阈值与平滑处理⑤

阈值与平滑处理

阈值处理

阈值处理是指提出图像内像素值高于一定值或者低于一定值的像素点。在opencv中提供了函数cv2.threshold()和函数cv2.adaptiveThreshold(),用于实现阈值处理。

ret, dst = cv2.threshold(src, thresh, maxval, type)

src： 输入图，只能输入单通道图像，通常来说为灰度图
dst： 输出图
thresh： 阈值
maxval： 当像素值超过了阈值（或者小于阈值，根据type来决定），所赋予的值

type：二值化操作的类型，包含以下5种类型： cv2.THRESH_BINARY； cv2.THRESH_BINARY_INV； cv2.THRESH_TRUNC； cv2.THRESH_TOZERO；cv2.THRESH_TOZERO_INV

cv2.THRESH_BINARY 超过阈值部分取maxval（最大值），否则取0
cv2.THRESH_BINARY_INV THRESH_BINARY的反转
cv2.THRESH_TRUNC 大于阈值部分设为阈值，否则不变
cv2.THRESH_TOZERO 大于阈值部分不改变，否则设为0
cv2.THRESH_TOZERO_INV THRESH_TOZERO的反转

import cv2
import numpy as np
peppa = cv2.imread('peppa.jpg')
img=cv2.cvtColor(peppa,cv2.COLOR_BGR2GRAY)
cv2.imshow('Peppa',img)
ret,thresh1 = cv2.threshold(img,200,255,cv2.THRESH_BINARY)
ret,thresh2 = cv2.threshold(img,200,255,cv2.THRESH_BINARY_INV)
ret,thresh3 = cv2.threshold(img,200,255,cv2.THRESH_TRUNC)
ret,thresh4 = cv2.threshold(img,200,255,cv2.THRESH_TOZERO)
ret,thresh5 = cv2.threshold(img,200,255,cv2.THRESH_TOZERO_INV)
cv2.imshow('BINARY',thresh1)
cv2.imshow('BINARY_INV',thresh2)
# cv2.imshow('TRUNC',thresh3)
# cv2.imshow('TOZERO',thresh4)
# cv2.imshow('TOZERO_INV',thresh5)
peppa_body=cv2.bitwise_and(peppa,peppa,mask=thresh2)
cv2.imshow('peppa_body',peppa_body)
cv2.waitKey()
cv2.destroyAllWindows()

import cv2
Type=0  #阈值处理类型值
Value=0 #使用的阈值
def onType(a):
    Type= cv2.getTrackbarPos(tType, windowName)
    Value= cv2.getTrackbarPos(tValue, windowName)
    ret, dst = cv2.threshold(img, Value,255, Type) 
    cv2.imshow(windowName,dst)
    
def onValue(a):
    Type= cv2.getTrackbarPos(tType, windowName)
    Value= cv2.getTrackbarPos(tValue, windowName)
    ret, dst = cv2.threshold(img, Value, 255, Type) 
    cv2.imshow(windowName,dst)

img = cv2.imread("peppa.jpg",0)
windowName = "Peppa"  #窗体名
cv2.namedWindow(windowName)
cv2.imshow(windowName,img)
#创建两个滑动条
tType = "Type"  #用来选取阈值处理类型的滚动条
tValue = "Value"    #用来选取阈值的滚动条
cv2.createTrackbar(tType, windowName, 0, 4, onType)
cv2.createTrackbar(tValue, windowName,0, 255, onValue) 
cv2.waitKey()
cv2.destroyAllWindows()

自适应阈值

cv2.adaptiveThreshold()

img=cv2.imread('peppa.jpg',0)
athdMEAN=cv2.adaptiveThreshold(img,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,7,5)
athdGAUS=cv2.adaptiveThreshold(img,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,5,3)
cv2.imshow("athMEAN",athdMEAN)
cv2.imshow("athGAUS",athdGAUS)
cv2.waitKey(0)
cv2.destroyAllWindows()

平滑处理

在尽量保留图像原有信息的情况下，过滤掉图像内部的噪声，这一过程称为对图像的平滑处理，所得的图像称为平滑图像。通过平滑处理，可以有效地过滤掉图像内的噪声信息。

模糊处理

import cv2
import numpy as np
img = cv2.imread("peppa_gaussian.jpg")
blur = cv2.blur(img, (7, 7))
box = cv2.boxFilter(img,-1,(7,7), normalize=True)  
gaussian = cv2.GaussianBlur(img, (7, 7), 10) 
median = cv2.medianBlur(img, 7)
bilater=cv2.bilateralFilter(img,9,75,75)
kernel = np.array((
        [-2, -1, 0],
        [-1,1,1],
        [0, 1, 2]), dtype="float32")
filter2D=cv2.filter2D(img,-1,kernel)#https://my.oschina.net/u/4306156/blog/3598055
cv2.imshow('img',img)
cv2.imshow('blur',blur)
cv2.imshow('box',box)
cv2.imshow('gaussian',gaussian)
cv2.imshow('median',median)
cv2.imshow('bilater',bilater)
cv2.imshow('filter2D',filter2D)
cv2.waitKey()
cv2.destroyAllWindows()

拓展：中值滤波的滑块调整

import cv2
Value=0 #使用的阈值
def onValue(a):
    Value= cv2.getTrackbarPos(tValue, windowName)
    median = cv2.medianBlur(img, 2*Value+1) 
    cv2.imshow(windowName,median)
    
img = cv2.imread("peppa.jpg",0)
windowName = "Peppa"  
cv2.namedWindow(windowName)
cv2.imshow(windowName,img)
tValue = "Value"
v=cv2.createTrackbar(tValue, windowName,1, 100, onValue)
cv2.waitKey()
cv2.destroyAllWindows()

© Fu Xianjun. All Rights Reserved.