【OpenCV】图像直方图及操作

一、图像像素值直方图

1.1 图像

import cv2 as cv
import numpy as np
from matplotlib import pyplot as plt
def plot_demo(image):
    '''统计每个像素值在图片中出现的次数'''
    plt.hist(image.ravel(),256,[0,256])#参数：bin是256，范围【0,256】
    plt.show()
    
src=cv.imread(r'D:\Project\Opencv\Learning01\angelababy.jpg')
cv.imshow("origin image",src)

plot_demo(src)
cv.waitKey(0)

1.2 图像通道cv.calcHist

import cv2 as cv
import numpy as np
from matplotlib import pyplot as plt
'''绘制每个通道的像素值的直方图：既每个通道下，每个像素出现的频次'''
def image_hist_demo(image):
    ''''''
    color=('blue','green','red')
    for i ,color in enumerate(color):
        hist=cv.calcHist(image,[i],None,[256],[0,255])#图像，通道，mask,bins(直方图size)、range[0,256]
        plt.plot(hist,color=color)
        plt.xlim([0,256])
    plt.show()


src=cv.imread(r'D:\Project\Opencv\Learning01\angelababy.jpg')
cv.imshow("origin image",src)

image_hist_demo(src)
cv.waitKey(0)

二、直方图应用——均衡化

  直方图是图像的统计学特征。直方图举例：像素值直方图，边缘、方向，梯度直方图等等。
  直方图有很多应用：均衡化（增强对比度）、有颜色物体跟踪…

  图像直方图均值化都是基于灰度图。直方图均值化自动调整图像对比度，使图像清晰，是图像增强的一个手段

2.1 全局均衡化

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

def equalHist_demo(image):
    ''''''
    gray=cv.cvtColor(image,cv.COLOR_BGR2GRAY)#图像转灰度图
    dst=cv.equalizeHist(gray)
    cv.imshow('equalHist',dst)


src=cv.imread(r'D:\Project\Opencv\Learning01\angelababy.jpg')
cv.imshow("origin image",src)

equalHist_demo(src)
cv.waitKey(0)

2.2 自适应（局部）均值化

import cv2 as cv
import numpy as np
from matplotlib import pyplot as plt
"""全局均衡化"""
def equalHist_demo(image):
    ''''''
    gray=cv.cvtColor(image,cv.COLOR_BGR2GRAY)#图像转灰度图
    dst=cv.equalizeHist(gray)
    cv.imshow('equalHist',dst)

"""自适应（局部）均衡化"""
def clahe_demo(image):
    ''''''
    gray=cv.cvtColor(image,cv.COLOR_BGR2GRAY)#图像转灰度图
    clahe=cv.createCLAHE(clipLimit=2.0,tileGridSize=(8,8))
    dst=clahe.apply(gray)
    cv.imshow('clahe lHist',dst)


src=cv.imread(r'D:\Project\Opencv\Learning01\angelababy.jpg')
cv.imshow("origin image",src)
equalHist_demo(src)
clahe_demo(src)
cv.waitKey(0)

  （左）原图、（中）全局均衡化、（右）自适应或局部均衡化

三、直方图反向投影

  在HSV色彩空间进行反向投影。直方图反向投影是基于2D直方图。
          

3.1 2D直方图

import cv2 as cv
from matplotlib import pyplot as plt

"""计算图像的2D直方图"""
def hist2d_demo(image):
    hsv=cv.cvtColor(image,cv.COLOR_BGR2HSV)
    hist=cv.calcHist([hsv],[0,1],None,[750,500],[0,750,0,500])#计算两个通道的直方图：图像、通道、mask、binsize、范围
    ####方法1
    cv.imshow('hist 2d',hist)
    ####方法2
    # plt.imshow(hist,interpolation='nearest')
    # plt.title('hist2D hist')
    # plt.show()

src=cv.imread(r'D:\Project\Opencv\Learning01\angelababy.jpg')
print(src.shape)
cv.imshow("origin image",src)
hist2d_demo(src)

cv.waitKey(0)

3.2 直方图反向投影

import cv2 as cv
from matplotlib import pyplot as plt

"""计算图像的2D直方图"""
def hist2d_demo(image):
    hsv=cv.cvtColor(image,cv.COLOR_BGR2HSV)
    hist=cv.calcHist([hsv],[0,1],None,[750,500],[0,750,0,500])#计算两个通道的直方图：图像、通道、mask、binsize、范围
    ####方法1
    cv.imshow('hist 2d',hist)
    ####方法2
    # plt.imshow(hist,interpolation='nearest')
    # plt.title('hist2D hist')
    # plt.show()
"""直方图反向投影"""
def back_projection_demo():
    target=cv.imread(r'D:\Project\Opencv\Learning01\target.png')
    roi=cv.imread(r'D:\Project\Opencv\Learning01\kuai.png')
    targev_hsc=cv.cvtColor(target,cv.COLOR_BGR2HSV)
    roi_hsv = cv.cvtColor(roi, cv.COLOR_BGR2HSV)
    # 计算2d直方图
    roiHist=cv.calcHist([roi_hsv],[0,1],None,[32,32],[0,180,0,256])#[32,32]代表直方图bin的个数，数越大，代表像素越细分，更加碎片化
    cv.normalize(roiHist,roiHist,0,255,cv.NORM_MINMAX)
    dst=cv.calcBackProject([targev_hsc],[0,1],roiHist,[0,180,0,256],1)
    cv.imshow('backprojection',dst)
    cv.imshow('roi',roi)


src=cv.imread(r'D:\Project\Opencv\Learning01\target.png')
print(src.shape)
cv.imshow("origin image",src)
back_projection_demo()
cv.waitKey(0)