Python计算图片数据集的均值方差示例详解

前言

在做图像处理的时候，有时候需要得到整个数据集的均值方差数值，以下代码可以解决你的烦恼：

（做这个之前一定保证所有的图片都是统一尺寸，不然算出来不对，我的代码里设计的是512*512，可以自己调整，同一尺寸的代码我也有：

Python批量reshape图片

# -*- coding: utf-8 -*-
"""
Created on Thu Aug 23 16:06:35 2018
@author: libo
"""
from PIL import Image
import os
def image_resize(image_path, new_path):           # 统一图片尺寸
   print('============>>修改图片尺寸')
   for img_name in os.listdir(image_path):
       img_path = image_path + "/" + img_name    # 获取该图片全称
       image = Image.open(img_path)              # 打开特定一张图片
       image = image.resize((512, 512))          # 设置需要转换的图片大小
       # process the 1 channel image
       image.save(new_path + '/'+ img_name)
   print("end the processing!")
if __name__ == '__main__':
   print("ready for ::::::::  ")
   ori_path = r"Z:\pycharm_projects\ssd\VOC2007\JPEGImages"                # 输入图片的文件夹路径
   new_path = 'Z:/pycharm_projects/ssd/VOC2007/reshape'                   # resize之后的文件夹路径
   image_resize(ori_path, new_path)

import os
from PIL import Image
import matplotlib.pyplot as plt
import numpy as np
from scipy.misc import imread
filepath = r'Z:\pycharm_projects\ssd\VOC2007\reshape'  # 数据集目录
pathDir = os.listdir(filepath)
R_channel = 0
G_channel = 0
B_channel = 0
for idx in range(len(pathDir)):
   filename = pathDir[idx]
   img = imread(os.path.join(filepath, filename)) / 255.0
   R_channel = R_channel + np.sum(img[:, :, 0])
   G_channel = G_channel + np.sum(img[:, :, 1])
   B_channel = B_channel + np.sum(img[:, :, 2])
num = len(pathDir) * 512 * 512  # 这里（512,512）是每幅图片的大小，所有图片尺寸都一样
R_mean = R_channel / num
G_mean = G_channel / num
B_mean = B_channel / num
R_channel = 0
G_channel = 0
B_channel = 0
for idx in range(len(pathDir)):
   filename = pathDir[idx]
   img = imread(os.path.join(filepath, filename)) / 255.0
   R_channel = R_channel + np.sum((img[:, :, 0] - R_mean) ** 2)
   G_channel = G_channel + np.sum((img[:, :, 1] - G_mean) ** 2)
   B_channel = B_channel + np.sum((img[:, :, 2] - B_mean) ** 2)
R_var = np.sqrt(R_channel / num)
G_var = np.sqrt(G_channel / num)
B_var = np.sqrt(B_channel / num)
print("R_mean is ％f, G_mean is ％f, B_mean is ％f" ％ (R_mean, G_mean, B_mean))
print("R_var is ％f, G_var is ％f, B_var is ％f" ％ (R_var, G_var, B_var))

可能有点慢，慢慢等着就行。。。。。。。

最后得到的结果是介个

参考

计算数据集均值和方差

import os
from PIL import Image  
import matplotlib.pyplot as plt
import numpy as np
from scipy.misc import imread
filepath = ‘/home/JPEGImages‘ # 数据集目录
pathDir = os.listdir(filepath)
R_channel = 0
G_channel = 0
B_channel = 0
for idx in xrange(len(pathDir)):
   filename = pathDir[idx]
   img = imread(os.path.join(filepath, filename))
   R_channel = R_channel + np.sum(img[:,:,0])
   G_channel = G_channel + np.sum(img[:,:,1])
   B_channel = B_channel + np.sum(img[:,:,2])
num = len(pathDir) * 384 * 512 # 这里（384,512）是每幅图片的大小，所有图片尺寸都一样
R_mean = R_channel / num
G_mean = G_channel / num
B_mean = B_channel / num

R_channel = 0
G_channel = 0
B_channel = 0

for idx in xrange(len(pathDir)):
   filename = pathDir[idx]
   img = imread(os.path.join(filepath, filename))
   R_channel = R_channel + np.sum((img[:,:,0] - R_mean)**2)
   G_channel = G_channel + np.sum((img[:,:,1] - G_mean)**2)
   B_channel = B_channel + np.sum((img[:,:,2] - B_mean)**2)
R_var = R_channel / num
G_var = G_channel / num
B_var = B_channel / num
print("R_mean is ％f, G_mean is ％f, B_mean is ％f" ％ (R_mean, G_mean, B_mean))
print("R_var is ％f, G_var is ％f, B_var is ％f" ％ (R_var, G_var, B_var))

来源：https://blog.csdn.net/weixin_41765699/article/details/100118660