数据扩增的概念

数据扩增是指不实际增加原始数据，只是对原始数据做一些变换，从而创造出更多的数据。

数据扩增的目的

数据扩增的目的是增加数据量、丰富数据多样性、提高模型的泛化能力。

数据扩增的方法

单样本扩增包括：图像翻转、图像旋转、图像扭曲、图像仿射变换、图像缩放、图像压缩、图像随机crop、图像随机padding、图像对比度调整、亮度调整、色度调整、饱和度调整、色彩抖动、添加噪声、图像模糊、图像区域随机擦除、风格转换、生成对抗网络生成等。

多样本扩增包括：SamplePairing和mixup，二者思路很相近

本文扩增的方法

校验噪声、高斯噪声、调节亮度。

代码

# -*- coding: utf-8 -*-

import cv2
import numpy as np
import os.path
import shutil


# 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


# 亮度
def brightness(image, percetage):
    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


if __name__ == '__main__':
    # 图片文件夹路径
    input_jpg = '../dataset/JPEGImages'
    input_xml = '../dataset/Annotations'
    output_jpg = '../augmentation_data/JPEGImages'
    output_xml = '../augmentation_data/Annotations'

    for img_name in os.listdir(input_jpg):
        name = img_name.split('.')[0]
        print(name)
        print(img_name)
        img_path = os.path.join(input_jpg, img_name)
        img = cv2.imread(img_path)
        xml_src_path = os.path.join(input_xml, name + '.xml')
        xml_dst_path = os.path.join(output_xml, name)

        # 增加噪声
        img_gauss = addGaussianNoise(img, 0.3)
        cv2.imwrite(os.path.join(output_jpg, name + '_noise.jpg'), img_gauss)
        shutil.copyfile(xml_src_path, xml_dst_path + '_noise.xml')
        print("Save " + os.path.join(output_jpg, name + '_noise.jpg') + " Successfully!")

        # 变暗
        img_darker = brightness(img, 0.6)
        cv2.imwrite(os.path.join(output_jpg, name + '_darker.jpg'), img_darker)
        shutil.copyfile(xml_src_path, xml_dst_path + '_darker.xml')
        print("Save " + os.path.join(output_jpg, name + '_darker.jpg') + " Successfully!")
        
        # 变亮
        img_brighter = brightness(img, 1.5)
        cv2.imwrite(os.path.join(output_jpg, name + '_brighter.jpg'), img_brighter)
        shutil.copyfile(xml_src_path, xml_dst_path + '_brighter.xml')
        print("Save " + os.path.join(output_jpg, name + '_brighter.jpg') + " Successfully!")

        blur = cv2.GaussianBlur(img, (7, 7), 1.5)
        #      cv2.GaussianBlur(图像，卷积核，标准差）
        cv2.imwrite(os.path.join(output_jpg, name + '_blur.jpg'), blur)
        shutil.copyfile(xml_src_path, xml_dst_path + '_blur.xml')
        print("Save " + os.path.join(output_jpg, name + '_blur.jpg') + " Successfully!")