The following is a simple example code for image data augmentation using Python and the OpenCV library, which includes common data augmentation operations:
import cv2
import numpy as np
import os
# 水平翻转
def horizontal_flip(image):
return cv2.flip(image, 1)
# 垂直翻转
def vertical_flip(image):
return cv2.flip(image, 0)
# 随机旋转
def random_rotation(image, angle_range=(-10, 10)):
angle = np.random.randint(angle_range[0], angle_range[1])
height, width = image.shape[:2]
rotation_matrix = cv2.getRotationMatrix2D((width / 2, height / 2), angle, 1)
rotated_image = cv2.warpAffine(image, rotation_matrix, (width, height))
return rotated_image
# 随机裁剪
def random_crop(image, crop_size=(224, 224)):
height, width = image.shape[:2]
left = np.random.randint(0, width - crop_size[0])
top = np.random.randint(0, height - crop_size[1])
right = left + crop_size[0]
bottom = top + crop_size[1]
cropped_image = image[top:bottom, left:right]
return cropped_image
# 添加随机噪声
def random_noise(image, noise_range=(20, 50)):
noise = np.random.randint(noise_range[0], noise_range[1], size=image.shape, dtype=np.uint8)
noisy_image = cv2.add(image, noise)
return np.clip(noisy_image, 0, 255)
# 设置原始图像路径和增强后图像保存路径
original_path = "original_images"
augmented_path = "augmented_images"
# 确保存储路径存在
os.makedirs(augmented_path, exist_ok=True)
# 遍历原始图像路径下的所有图像
for filename in os.listdir(original_path):
if filename.endswith(".jpg") or filename.endswith(".png"):
image_path = os.path.join(original_path, filename)
image = cv2.imread(image_path)
# 水平翻转
h_flip = horizontal_flip(image)
cv2.imwrite(os.path.join(augmented_path, f"flip_h_{
filename}"), h_flip)
# 垂直翻转
v_flip = vertical_flip(image)
cv2.imwrite(os.path.join(augmented_path, f"flip_v_{
filename}"), v_flip)
# 随机旋转
rotated = random_rotation(image)
cv2.imwrite(os.path.join(augmented_path, f"rotated_{
filename}"), rotated)
# 随机裁剪
cropped = random_crop(image)
cv2.imwrite(os.path.join(augmented_path, f"crop_{
filename}"), cropped)
# 添加随机噪声
noisy = random_noise(image)
cv2.imwrite(os.path.join(augmented_path, f"noisy_{
filename}"), noisy)
In this sample code, we use the OpenCV library to load and process images. We define several commonly used data augmentation operation functions, including horizontal flipping, vertical flipping, random rotation, random cropping and adding random noise. Then, we traverse all images under the original image path, perform data enhancement operations on each image, and save it to the enhanced image storage path.
Please note that in order to run this code you need to install the OpenCV library. It can be pip install opencv-pythoninstalled using the command. At the same time, make sure to place the original image under the specified original image path, and set the saving path of the enhanced image.
# -*- coding: utf-8 -*-
import cv2
import numpy as np
import os.path
import copy
# 椒盐噪声
def SaltAndPepper(src, percetage):
SP_NoiseImg = src.copy()
SP_NoiseNum = int(percetage*src.shape[0]*src.shape[1])
for i in range(SP_NoiseNum):
randR = np.random.randint(0, src.shape[0]-1)
randG = np.random.randint(0, src.shape[1]-1)
randB = np.random.randint(0, 3)
if np.random.randint(0, 1) == 0:
SP_NoiseImg[randR, randG, randB] = 0
else:
SP_NoiseImg[randR, randG, randB] = 255
return SP_NoiseImg
# 高斯噪声
def addGaussianNoise(image, percetage):
G_Noiseimg = image.copy()
w = image.shape[1]
h = image.shape[0]
G_NoiseNum = int(percetage*image.shape[0]*image.shape[1])
for i in range(G_NoiseNum):
temp_x = np.random.randint(0, h)
temp_y = np.random.randint(0, w)
G_Noiseimg[temp_x][temp_y][np.random.randint(3)] = np.random.randn(1)[
0]
return G_Noiseimg
# 昏暗
def darker(image, percetage=0.9):
image_copy = image.copy()
w = image.shape[1]
h = image.shape[0]
# get darker
for xi in range(0, w):
for xj in range(0, h):
image_copy[xj, xi, 0] = int(image[xj, xi, 0]*percetage)
image_copy[xj, xi, 1] = int(image[xj, xi, 1]*percetage)
image_copy[xj, xi, 2] = int(image[xj, xi, 2]*percetage)
return image_copy
# 亮度
def brighter(image, percetage=1.5):
image_copy = image.copy()
w = image.shape[1]
h = image.shape[0]
# get brighter
for xi in range(0, w):
for xj in range(0, h):
image_copy[xj, xi, 0] = np.clip(
int(image[xj, xi, 0]*percetage), a_max=255, a_min=0)
image_copy[xj, xi, 1] = np.clip(
int(image[xj, xi, 1]*percetage), a_max=255, a_min=0)
image_copy[xj, xi, 2] = np.clip(
int(image[xj, xi, 2]*percetage), a_max=255, a_min=0)
return image_copy
# 旋转
def rotate(image, angle, center=None, scale=1.0):
(h, w) = image.shape[:2]
# If no rotation center is specified, the center of the image is set as the rotation center
if center is None:
center = (w / 2, h / 2)
m = cv2.getRotationMatrix2D(center, angle, scale)
rotated = cv2.warpAffine(image, m, (w, h))
return rotated
# 翻转
def flip(image):
flipped_image = np.fliplr(image)
return flipped_image
# 图片文件夹路径
file_dir = r'test/img/'
for img_name in os.listdir(file_dir):
img_path = file_dir + img_name
img = cv2.imread(img_path)
# cv2.imshow("1",img)
# cv2.waitKey(5000)
# 旋转
rotated_90 = rotate(img, 90)
cv2.imwrite(file_dir + img_name[0:-4] + '_r90.jpg', rotated_90)
rotated_180 = rotate(img, 180)
cv2.imwrite(file_dir + img_name[0:-4] + '_r180.jpg', rotated_180)
for img_name in os.listdir(file_dir):
img_path = file_dir + img_name
img = cv2.imread(img_path)
# 镜像
flipped_img = flip(img)
cv2.imwrite(file_dir + img_name[0:-4] + '_fli.jpg', flipped_img)
# 增加噪声
# img_salt = SaltAndPepper(img, 0.3)
# cv2.imwrite(file_dir + img_name[0:7] + '_salt.jpg', img_salt)
img_gauss = addGaussianNoise(img, 0.3)
cv2.imwrite(file_dir + img_name[0:-4] + '_noise.jpg', img_gauss)
# 变亮、变暗
img_darker = darker(img)
cv2.imwrite(file_dir + img_name[0:-4] + '_darker.jpg', img_darker)
img_brighter = brighter(img)
cv2.imwrite(file_dir + img_name[0:-4] + '_brighter.jpg', img_brighter)
blur = cv2.GaussianBlur(img, (7, 7), 1.5)
# cv2.GaussianBlur(图像，卷积核，标准差）
cv2.imwrite(file_dir + img_name[0:-4] + '_blur.jpg', blur)