1、图片的读取与展示
import cv2
img = cv2.imread('test.jpg',0) # 1:图片名称 2: 1彩色图片 0 灰度图片
cv2.imshow('image',img) # 1：展示窗口名称
cv2.waitKey(0)
2、图片的写入
#图片的写入
import cv2
img = cv2.imread('test.jpg',1)
# 1:要保存的文件名 2；保存文件的原始数据
#3: 保存JPG图片 有损压缩 0--100 ：数字越大，质量越高
# cv2.imwrite('copytest.jpg',img,[cv2.IMWRITE_JPEG_QUALITY,5])
#保存png图片 无损 透明操作 0--9 ：数字越大，质量越低
cv2.imwrite('copytest.jpg',img,[cv2.IMWRITE_PNG_COMPRESSION,0])
像素.png
像素的写入
import cv2
img = cv2.imread('test.jpg',1)
(b,g,r) = img[100, 100]#返回的是bgr的值
print(b,g,r)
for i in range(100):
img[10+i,100]=(255,0,0)
cv2.imshow('image',img)
cv2.waitKey(0)
tf的变量定义
import tensorflow as tf
#常量
data1 = tf.constant(10,dtype=tf.int32)
sess = tf.Session()
print(sess.run(data1))
#变量
data2 = tf.Variable(20,name='val')
init = tf.global_variables_initializer()
sess.run(init)
print(sess.run(data2))
tf 的四则运算
import tensorflow as tf
data1 = tf.constant(2)
data2 = tf.Variable(6)
dataAdd = tf.add(data1,data2)
datacopy = tf.assign(data2,dataAdd) # dataAdd-->data2
dataMui = tf.multiply(data1,data2)
dataSub = tf.subtract(data1,data2)
dataDvi = tf.divide(data1,data2)
with tf.Session() as sess:
init = tf.global_variables_initializer()
sess.run(init)
print(sess.run(dataAdd))
print(sess.run(dataMui))
print(sess.run(dataSub))
print(sess.run(dataDvi))
print(sess.run(datacopy)) # 8
print(datacopy.eval()) #10
print(tf.get_default_session().run(datacopy)) #12
placeholder的使用
import tensorflow as tf
data1 = tf.placeholder(tf.float32)
data2 = tf.placeholder(tf.float32)
dataAdd = tf.add(data1,data2)
with tf.Session() as sess:
print(sess.run(dataAdd,feed_dict={data1:2,data2:6}))
矩阵的定义和读取操作
import tensorflow as tf
data = tf.constant([[1,2],
[3,4],
[5,6]])
with tf.Session() as sess:
print(data.shape)
print(sess.run(data))
print(sess.run(data[0]))#打印第一行
print(sess.run(data[:,0]))#答应第一列
矩阵的加法和除法
import tensorflow as tf
data1 = tf.constant([[1,2]])
data2 = tf.constant([[3],
[4]])
datamul = tf.matmul(data1,data2)#矩阵乘法
datamulti = tf.multiply(data1,data2)#数乘
dataadd = tf.add(data1,data1)
with tf.Session() as sess:
print(sess.run(datamul))
print(sess.run(dataadd))
print(sess.run(datamulti))
print(sess.run([datamul,dataadd,datamulti]))
特殊矩阵
import tensorflow as tf
mat0 = tf.constant([[1, 2,3]])
mat1 = tf.zeros([3,2]) #3行2列的全0 矩阵
mat2 = tf.ones([3,3]) #3行3列的全1 矩阵
mat3 = tf.fill([2,1],15) #2行1列的全15 矩阵
mat4 = tf.linspace(1.0,2.0,11)#将1 到2 10等分
mat5 = tf.random_uniform([3,2],1,100)#3行2列的矩阵 1--100的水机数
mat6 = tf.zeros_like(mat0) #和mat0一样的形状的全0矩阵
with tf.Session() as sess:
print(sess.run(mat6))
malplotlib的使用
import numpy as np
import matplotlib.pyplot as plt
x = np.array([1,2,3,4,5,6])
y = np.array([4,5,7,4,9,1])
plt.plot(x,y,color='blue',lw=10)
x = np.array([1,2,3,4,5,6])
y = np.array([4,5,7,4,9,1])
plt.bar(x,y,0.5,color='red',alpha=1,)#0.5表示条形图填充的面积百分比
plt.show()
股票逼近案例
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
day = np.linspace(1,7,7)
open = np.array([2673.07,2723.89,2691.43,2777.25,2780.74,2769.02,2791.40])
close = np.array([2698.47,2668.97,2705.19,2723.26,2780.96,2785.87,2795.31])
for i in range(0,7):
onday = np.zeros([2])
onday[0] = i+1
onday[1] = i+1
price = np.zeros([2])
price[0] = open[i]
price[1] = close[i]
if open[i]图片的缩放
import cv2
img = cv2.imread('test.jpg',1)
shape = img.shape
print(shape)
height = shape[0]
width = shape[1]
dstheight = int(height*0.5)
dstwidth = int(width*0.5)
dst = cv2.resize(img,(dstwidth,dstheight))
cv2.imshow('resize',dst)
cv2.waitKey(0)
图片缩放的源代码实现
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
dstheight = int(height/2)
dstwidth = int(width/2)
dst = np.zeros((dstheight,dstwidth,3),np.uint8)#0-255
for i in range(0,dstheight):
for j in range(0,dstwidth):
iNew = int(i*(width/dstwidth))
jNew = int(j*(height/dstheight))
dst[i,j] = img[iNew,jNew]
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的剪切
import cv2
img = cv2.imread('test.jpg',1)
print(img.shape)
dst = img[100:300,200:400]
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的移位
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
cv2.imshow('original',img)
shape = img.shape
height = shape[0]
width = shape[1]
matShift = np.float32([[1,0,100],[0,1,200]])
dst = cv2.warpAffine(img,matShift,(height,width))
cv2.imshow('dst',dst)
cv2.waitKey(0)
移位的源码实现
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
shape = img.shape
dst = np.zeros(img.shape,np.uint8)
height = shape[0]
width = shape[1]
for i in range(height):
for j in range(width-100):
dst[i,j+100] = img[i,j]
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的翻转
import cv2
import numpy as np
img = cv2.imread('test.png',1)
shape = img.shape
height = shape[0]
width = shape[1]
deep = shape[2]
newImg = (height*2,width,deep)
dst = np.zeros(newImg,np.uint8)
for i in range(0,height):
for j in range(0,width):
dst[i,j] = img[i][j]
dst[2*height-i-1,j] = img[i,j]
for i in range(0,width):
dst[height,i] = (0,0,255)
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的放缩
import numpy as np
import cv2
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
maxScale = np.float32([[0.5,0,0],[0,0.5,0]])
dst = cv2.warpAffine(img,maxScale,(int(width/2),int(height/2)))
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的边缘检测
import cv2
import numpy as np
img = cv2.imread('test.jpg',0)
# gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
imgG = cv2.GaussianBlur(img,(3,3),0)
dst = cv2.Canny(img,50,50)
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的旋转
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
#旋转矩阵
# 1 旋转中心 2、旋转角度 3、图片缩放
matRoate = cv2.getRotationMatrix2D((width*0.5,height*0.5),45,0.5)
dst = cv2.warpAffine(img,matRoate,(width,height))
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的灰度处理
'''方法1
import cv2
img = cv2.imread('test.jpg',1)
dst = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
cv2.imshow('dst',dst)
cv2.waitKey(0)
'''
#方法2 心理学计算公式 gray = r*0.299 + g*0.587 +b*0.114
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
dst = np.zeros((height,width,3),np.uint8)
for i in range(0,height):
for j in range(0,width):
(b,g,r) = img[i,j]
b = int (b)
g = int (g)
r = int (r)
#通过移位熟读更快
#gray = (b+(g<<1)+r)>>2
gray = r*0.299 + g*0.587 +b*0.114
dst[i,j]=gray
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的马赛克效果
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
for i in range(100,1000):
for j in range(100,700):
if i%10==0 and j%10==0:
(b, g, r) = img[i, j]
#填充10*10的小方块
for m in range(1,10):
for n in range(1,10):
img[i+m,n+j]=(b,g,r)
cv2.imshow('dst',img)
cv2.waitKey(0)
毛玻璃效果
import cv2
import numpy as np
import random
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
dst = np.zeros((height,width,3),np.uint8)
mm = 8 #上下8像素随机距离
for i in range(height-mm):
for j in range(width-mm):
index = int(random.random()*8)#随机生成0-8
(b,g,r)=img[i+index,j+index]
dst[i,j]=(b,g,r)
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的融合
import cv2
import numpy as np
img1 = cv2.imread('test.jpg',1)
img2 = cv2.imread('chenw.jpg',1)
t1 = img1[0:500,0:700]
t2 = img2[0:500,0:700]
dst = np.zeros((500,700,3),np.uint8)
# dst = t1*0.5+t2*0.5
dst = cv2.addWeighted(t1,0.5,t2,0.5,0)
cv2.imshow('dst',dst)
cv2.waitKey(0)
边缘检测
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
#灰度处理
img = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
#高斯滤波
imgG = cv2.GaussianBlur(img,(3,3),0)
#Canny处理 大于50认为为结果点
dst = cv2.Canny(imgG,50,50)
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片卷积源码实现
import cv2
import numpy as np
import math
img = cv2.imread('test.jpg',0)
img = cv2.resize(img,(400,300))
shape = img.shape
height = shape[0]
width = shape[1]
dst = np.zeros((height,width,1),np.uint8)
for i in range(0,height-2):
for j in range(0,width-2):
'''
[1,2,1
0.,0,
-1,-2,-1
]
[
1,0,-1
2,0,-2
1,0,-1
]
'''
gy = img[i,j]*1+img[i,j+1]*2+img[i,j+2]*1-(img[i+2,j]*1+img[i+2,j+1]*2+img[i+2,j+2]*1)
gx = img[i,j]*1+img[i+1,j]+2+img[i+2,j]*1-(img[i,j+2]*1+img[i+1,j+2]+2+img[i+2,j+2]*1)
gr = math.sqrt(gx*gx+gy*gy)
if gr>50:
dst[i,j] = 255
else :
dst[i,j] = 0
cv2.imshow('dst',dst)
cv2.waitKey(0)
图片的浮雕效果
import cv2
import numpy as np
import math
img = cv2.imread('test.jpg',0)
shape = img.shape
height = shape[0]
width = shape[1]
dst = np.zeros((height,width,1),np.uint8)
for i in range(0,height):
for j in range(0,width-1):
gray0 = int(img[i,j])
gray1 = int(img[i,j+1])
newGrey = gray0-gray1+150
if newGrey>255:
newGrey=255
elif newGrey<0:
newGrey=0
dst[i,j]=newGrey
cv2.imshow('dst',dst)
cv2.waitKey(0)
线段的绘制
import cv2
import numpy as np
imginfo = (500,500,3)
img = np.zeros(imginfo,np.uint8)
# begin end cololr line_width line_typr
cv2.line(img,(200,200),(400,400),(0,255,255),20,cv2.LINE_AA)
cv2.imshow('img',img)
cv2.waitKey(0)
基本图形的绘制
import cv2
import numpy as np
imgInfo = (500,400,3)
dst = np.zeros(imgInfo,np.uint8)
# 2:左上角 3:右下角 4：颜色 5:-1表示填充 大于0表示线条宽度
cv2.rectangle(dst,(100,100),(200,200),(255,0,0),-1)
#
cv2.circle(dst,(50,50),(50),(0,255,0),-1)
#椭圆 0-360度
cv2.ellipse(dst,(300,300),(100,50),0,0,360,(0,0,255),-1)
cv2.imshow('dst',dst)
cv2.waitKey(0)
文字的写入
'''
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
font = cv2.FONT_HERSHEY_PLAIN
cv2.putText(img,'this a girl',(20,200),font,2,(0,255,20),3,cv2.LINE_AA)
cv2.imshow('img',img)
cv2.waitKey(0)
'''
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
height = int(img.shape[0]*0.2)
width = int(img.shape[1]*0.2)
tem = cv2.resize(img,(width,height))
for i in range(0,height):
for j in range(0,width):
img[i+200,j+200] = tem[i,j]
cv2.imshow('img',img)
cv2.waitKey(0)
彩色图片直方图
import cv2
import numpy as np
def getHist(img,type):
color = (255,255,255)
if type == 33:
color = (255,0,0)
elif type ==34:
color = (0,255,0)
elif type == 35:
color = (0,0,255)
#注意全部中括号 2：灰度 3:mask 4:范围 5：颜色范围
hist = cv2.calcHist([img],[0],None,[255],[0.0,255.0])
maV,maxV,minL,maxL = cv2.minMaxLoc(hist)
histIng = np.zeros([255,255,3],np.uint8)
for n in range(0,255):
interNormal = int(hist[n]*256/maxV)
cv2.line(histIng,(n,256),(n,256-interNormal),color)
cv2.imshow('name',histIng)
cv2.waitKey(0)
img = cv2.imread('test.jpg',1)
channels = cv2.split(img)
for i in range(0,3):
getHist(channels[i],33+i)
直方图均衡化
# import cv2
# img = cv2.imread('test.jpg',0)
# cv2.imshow('img',img)
# #dst = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
# dst = cv2.equalizeHist(img)
# cv2.imshow('dst',dst)
# cv2.waitKey(0)
# import cv2
# img = cv2.imread('test.jpg',1)
# (b,g,r) = cv2.split(img)
# bH = cv2.equalizeHist(b)
# gH = cv2.equalizeHist(g)
# rH = cv2.equalizeHist(r)
# res = cv2.merge((bH,gH,rH))
# cv2.imshow('res',res)
# cv2.waitKey(0)
import cv2
img = cv2.imread('test.jpg',1)
tem = cv2.cvtColor(img,cv2.COLOR_BGR2YCrCb)
channelYUV = cv2.split(tem)
channelYUV[0] = cv2.equalizeHist(channelYUV[0])
channel = cv2.merge(channelYUV)
res = cv2.cvtColor(channel,cv2.COLOR_YCR_CB2BGR)
cv2.imshow('res',res)
cv2.waitKey(0)
图片的修补
import cv2
import numpy as np
img = cv2.imread('bandgirl.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
paint = np.zeros((height,width,1),np.uint8)
for i in range(200,300):
paint[i,250] = 255
paint[i,250+1] = 255
paint[i,250-1] = 255
for i in range(200,300):
paint[250,i] = 255
paint[250+1,i] = 255
paint[250-1,i] = 255
# 坏图片 蒙版数组
dst = cv2.inpaint(img,paint,3,cv2.INPAINT_TELEA)
cv2.imshow('dst',dst)
cv2.waitKey(0)
双边滤波
import cv2
img = cv2.imread('test.jpg',1)
dst = cv2.bilateralFilter(img,15,50,50)
cv2.imshow('img',img)
cv2.imshow('dst',dst)
cv2.waitKey(0)
中值滤波：对6*6的范围求均值
import cv2
import numpy as np
img = cv2.imread('test.jpg',1)
shape = img.shape
height = shape[0]
width = shape[1]
dst = np.zeros((height,width,3),np.uint8)
for i in range(3,height-3):
for j in range(3,width-3):
sum_b = int(0)
sum_g = int(0)
sum_r = int(0)
for m in range(-3,3):
for n in range(-3,3):
(b,g,r) = img[i+m,j+n]
sum_b = sum_b+int(b)
sum_g = sum_g+int(g)
sum_r = sum_r+int(r)
b = np.uint8(sum_b/36)
g = np.uint8(sum_g/36)
r = np.uint8(sum_r/36)
dst[i,j] = (b,g,r)
cv2.imshow('dst',dst)
cv2.waitKey(0)
视频截取图片
import cv2
cap = cv2.VideoCapture('test.mp4')
isOpen = cap.isOpened
fps = cap.get(cv2.CAP_PROP_FPS)#帧数
heigth = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
print(fps,heigth,width)
i = 0
while(isOpen):
imgName = 'img'+str(i)+'.jpg'
(flag,frame) =cap.read()#读取图片
if flag==True:
cv2.imwrite(imgName,frame)
i=i+1
if i==10:
break
图片合成视频
import cv2
img = cv2.imread('img0.jpg')
size = (img.shape[1],img.shape[0])
print(size)
videoWriter = cv2.VideoWriter('chini.avi',-1,15,size)
for i in range(1,10):
fname = 'img'+str(i)+'.jpg'
img=cv2.imread(fname)
videoWriter.write(img)
print('end!')
人脸识别
import cv2
img = cv2.imread('test.jpg')
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
face_xml = cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
eye_xml = cv2.CascadeClassifier('haarcascade_eye.xml')
#检测人脸 缩放量 脸的最小尺寸>5
face = face_xml.detectMultiScale(gray,1.1,5)
print('face',len(face))
for (x,y,w,h) in face:
cv2.rectangle(img,(x,y),(y+h,x+w),(255,0,0),2)
eye_gray = gray[y:y+h,x:x+w]
eye_color = img[y:y+h,x:x+w]
eyes = eye_xml.detectMultiScale(eye_gray,1.1,5)
print('eyes',len(eyes))
for (e_x,e_y,e_w,e_h) in eyes:
cv2.rectangle(eye_color, (e_x, e_y), (e_y + e_h, e_x + e_w), (0, 255, 0), 2)
cv2.imshow('face',img)
cv2.waitKey(0)
print('end!')
SVM
import cv2
import numpy as np
import matplotlib.pyplot as plt
ran1 = np.array([[150,50],[153,49],[160,55],[165,58],[170,60]])
ran2 = np.array([[152,60],[160,63],[165,68],[170,69],[180,75]])
label = np.array([[0],[0],[0],[0],[0],[1],[1],[1],[1],[1]])
ran = np.vstack((ran1,ran2))
ran = np.array(ran,dtype='float32')
print(ran)
#训练
svm = cv2.ml.SVM_create()
#属性
svm.setType(cv2.ml.SVM_C_SVC)
svm.setKernel(cv2.ml.SVM_LINEAR)#线性分类
svm.setC(0.01)
result = svm.train(ran,cv2.ml.ROW_SAMPLE,label)
#预测数据
pr_data = np.vstack([[170,69],[150,49]])
pr_data = np.array(pr_data,dtype='float32')
(res1,res2) = svm.predict(pr_data)
print(res2)
hog特征
hog1
image.png