))
plt.subplot(1, 2, 1)
plt.imshow(cv2.flip(img, 0))
plt.subplot(1, 2, 2)
plt.imshow(cv2.flip(mask, 0))

# 水平翻转
plt.figure(figsize=(16, 8))
plt.subplot(1, 2, 1)
plt.imshow(cv2.flip(img, 0))
plt.subplot(1, 2, 2)
plt.imshow(cv2.flip(mask, 0))

# 随机裁剪
x, y = np.random.randint(0, 256), np.random.randint(0, 256)
plt.figure(figsize=(16, 8))
plt.subplot(1, 2, 1)
plt.imshow(img[x:x+256, y:y+256])
plt.subplot(1, 2, 2)
plt.imshow(mask[x:x+256, y:y+256])

import albumentations as A
# 水平翻转
augments = A.HorizontalFlip(p=1)(image=img, mask=mask)
img_aug, mask_aug = augments['image'], augments['mask']
# 随机裁剪
augments = A.RandomCrop(p=1, height=256, width=256)(image=img, mask=mask)
img_aug, mask_aug = augments['image'], augments['mask']
# 旋转
augments = A.ShiftScaleRotate(p=1)(image=img, mask=mask)
img_aug, mask_aug = augments['image'], augments['mask']

trfm = A.Compose([
    A.Resize(256, 256),
    A.HorizontalFlip(p=0.5),
    A.VerticalFlip(p=0.5),
    A.RandomRotate90(),
])
augments = trfm(image=img, mask=mask)
img_aug, mask_aug = augments['image'], augments['mask']
plt.figure(figsize=(16, 8))
plt.subplot(1, 2, 1)
plt.imshow(augments['image'])
plt.subplot(1, 2, 2)
plt.imshow(augments['mask'])aug

import torch.utils.data as D
class TianChiDataset(D.Dataset):
    def __init__(self, paths, rles, transform):
        self.paths = paths
        self.rles = rles
        self.transform = transform
        self.len = len(paths)
        self.as_tensor = T.Compose([
          #要用torchvision.transform变换，先要转为PIL_img类型
            T.ToPILImage(),
            T.Resize(IMAGE_SIZE),
            T.ToTensor(),
            T.Normalize([0.625, 0.448, 0.688],
                        [0.131, 0.177, 0.101]),
        ])
    def __getitem__(self, index):
        img = cv2.imread(self.paths[index])
        mask = rle_decode(self.rles[index])
        augments = self.transform(image=img, mask=mask)
        return self.as_tensor(augments['image']), augments['mask'][None]
    def __len__(self):
        return self.len

trfm = A.Compose([
    A.Resize(IMAGE_SIZE, IMAGE_SIZE),
    A.HorizontalFlip(p=0.5),
    A.VerticalFlip(p=0.5),
    A.RandomRotate90(),
])
dataset = TianChiDataset(
    train_mask['name'].values,
    train_mask['mask'].fillna('').values,
    trfm
)

loader = D.DataLoader(dataset, batch_size=10, shuffle=True, num_workers=0)

from albumentations import (
    HorizontalFlip, IAAPerspective, ShiftScaleRotate, CLAHE, RandomRotate90,
    Transpose, ShiftScaleRotate, Blur, OpticalDistortion, GridDistortion, HueSaturationValue,
    IAAAdditiveGaussianNoise, GaussNoise, MotionBlur, MedianBlur, IAAPiecewiseAffine,
    IAASharpen, IAAEmboss, RandomBrightnessContrast, Flip, OneOf, Compose
)
import numpy as np
def strong_aug(p=0.5):
    return Compose([
        RandomRotate90(),
        Flip(),
        Transpose(),
        OneOf([
            IAAAdditiveGaussianNoise(),
            GaussNoise(),
        ], p=0.2),
        OneOf([
            MotionBlur(p=0.2),
            MedianBlur(blur_limit=3, p=0.1),
            Blur(blur_limit=3, p=0.1),
        ], p=0.2),
        ShiftScaleRotate(shift_limit=0.0625, scale_limit=0.2, rotate_limit=45, p=0.2),
        OneOf([
            OpticalDistortion(p=0.3),
            GridDistortion(p=0.1),
            IAAPiecewiseAffine(p=0.3),
        ], p=0.2),
        OneOf([
            CLAHE(clip_limit=2),
            IAASharpen(),
            IAAEmboss(),
            RandomBrightnessContrast(),
        ], p=0.3),
        HueSaturationValue(p=0.3),
    ], p=p)
image = np.ones((300, 300, 3), dtype=np.uint8)
mask = np.ones((300, 300), dtype=np.uint8)
whatever_data = "my name"
augmentation = strong_aug(p=0.9)
data = {"image": image, "mask": mask, "whatever_data": whatever_data, "additional": "hello"}
augmented = augmentation(**data)
image, mask, whatever_data, additional = augmented["image"], augmented["mask"], augmented["whatever_data"], augmented["additional"]

OneOf([
            OpticalDistortion(p=0.3),
            GridDistortion(p=0.1),
            IAAPiecewiseAffine(p=0.3),
        ], p=0.2),

trfm = A.Compose([
     #改变HSV
     A.ColorJitter(p=0.5),
     A.HueSaturationValue(p=0.5),
     #resize
     A.Resize(IMAGE_SIZE, IMAGE_SIZE),
    #旋转与缩放
     A.HorizontalFlip(p=0.5),
     A.VerticalFlip(p=0.5),
     A.RandomRotate90(),
     A.ShiftScaleRotate(scale_limit=(-0.3,-0.05),p=0.5),#随机旋转(-45~45),随机缩小0.3~0.05
     #透视变换
     A.IAAPerspective(scale=(0.03,0.05),p=0.5),#透视变化，随机放大0.03~0.05
     #加噪声
     A.ISONoise(p=0.5),
     A.GaussNoise(p=0.5),
     # 平滑处理
     A.GaussianBlur(blur_limit=3,p=0.5),#核大小给3，设置成更大，根本看不清了
 ])