python opencv图像阀值

opencv阀值应用场景主要用于从黑白相片中提取出黑色目标或者是白色目标。
全局阀值：
阀值类型

全局阀值的函数为cv2.threshold (src, thresh, maxval, type)

import cv2
from matplotlib import pyplot as plt

img = cv2.imread('sunshine.jpg',0)
ret,res1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
ret,res2 = cv2.threshold(img,127,255,cv2.THRESH_BINARY_INV)
ret,res3 = cv2.threshold(img,127,255,cv2.THRESH_TRUNC)
ret,res4 = cv2.threshold(img,127,255,cv2.THRESH_TOZERO)
ret,res5 = cv2.threshold(img,127,255,cv2.THRESH_TOZERO_INV)
titles = ['original','binary','ninary_inv','trunc','tozero','tozero_inv']
images = [img,res1,res2,res3,res4,res5]
for i in range(6):
    plt.subplot(2,3,i+1),plt.imshow(images[i],'gray'),plt.title(titles[i])
    plt.xticks([]),plt.yticks([])
plt.show()

自适应阀值，主要是利用算法解决因阳光导致照片局部亮度不同，从而引起的提取误差。
主要函数是dst = cv2.adaptiveThreshold(src, maxval, thresh_type, type, Block Size)

import cv2
from matplotlib import pyplot as plt
#
img = cv2.imread('sunshine.jpg',0)
ret,res1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
res2 = cv2.adaptiveThreshold(img,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,13,2)
res3 = cv2.adaptiveThreshold(img,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,13,2)
images = [img, res1, res2,res3]
titles = ['original','binary','mean','gaussian']
for i in range(4):
    plt.subplot(2,2,i+1),plt.imshow(images[i],'gray')
    plt.title(titles[i]),plt.xticks([]),plt.yticks([])
plt.show()

OTSU二值化，
otsu二值化相比全局阀值不同之处是自动确定阀值
这里效果并不明显，可能是因为图片比较完好，没有高斯噪声：

import cv2
from matplotlib import pyplot as plt
#
img = cv2.imread('sunshine.jpg',0)
ret,res1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
res2 = cv2.adaptiveThreshold(img,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,13,2)
ret,res3 = cv2.threshold(img,0,255,cv2.THRESH_BINARY + cv2.THRESH_OTSU)
# 高斯滤波后再采用Otsu阈值
res4 = cv2.GaussianBlur(img,(5,5),0)
ret,res5 = cv2.threshold(img,0,255,cv2.THRESH_BINARY + cv2.THRESH_OTSU)
images = [img,res1,res2,res3,res4,res5]
titles = ['original','binary','gaussian','otsu','gaussian_filter','otsu_gaussian_filter']
for i in range(6):
    plt.subplot(2,3,i+1),plt.imshow(images[i],'gray')
    plt.title(titles[i]),plt.xticks([]),plt.yticks([])
plt.show()