# 图像二值化binary(将图像的每个像素变成两种值,如0,255)
分为两种API:
全局二值化:
threshold(img,thresh,maxVal,type)
img:图像,最好拿灰度图进行二值化
thresh:阈值
maxVal:超过阈值,则替换成maxVal
type:THRESH_BINARY和THRESH_BINARY_INV,这两个api相反效果,第一个是超过阈值变成最大值,第二个是低于
局部二值化adaptiveThreshold API:
adaptiveThreshold(img,maxVal,adaptiveMethod,
type,blockSize,C)
adaptiveMethod(适应方法):
计算阈值的方法:ADAPTIVE_THRESH_MEAN_C:计算邻近区域的平均值
ADAPTIVE_THRESH_GAUSSIAN_C:高斯窗口甲醛平均值,一般选择这个
type:THRESHOLD_BINARY、THRESHOLD_BINARY_INV用法与全局二值化一样
blockSize:邻近区域的大小
C:常量,应从计算出的平均值或加权平均值中减去,一般设置为0
import cv2 as cv
import numpy as np
img = cv.imread('./img/houmo.png')
img1 = cv.cvtColor(img,cv.COLOR_BGR2GRAY)
ret,dst = cv.threshold(img1,127,255,cv.THRESH_BINARY)
# 与上面效果相反
# ret,dst = cv.threshold(img1,127,255,cv.THRESH_BINARY_INV)
print(ret,dst.shape)
print(dst)
cv.imshow('img1',img1)
cv.imshow('dst',dst)
cv.waitKey(0)
局部二值化adaptiveThreshold
import cv2 as cv
import numpy as np
img = cv.imread('./img/houmo.png')
img1 = cv.cvtColor(img,cv.COLOR_BGR2GRAY)
dst = cv.adaptiveThreshold(img1,255,cv.ADAPTIVE_THRESH_GAUSSIAN_C,cv.THRESH_BINARY,7,0)
print(dst)
cv.imshow('img1',img1)
cv.imshow('dst',dst)
cv.waitKey(0)