计算一个数据集的mean和std

先将这个数据集的所有图片的路径写到一个文件中

import os
import re

dirs = os.listdir(r"F:\美食分类比赛\food_test\images")

# dress="F:/tensorflow4/src/train_data/"
with open(r"F:\美食分类比赛\train.txt","w") as f:
	for file_ in dirs:
	    for root,dirs,files in os.walk(os.path.join(dirs, file_)):
	        # root = root.replace(dress,'')
	        for file in files:
	            if re.search('.jpg', file):
	                f.write(os.path.join(root, file) + "\n")

然后再去这个文件读取图片的路径，并计算数据集的mean和std

**# -*- coding: utf-8 -*-**
import numpy as np
import cv2
import random
import os

**# calculate means and std  注意换行\n符号**
**# train.txt中每一行是图像的位置信息**
path = 'train.txt'
means = [0, 0, 0]
stdevs = [0, 0, 0]

index = 1
num_imgs = 0
with open(path, 'r') as f:
    lines = f.readlines()
    # random.shuffle(lines)
    print(lines)
    for line in lines:
        print(line)
        print('{}/{}'.format(index, len(lines)))
        index += 1
        a = os.path.join(line)
        # print(a[:-1])
        num_imgs += 1
        img = cv2.imread(a[:-1])
        img = np.asarray(img)
        print(img)
        img = img.astype(np.float32) / 255.
        for i in range(3):
            means[i] += img[:, :, i].mean()
            stdevs[i] += img[:, :, i].std()
print(num_imgs)
means.reverse()
stdevs.reverse()

means = np.asarray(means) / num_imgs
stdevs = np.asarray(stdevs) / num_imgs

print("normMean = {}".format(means))
print("normStd = {}".format(stdevs))
print('transforms.Normalize(normMean = {}, normStd = {})'.format(means, stdevs))