学习时，将每块注释分别去掉即为一个模块知识点

图像数字化

import numpy as np
"""
#二维的ndarry
#两行四列全是0的uchar类型的二维数组
z = np.zeros((2,4),np.uint8)
print(type(z))
print(z)
print("\n")

#两行四列全是1的整型矩阵
o = np.ones((2,4),np.int32)
print(o)
print("\n")

#浮点型矩阵
m = np.array([[4,12,3,1],[10,12,14,29]],np.float32)
print(m)
print("\n")
"""

"""
#二维的ndarry
#2×2×4的32为浮点型数组
m = np.array([
        [[1,2,3,4],[5,6,7,8]],
        [[10,11,12,14],[15,16,17,18]]
        ],np.float32)
print(m)
"""

"""
#ndarry成员变量
m = np.array([[4,12,3,1],[10,12,14,29]],np.float32)
#m的尺寸
print(m.shape)
#m的数据类型
print(m.dtype)
"""

"""
#访问ndarry中的值
#访问二维ndarry中的值
m = np.array([[14,12,3,1],[10,12,114,29],[67,23,534,2]],np.float32)
#从第0行第0列开始，：表示整个，左开右闭
print(m)
print(m[1,3])#第1行第3列
print(m[2,:])#整个第2行
print(m[:,3])#整个第3列
print(m[0:2,1:3])#第0行到第2行 ∩ 第1列和第3列的
"""

"""
#访问三维ndarry中的值
m = np.array([
        [[1,2,3,4],[5,6,7,8]],
        [[10,11,12,14],[15,16,17,18]],
        [[11,12,43,32],[1,5,10,23]]
        ],dtype=float)
print(m)
#所有二维数组的第0列
print(m[:,:,0])
#第0个二维数组
print(m[0,:,:])
"""

'''
Mat(Size(int cols,int rows),int type)
其中cols列（宽）
rows行（高）
type类型，包括通道数及其数据类型
    CV_8UC(n)
    CV_16SC(n)
    CV_16UC(n)
    CV_32SC(n)
    CV_32FC(n)
    CV_64FC(n)
8U、8S、16S、16U、32S、32F、64F代表一个数值所占的bit数，1byte=8bit，都应该除以8再看
1U、1S、2S、2U、4S、4F、8F
uchar、S为int、32F为4字节的float，64F为8字节的double
C(n)表示通道数，n=1即二维矩阵，n=2即三维矩阵...
'''

"""
#ndarray的加法
src1 = np.array([[23,123,90],[100,250,0]],np.uint8)
src2 = np.array([[125,150,60],[100,10,40]],np.uint8)
print(src1)
print(src2)
dst = src1 + src2
print(dst)
print(dst.dtype)
#仔细看123+150=273，最后的结果为273%255-1=17
"""
'''
src1 = np.array([[23,123,90],[100,250,0]],np.uint8)
src2 = np.array([[125,150,60],[100,10,40]],np.uint8)
dst = cv2.add(src1,src2,dtype=cv2.CV_32F)
print(dst)
'''

"""
#ndarry的减法
src1 = np.array([[23,123,90],[100,250,0]],np.uint8)
src2 = np.array([[125,150,60],[100,10,40]],np.uint8)
dst = src1 - src2
print(dst)
#23-125=-102 -102%255+1=154
"""

"""
#ndarray的点乘
src1 = np.array([[23,123,90],[100,250,0]],np.uint8)
src2 = np.array([[125,150,60],[100,10,40]],np.uint8)
dst = src1 * src2
print(dst)
dst1 =np.multiply(src1,src2)
print(dst1)
"""

"""
#ndarray的点除
src1 = np.array([[23,123,90],[100,250,0]],np.uint8)
src2 = np.array([[125,150,60],[100,10,40]],np.uint8)
src3 = np.array([[23,123,90],[100,250,0]],np.float32)
dst = src2 / src1
dst1 = src2 / src3
print(dst)
print(dst1)
#Numpy在处理分母为0的时候，如果两个ndarray都是uint8类型则返回0，其他情况返回inf
"""

"""
#ndarray的乘法
src3 = np.array([[1,2,3],[4,5,6]],np.uint8)
src4 = np.array([[6,5],[4,3],[2,1]],np.uint8)
dst = np.dot(src3,src4)
print(dst)
"""

"""
#ndarray的指数和对数运算
src5 = np.array([[6,5],[4,3]],np.uint8)
dst2 = np.log(src5)
dst3 = np.exp(src5)
print(dst2)
print(dst2.dtype)
print(dst3)
print(dst3.dtype)
#log对数，exp指数，矩阵类型只能是CV_32F或者CF_64F
"""

"""
#ndarray的幂指数和开平方运算
src = np.array([[25,40],[10,100]],np.uint8)
dst1 = np.power(src,2)
print(dst1)
print(dst1.dtype)
dst2 = np.power(src,2.0)
print(dst2)
print(dst2.dtype)
#power中指数是2和2.0的时候返回值会不同，精度不同，这点要注意
"""

例子

加载一个灰度图(E:\Python-workspace\yanyu.png)，显示图片
按下’s’键保存(beyond.png)(保存后的路径和该程序所在路径一致)后退出，或者按下 ESC 键退出不保存

import cv2

img = cv2.imread('E:\Python-workspace/yanyu.png',0)#读取该灰度照片0，彩色照片1
cv2.imshow("image",img)#显示该照片
k = cv2.waitKey(0)#获取所按的键盘反馈
if k == 27:#是Esc
    cv2.destoryAllWindows()#退出
elif k == ord('s'):#是's'
    cv2.imwrite("beyond.png",img)#保存该照片
    cv2.destroyAllWindows()#退出

对照片(E:\Python-workspace\OpenCV/water1.png)进行边缘检测，并将边缘化的照片进行保存

import cv2
import numpy as np
from matplotlib import pyplot as plt
img = cv2.imread('E:\Python-workspace\OpenCV/water1.png',1)
edges = cv2.Canny(img,100,200)
plt.subplot(121),plt.imshow(img,cmap = 'gray')
plt.title('Original Image'), plt.xticks([]), plt.yticks([])
plt.subplot(122),plt.imshow(edges,cmap = 'gray')
plt.title('Edge Image'), plt.xticks([]), plt.yticks([])
plt.show()
cv2.imwrite("beyond.png",edges)#保存照片
cv2.destroyAllWindows()#退出

找取图片上的25个特征点

import numpy as np
import cv2
from matplotlib import pyplot as plt
img = cv2.imread('water1.png')
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
corners = cv2.goodFeaturesToTrack(gray,25,0.01,10) # 返回的结果是 [[ 311., 250.]] 两层括号的数组。
corners = np.int0(corners)
for i in corners:
    x,y = i.ravel()
    cv2.circle(img,(x,y),3,255,-1)
plt.imshow(img),plt.show()

显示照片的直方图

import cv2
import numpy as np
from matplotlib import pyplot as plt
img = cv2.imread('water2.png',0)
plt.hist(img.ravel(),256,[0,256]);
plt.show()

将两照片进行加权混合对比(1.png和2.jpg的大小完全相同才行)

import cv2
import numpy as np
img1=cv2.imread('E:\Python-workspace\OpenCV/1.png',0)
img2=cv2.imread('E:\Python-workspace\OpenCV/2.jpg',0)
Image=cv2.addWeighted(img1,0.7,img2,0.3,0) #img1的比重因子为0.7，img2的比重因子0.3，伽马为0
cv2.imshow('Image',Image)
cv2.waitKey(0)
cv2.destroyAllWindow()

比较俩照片的差异，返回数值

from PIL import Image
import math
import operator
from functools import reduce

def image_contrast(img1, img2):
    image1 = Image.open(img1)
    image2 = Image.open(img2)
    h1 = image1.histogram()
    h2 = image2.histogram()
    result = math.sqrt(reduce(operator.add,  list(map(lambda a,b: (a-b)**2, h1, h2)))/len(h1) )
    return result
if __name__ == '__main__':
    img1 = "E:\Python-workspace\OpenCV/water1.png"  # 指定图片路径
    img2 = "E:\Python-workspace\OpenCV/water1.png"
    result = image_contrast(img1,img2)
    print(result)
    if result==0:print("OK")
    else: print("NO")