帧间差分、对比度--评价镜头的防抖性。EIS

import matplotlib.pyplot as plt
import cv2
import numpy as np
import os
import pandas as pd
import time
start = time.process_time()
inputs1 = './inputs1'
inputs2 = './inputs2'
path0 = './txt-out' 
if not os.path.exists(path0):
    os.mkdir(path0)
path1 = './Last-Result' 
if not os.path.exists(path1):
    os.mkdir(path1)
f3 = open('./txt-out/absdiff.txt','w+')
f4 = open('./txt-out/resolution.txt','w+')
filepath1 = './inputs1'
filepath2 = './inputs2'
pathDir1 = os.listdir(filepath1) 
pathDir1.sort(key=lambda x: int(x.split('-')[0]))  
pathDir2 = os.listdir(filepath2)
pathDir2.sort(key=lambda x: int(x.split('-')[0]))  
for allDir1,allDir2 in zip(pathDir1,pathDir2):
    picpath1 = r'./inputs1/' + allDir1
    picpath2 = r'./inputs2/' + allDir2
    img1 = cv2.imdecode(np.fromfile(picpath1, dtype=np.uint8), 1)
    img2 = cv2.imdecode(np.fromfile(picpath2, dtype=np.uint8), 1)
    H1 = cv2.calcHist([img1], [0], None, [256],[0,256])
    H1 = cv2.normalize(H1, H1, 0, 1, cv2.NORM_MINMAX, -1) 
    H2 = cv2.calcHist([img2], [0], None, [256],[0,256])
    H2 = cv2.normalize(H2, H2, 0, 1, cv2.NORM_MINMAX, -1)
    a_gray = cv2.cvtColor(img1,cv2.COLOR_BGR2GRAY)
    b_gray = cv2.cvtColor(img2,cv2.COLOR_BGR2GRAY)
    a_Gauss = cv2.GaussianBlur(a_gray,(3,3),0)
    b_Gauss = cv2.GaussianBlur(b_gray,(3,3),0)
    diff = cv2.absdiff(a_Gauss,b_Gauss).mean()
    resolution = cv2.Laplacian(a_gray, cv2.CV_64F).var()
    '''# sobel算子
    sobelx = cv2.Sobel(a_gray,cv2.CV_64F,1,0,ksize=5)
    sobely = cv2.Sobel(a_gray,cv2.CV_64F,0,1,ksize=5)
    absX = cv2.convertScaleAbs(sobelx)   # 转回uint8
    absY = cv2.convertScaleAbs(sobely)
    '''
    print('当前正在计算：'+allDir1+' <———VS———> '+allDir2+'	'+'请稍等………')
    print('清晰度：',resolution)
    print('帧间差分：',diff,'
')
    print(allDir1 + "--Vs--"+allDir2+':',diff,file = f3)
    print(allDir1 + ':',resolution,file = f4)
f3.close()
f4.close()
print('
','恭喜，计算完成！')
file3 = open('.\txt-out\absdiff.txt', 'r')  
df3 = file3.read().split('
')
df3 = pd.DataFrame(df3,columns=['Compare-Name'])  
dt3 = df3['Compare-Name'].str.split(':',expand=True) 
df3[['Compare-Name','absdiff']]=dt3  # 设置列名
df3['absdiff'] = df3['absdiff'].astype(np.float) 
df3.to_excel('./Last-Result/'+time.strftime('%Y%m%d%H%M%S'+'-',time.localtime(time.time()))+'帧间差分.xls','帧间差分',index = False)  
print('帧间差分结果存储完成,请移步----→ Last-Result（ 帧间差分.xls ）中查看结果。')
file4 = open('.\txt-out\resolution.txt', 'r')  
df4 = file4.read().split('
')
df4 = pd.DataFrame(df4,columns=['Compare-Name'])  
dt4 = df4['Compare-Name'].str.split(':',expand=True) 
df4[['Compare-Name','resolution']]=dt4  # 设置列名
df4['resolution'] = df4['resolution'].astype(np.float) 
df4.to_excel('./Last-Result/'+time.strftime('%Y%m%d%H%M%S'+'-',time.localtime(time.time()))+'清晰度.xls','清晰度-Laplacian',index = False)  
print('清晰度结果存储完成,请移步----→ Last-Result（ 清晰度.xls ）中查看结果。')
end = time.process_time()
print('
','程序运行耗费时间: %s Seconds'%(end-start))

'''
# 两张图对比
a = cv2.imread('./m.jpg')
b = cv2.imread('./n.jpg')
H1 = cv2.calcHist([a], [0], None, [256],[0,256])
H1 = cv2.normalize(H1, H1, 0, 1, cv2.NORM_MINMAX, -1) # 对图片进行归一化处理
H2 = cv2.calcHist([b], [0], None, [256],[0,256])
H2 = cv2.normalize(H2, H2, 0, 1, cv2.NORM_MINMAX, -1)
similarity0 = cv2.compareHist(H1, H2,cv2.HISTCMP_CHISQR)  # 卡方检验
similarity1 = cv2.compareHist(H1, H2,cv2.HISTCMP_CORREL)  # 相关性
similarity2 = cv2.compareHist(H1, H2,cv2.HISTCMP_BHATTACHARYYA)  # 巴氏距离
print('chi:',similarity0)
print('correl:',similarity1)
print('dis:',similarity2)
'''