开发: C++知识库 Java知识库 JavaScript Python PHP知识库人工智能区块链大数据移动开发嵌入式开发工具数据结构与算法开发测试游戏开发网络协议系统运维
教程: HTML教程 CSS教程 JavaScript教程 Go语言教程 JQuery教程 VUE教程 VUE3教程 Bootstrap教程 SQL数据库教程 C语言教程 C++教程 Java教程 Python教程 Python3教程 C#教程
数码: 电脑笔记本显卡显示器固态硬盘硬盘耳机手机 iphone vivo oppo 小米华为单反装机图拉丁

-> 人工智能 -> 十一、图像二值化 -> 正文阅读

[人工智能]十一、图像二值化

一、二值图像

其实就是把图像转换为只有黑白的两种颜色图像，即像素值非零即一

三角阈值二值化

对一个图像进行操作，获取图像的直方图，找到波峰和波谷进行连线设为线段A，每个点做有关线段A的垂线垂足在线段A上，最后将所有的垂足的横坐标累加再加上一个偏量位移得到的值也就是图像的阈值，通过这个阈值对整个图像进行二值化操作。

二、图像二值化的方法

1、全局阈值

OTSU：内方差最小，外方差最大（常用）

import cv2
import numpy as np
from matplotlib import pyplot as plt


def threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    ret,binary = cv2.threshold(gray,0,255,cv2.THRESH_BINARY|cv2.THRESH_OTSU)
    print("threshold value is:%s"%ret)
    cv2.imshow("OTSU",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

Triangle

import cv2
import numpy as np
from matplotlib import pyplot as plt


def threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    ret,binary = cv2.threshold(gray,0,255,cv2.THRESH_BINARY|cv2.THRESH_TRIANGLE)
    print("threshold value is:%s"%ret)
    cv2.imshow("TRIANGLE",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

在这里插入图片描述

自动与手动设置阈值

①THRESH_BINARY手动设置阈值为127，像素值小于127为0黑色，大于127为255白色

import cv2
import numpy as np
from matplotlib import pyplot as plt


def threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    ret,binary = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
    print("threshold value is:%s"%ret)
    cv2.imshow("binary",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

②THRESH_BINARY_INV手动设置阈值为127，取反效果，像素值小于127为255白色，大于127为0黑色

import cv2
import numpy as np
from matplotlib import pyplot as plt


def threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    ret,binary = cv2.threshold(gray,127,255,cv2.THRESH_BINARY_INV)
    print("threshold value is:%s"%ret)
    cv2.imshow("binary",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

③THRESH_TRUNC手动设置阈值为127，截断，像素值大于127为127，小于127为0黑色

import cv2
import numpy as np
from matplotlib import pyplot as plt


def threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    ret,binary = cv2.threshold(gray,127,255,cv2.THRESH_TRUNC)
    print("threshold value is:%s"%ret)
    cv2.imshow("TRUNC",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

④THRESH_TOZERO手动设置阈值为127，截断，像素值小于127为0黑色，大于127像素值不变

import cv2
import numpy as np
from matplotlib import pyplot as plt


def threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    ret,binary = cv2.threshold(gray,127,255,cv2.THRESH_TOZERO)
    print("threshold value is:%s"%ret)
    cv2.imshow("TOZERO",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

在这里插入图片描述

2、局部阈值

Ⅰ局部自适应阈值方法

①THRESH_MEAN_C

MEAN_C：将图片分为若干个小方格，求出每一个方格中像素的均值，判断均值是否大于自定义的阈值，大于阈值为白色，小于阈值为黑色
adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,25,10)
gray必须为灰度图
255为maxValue
25为blockSize，必须是奇数
10为C，是一个常量，每个像素块的均值与计算出来的均值差不能大于10，否则视为无效，其目的为了防止噪声的影响

import cv2
import numpy as np
from matplotlib import pyplot as plt


def local_threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    binary = cv2.adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,25,10)
    cv2.imshow("MEAN_C",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

local_threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：
在这里插入图片描述

②THRESH_GAUSSIAN_C

GAUSSIAN_C：将图片分为若干个小方格，求出每一个方格中像素的均值并乘以相应的高斯权重(中间的权重更大)，判断均值是否大于自定义的阈值，大于阈值为白色，小于阈值为黑色
adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,25,10)
gray必须为灰度图
255为maxValue
25为blockSize，必须是奇数
10为C，是一个常量，每个像素块的均值与计算出来的均值差不能大于10，否则视为无效，其目的为了防止噪声的影响

import cv2
import numpy as np
from matplotlib import pyplot as plt


def local_threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    binary = cv2.adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,25,10)
    cv2.imshow("binary",binary)

src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)
local_threshold(src)
cv2.waitKey(0)
cv2.destroyAllWindows()**

效果图如下：
在这里插入图片描述

Ⅱ自己计算均值然后对图像进行分割从而达到图像二值化效果

import cv2
import numpy as np
from matplotlib import pyplot as plt


def custom_threshold(image):
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    h,w = gray.shape[:2]
    m = np.reshape(gray,[1,h*w])
    mean = m.sum()/h*w
    print("mean:",mean)
    ret ,binary = cv2.threshold(gray,mean,255,cv2.THRESH_BINARY)
    cv2.imshow("binary",binary)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\a1.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)

custom_threshold(src)

cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：在这里插入图片描述

三、超大图像二值化

Ⅰ自动全局方法操作

思想：对超大图像进行分块二值化操作

import cv2
import numpy as np
from matplotlib import pyplot as plt


def big_image_binary(image):
    print(image.shape)
    cw = 256
    ch = 256
    h,w = image.shape[:2]
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    for row in range(0,h,ch):
        for col in range(0,w,cw):
            roi = gray[row:row+ch,col:col+cw]
            ret,dst = cv2.threshold(roi,0,255,cv2.THRESH_BINARY|cv2.THRESH_OTSU)
            gray[row:row+ch,col:col+cw] = dst
            print(np.std(dst),np.mean(dst))
    cv2.imwrite(r"G:\Juptyer_workspace\study\opencv\opencv3\big1.jpg",gray)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\big.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)
big_image_binary(src)


cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：
在这里插入图片描述

对图像噪声进行过滤之后进行全局二值化操作

import cv2
import numpy as np
from matplotlib import pyplot as plt


def big_image_binary(image):
    print(image.shape)
    cw = 256
    ch = 256
    h,w = image.shape[:2]
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    for row in range(0,h,ch):
        for col in range(0,w,cw):
            roi = gray[row:row+ch,col:col+cw]
            print(np.std(roi),np.mean(roi))
            dev = np.std(roi)
            if dev < 15:
                gray[row:row+ch,col:col+cw] = 255
            else :
                ret,dst = cv2.threshold(roi,0,255,cv2.THRESH_BINARY|cv2.THRESH_OTSU)
                gray[row:row+ch,col:col+cw] = dst
               
    cv2.imwrite(r"G:\Juptyer_workspace\study\opencv\opencv3\big3.jpg",gray)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\big.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)
big_image_binary(src)


cv2.waitKey(0)
cv2.destroyAllWindows()

效果图如下：
在这里插入图片描述

Ⅱ自适应局部方法操作

import cv2
import numpy as np
from matplotlib import pyplot as plt


def big_image_binary(image):
    print(image.shape)
    cw = 256
    ch = 256
    h,w = image.shape[:2]
    gray = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
    for row in range(0,h,ch):
        for col in range(0,w,cw):
            roi = gray[row:row+ch,col:col+cw]
            dst = cv2.adaptiveThreshold(roi,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY,127,20)
            gray[row:row+ch,col:col+cw] = dst
            print(np.std(dst),np.mean(dst))
    cv2.imwrite(r"G:\Juptyer_workspace\study\opencv\opencv3\big2.jpg",gray)


src = cv2.imread(r"G:\Juptyer_workspace\study\opencv\opencv3\big.jpg")
cv2.imshow("image",src)
cv2.namedWindow("image",cv2.WINDOW_AUTOSIZE)
big_image_binary(src)


cv2.waitKey(0)
cv2.destroyAllWindows()