Pixel operation
One, arithmetic operations
1.1 Add, subtract, multiply and divide
Opencv's own picture color processing functions:
add: add()
, subtract: subtract(),
multiply: multiply()
, divide: divide(), the
principle is: get two (only two at a time) Zhang) The color value of the same position of a picture to realize the calculation.
Operational requirements: the shape of the two pictures must be the same.
Example image:
Code:
import cv2 as cv #导入cv模块
import numpy as np #np科学计数的包,通过numpy对数据进行处理
def add_demo(m1, m2):
dst = cv.add(m1, m2)
cv.imshow("add_demo", dst)
def subtract_demo(m1, m2):
dst = cv.subtract(m1, m2)
cv.imshow("subtract_demo", dst)
def multiply_demo(m1, m2):
dst = cv.multiply(m1, m2)
cv.imshow("multiply_demo", dst)
def divide_demo(m1, m2):
dst = cv.divide(m1, m2)
cv.imshow("divide_demo", dst)
src1 = cv.imread("D:/opencv3.4.13/opencv/sources/samples/data/LinuxLogo.jpg") #读取图片位置
src2 = cv.imread("D:/opencv3.4.13/opencv/sources/samples/data/WindowsLogo.jpg") #读取图片位置
print(src1.shape)
print(src2.shape)
cv.namedWindow("image1", cv.WINDOW_AUTOSIZE) #创建一个GUI
cv.imshow("image1", src1) #对窗口图片进行展示
cv.imshow("image2", src2) #对窗口图片进行展示
add_demo(src1, src2)
subtract_demo(src1, src2)
multiply_demo(src1, src2)
divide_demo(src1, src2)
cv.waitKey(0)
cv.destroyAllWindows() #释放所有的内存
The code can be simplified:
import cv2 as cv
#数值运算：加减乘除
def shu_image(src11, src22):
src = cv.add(src11, src22)#加
cv.imshow("add", src)
src = cv.subtract(src11, src22)#减
cv.imshow("subtract", src)
src = cv.subtract(src11, src22)#乘
cv.imshow("subtract", src)
src = cv.divide(src11, src22)#除
cv.imshow("divide", src)
src1 = cv.imread("01.jpg")
src2 = cv.imread("02.jpg")
cv.imshow("image1", src1)
cv.imshow("image2", src2)
shu_image(src1, src2)
cv.waitKey(0)
cv.destroyAllWindows()
Run screenshot:
1.2 Adjust brightness and adjust contrast
Basic principle: Two pictures are combined.
First, create a new picture with all zero pigments according to the original picture format, and then synthesize a new picture according to the different proportions of the two pictures. Mainly used function: addWeighted function
code show as below:
import cv2 as cv #导入cv模块
import numpy as np #np科学计数的包,通过numpy对数据进行处理
# 粗略的调节对比度和亮度
def contrast_brightness_image(src1, a, g):
h, w, ch = src1.shape # 获取shape的数值，height和width、通道
# 新建全零图片数组src2,将height和width，类型设置为原图片的通道类型(色素全为零，输出为全黑图片)
src2 = np.zeros([h, w, ch], src1.dtype)
dst = cv.addWeighted(src1, a, src2, 1 - a, g) # addWeighted函数说明如下
cv.imshow("con-bri-demo", dst)
src = cv.imread("C:/Users/lenovo/Desktop/opencv/daima/banknum/template-matching-ocr/images/lena.jpg")
cv.namedWindow("image1", cv.WINDOW_AUTOSIZE) #创建一个GUI
cv.imshow("image1", src) #对窗口图片进行展示
contrast_brightness_image(src, 1.5, 10)#第一个1.2为对比度 第二个为亮度数值越大越亮
cv.waitKey(0)
cv.destroyAllWindows() #释放所有的内存
Run screenshot:
addWeighted function: official: calculate the weight of two image arrays and my understanding is to synthesize two images according to the proportion.
Opencv: cv2.addWeighted() image fusion
addWeighted(InputArray src1, double alpha, InputArray src2, double beta, double gamma, OutputArray dst, int dtype=-1);
There are a total of seven parameters: the first 4 are the two pictures to be synthesized and their proportions, the fifth double gamma plays a fine-tuning role, the sixth OutputArray dst is the synthesized picture, and the seventh output picture type (Optional parameter, default -1)
There is a formula to get the picture output by adding two pictures: dst = src1[I] * alpha + src2[I] * beta + gamma
Reference: Opencv advanced study notes 3: Pixel operation and image brightness and contrast adjustment
The larger the c, the brighter.
What works is the coefficient 1, and the amount of brightness adjustment.
The coefficient 2 multiplied by the array of all 0s is useless, and it is used as a parameter here.
Two, logical operations
2.1 AND, OR, NOT
Opencv comes with image color processing functions:
and: bitwise_add()
or: bitwise_or()
non: bitwise_not()
XOR: bitwise_xor() The
code is as follows:
import cv2 as cv #导入cv模块
import numpy as np #np科学计数的包,通过numpy对数据进行处理
def logic_demo(m1, m2):
dst = cv.bitwise_and(m1, m2)
dst = cv.bitwise_or(m1, m2)
image = cv.imread("D:/opencv3.4.13/opencv/sources/samples/data/LinuxLogo.jpg") # 读取图片位
dst = cv.bitwise_not(image)
dst = cv.bitwise_xor(m1, m2)
cv.imshow("logic_demo", dst)
src1 = cv.imread("D:/opencv3.4.13/opencv/sources/samples/data/LinuxLogo.jpg") #读取图片位置
src2 = cv.imread("D:/opencv3.4.13/opencv/sources/samples/data/WindowsLogo.jpg") #读取图片位置
print(src1.shape)
print(src2.shape)
cv.namedWindow("image1", cv.WINDOW_AUTOSIZE) #创建一个GUI
cv.imshow("image1", src1) #对窗口图片进行展示
cv.imshow("image2", src2) #对窗口图片进行展示
logic_demo(src1, src2)
cv.waitKey(0)
cv.destroyAllWindows() #释放所有的内存
Run screenshot:
2.2 Mask layer control The
code is as follows:
import cv2 as cv #导入cv模块
import numpy as np #np科学计数的包,通过numpy对数据进行处理
def extrace_object_demo():
capture = cv.VideoCapture("C:/Users/lenovo/Desktop/opencv/daima/banknum/test.mp4")
while(True):
ret, frame = capture.read()
if ret == False:
break;
hsv = cv.cvtColor(frame, cv.COLOR_BGR2HSV)
lower_hsv = np.array([37, 43, 46])
upper_hsv = np.array([77, 255, 255])
mask = cv.inRange(hsv, lowerb=lower_hsv, upperb=upper_hsv)
dst = cv.bitwise_and(frame, frame, mask=mask)
cv.imshow("video", frame)
cv.imshow("mask", dst)
c = cv.waitKey(40)
if c == 27:
break
extrace_object_demo()
cv.waitKey(0)
cv.destroyAllWindows() #释放所有的内存
Run screenshot:
explain the meaning of the parameters in res = cv2.bitwise_and (img, img, mask = mask)
src1: the first image (the merged first object) src2: the second image (the merged second object) mask : Understand the rules to be merged. If the area of the image (scaled in grayscale and then masked) has black (value 0), it is not merged (the merged area of the first image and the merged area of the second image). Vice versa, it will be executed