Python+OpenCV实现阈值分割的方法详解

一、全局阈值

原图：

整幅图采用一个阈值，与图片的每一个像素灰度进行比较，重新赋值；

1.效果图

2.源码

import cv2
import matplotlib.pyplot as plt
#设定阈值
thresh=130
#载入原图，并转化为灰度图像
img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)
img_original=cv2.resize(img_original,(0,0),fx=0.3,fy=0.3)
#采用5种阈值类型（thresholding type）分割图像
retval1,img_binary=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY)
retval2,img_binary_invertion=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY_INV)
retval3,img_trunc=cv2.threshold(img_original,thresh,255,cv2.THRESH_TRUNC)
retval4,img_tozero=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO)
retval5,img_tozero_inversion=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO_INV)
#采用plt.imshow()显示图像
imgs=[img_original,img_binary,img_binary_invertion,img_trunc,img_tozero,img_tozero_inversion]
titles=['original','binary','binary_inv','trunc','tozero','tozero_inv']
for i in range(6):
   plt.subplot(2,3,i+1)
   plt.imshow(imgs[i],'gray')
   plt.title(titles[i])
   plt.xticks([])
   plt.yticks([])
plt.show()

二、滑动改变阈值(滑动条)

1.效果图

2.源码

代码如下（示例）：

import cv2
import numpy as np
import matplotlib.pyplot as plt
#载入原图，转化为灰度图像,并通过cv2.resize()等比调整图像大小
img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)
img_original=cv2.resize(img_original,(0,0),fx=0.3,fy=0.3)
#初始化阈值，定义全局变量imgs
thresh=130
imgs=0
#创建滑动条回调函数，参数thresh为滑动条对应位置的数值
def threshold_segmentation(thresh):
   #采用5种阈值类型（thresholding type）分割图像
   retval1,img_binary=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY)
   retval2,img_binary_invertion=cv2.threshold(img_original,thresh,255,cv2.THRESH_BINARY_INV)
   retval3,img_trunc=cv2.threshold(img_original,thresh,255,cv2.THRESH_TRUNC)
   retval4,img_tozero=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO)
   retval5,img_tozero_inversion=cv2.threshold(img_original,thresh,255,cv2.THRESH_TOZERO_INV)
   #由于cv2.imshow()显示的是多维数组（ndarray）,因此我们通过np.hstack(数组水平拼接)
   #和np.vstack(竖直拼接)拼接数组，达到同时显示多幅图的目的
   img1=np.hstack([img_original,img_binary,img_binary_invertion])
   img2=np.hstack([img_trunc,img_tozero,img_tozero_inversion])
   global imgs
   imgs=np.vstack([img1,img2])
#新建窗口
cv2.namedWindow('Images')
#新建滑动条，初始位置为130
cv2.createTrackbar('threshold value','Images',130,255,threshold_segmentation)
#第一次调用函数
threshold_segmentation(thresh)
#显示图像
while(1):
   cv2.imshow('Images',imgs)
   if cv2.waitKey(1)==ord('q'):
       break
cv2.destroyAllWindows()

三、自适应阈值分割

1.效果图

2.源码

代码如下（示例）：

import cv2
import matplotlib.pyplot as plt
#载入原图
img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)
#全局阈值分割
retval,img_global=cv2.threshold(img_original,130,255,cv2.THRESH_BINARY)
#自适应阈值分割
img_ada_mean=cv2.adaptiveThreshold(img_original,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,15,3)
img_ada_gaussian=cv2.adaptiveThreshold(img_original,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,15,3)
imgs=[img_original,img_global,img_ada_mean,img_ada_gaussian]
titles=['Original Image','Global Thresholding(130)','Adaptive Mean','Adaptive Guassian',]
#显示图片
for i in range(4):
   plt.subplot(2,2,i+1)
   plt.imshow(imgs[i],'gray')
   plt.title(titles[i])
   plt.xticks([])
   plt.yticks([])
plt.show()

3.GaussianBlur()函数去噪

代码如下（示例）：

import cv2
import matplotlib.pyplot as plt
#载入原图
img_original=cv2.imread(r'E:\py\python3.7\test2\test14yuzhi\cell.png',0)
#高斯滤波
img_blur=cv2.GaussianBlur(img_original,(13,13),13)  #根据情况修改参数
#自适应阈值分割
img_thresh=cv2.adaptiveThreshold(img_original,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,15,3)
img_thresh_blur=cv2.adaptiveThreshold(img_blur,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,15,3)
#显示图像
imgs=[img_thresh,img_thresh_blur]
titles=['img_thresh','img_thresh_blur']
for i in range(2):
   plt.subplot(1,2,i+1)
   plt.imshow(imgs[i],'gray')
   plt.title(titles[i])
   plt.xticks([])
   plt.yticks([])
plt.show()

四、参数解释

1.cv2.threshold(src, thresh, maxval, type)

参数：

src:输入的图像

thresh:图像分割所用的阈值（threshold value）

maxval:当阈值类型（thresholding type）采用cv2.THRESH_BINARY和cv2.THRESH_BINARY_INV时像素点被赋予的新值

type:介绍6种类型：

cv2.THRESH_BINARY（当图像某点像素值大于thresh(阈值)时赋予maxval，反之为0。注：最常用）

cv2.THRESH_BINARY_INV（当图像某点像素值小于thresh时赋予maxval，反之为0）

cv2.THRESH_TRUNC（当图像某点像素值大于thresh时赋予thresh,反之不变。注：虽然maxval没用了，但是调用函数不能省略）

cv2.THRESH_TOZERO（当图像某点像素值小于thresh时赋予0，反之不变。注：同上）

cv2.THRESH_TOZERO_INV（当图像某点像素值大于thresh时赋予0，反之不变。注：同上）

cv2.THRESH_OTSU（该方法自动寻找最优阈值，并返回给retval,见下文）

返回值：

retval:设定的thresh值，或者是通过cv2.THRESH_OTSU算出的最优阈值

dst:阈值分割后的图像

来源：https://blog.csdn.net/qq_45365214/article/details/124750672