批量对数据添加噪声并生成新命名标注文件

本文链接：https://blog.youkuaiyun.com/weixin_43253464/article/details/125207927

该博客介绍了如何在图像处理中添加椒盐噪声和高斯噪声来增强数据集。通过随机选择并应用这两种噪声类型，创建了新的带有噪声的图像，并相应地更新了标注文件，用于训练机器学习模型，提高其鲁棒性。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

比赛中，通过随机添加高斯噪声和椒盐噪声并生成新命名的标注文件：

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

import cv2
import numpy as np
import os.path
import copy
import random
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

# 图片文件夹路径
file_dir = r'./images/' 
label_dir = r'/home/ubuntu/workspace/tao-experiments/data/data_augmentation/labels/'

cnt = 1

for img_name in os.listdir(file_dir):
    img_path = file_dir + img_name
    label_raw_name = './labels/' + img_name[:-4] + '.txt'
    label_new_name = label_dir + img_name[:-4]+ '_noise' + '.txt'
    img = cv2.imread(img_path)
    rand_num = random.randint(1,10)
    if rand_num >=5:
        img_gauss = addGaussianNoise(img, 0.3)
    else:
        img_gauss = SaltAndPepper(img, 0.3)
    cv2.imwrite(r'./data_augmentation/images/' + img_name[0:-4] + '_noise.jpg',img_gauss)
    shutil.copyfile(label_raw_name, label_new_name)
    if cnt == 210:
        break
    else:
        cnt += 1