winggyn的专栏

背景减除法：1.GMM MOG2算法，高斯混合模型分离算法，它为每个像素选择适当数量的高斯分布函数：cv2.createBackgroundSubtractorMOG2(int history=500,double varThread=16,bool detectShadows =true)2.KNN模型：cv2.createBackgroundSubtractorKNN()方法：通过视频中的背景进行建模，实现背景消除，生成mask图像，通过对mask二值图像分析实现对前景活动对象的区域

import cv2import numpy as npes = cv2.getStructuringElement(cv2.MORPH_ELLIPSE,(10,10))kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE,(3,3))cap = cv2.VideoCapture("anew.flv")frame1 = cap.read()[1]prvs = cv2.cvtColor(frame1,cv2.COLOR_BGR2GRAY)...

import cv2camera = cv2.VideoCapture('anew.flv')out_fps =12.0fourcc = cv2.VideoWriter_fourcc('M','P','4','2')out1 = cv2.VideoWriter('v1.avi',fourcc,out_fps,(500,400))out2 = cv2.VideoWriter('v2.avi',fourcc,out_fps,(500,400))lastFrame = Nonewhile came.

import cv2print(cv2.__file__)img = cv2.imread('test5.jpg')#加载人脸特征,该文件在Python安装目录下，下面是我的绝对路径face_cascade = cv2.CascadeClassifier(r'/Users/wing/anaconda3/lib/python3.6/site-packages/cv2/data/haarcascade_frontalface_default.xml')gray = cv2.cvtC...

"""参数1 即将保存的文件路径参数2 VideoWriter_fourcc为视频编解码器 fourcc意为四字符代码（Four-Character Codes），顾名思义，该编码由四个字符组成,下面是VideoWriter_fourcc对象一些常用的参数,注意：字符顺序不能弄混 cv2.VideoWriter_fourcc('I', '4', '2', '0'),该参数是YUV编码类型，文件名后缀为.avi cv2.VideoWriter_fourcc('P', 'I', .

import cv2import numpy as npdef template_demo(tpl,target): methods = [cv2.TM_SQDIFF_NORMED,cv2.TM_CCORR_NORMED,cv2.TM_CCOEFF_NORMED] th,tw =tpl.shape[:2] for md in methods: result = cv2.matchTemplate(target,tpl,md) min_val,ma.

import cv2import numpy as npdef LBP(src): height = src.shape[0] width = src.shape[1] dst = src.copy() lbp_value = np.zeros((1,8),dtype = np.uint8) neighbours = np.zeros((1,8),dtype =np.uint8) for x in range(1,width-1): fo.

Harris实现过程：计算图像在X和Y方向的梯度：计算图像两个方向梯度的乘积使用高斯函数对三者进行高斯加权，生成矩阵M的A，B，C；计算每个像素的Harris相应值R，并对小于某一阈值t的R置为零；在3*3或者5*5邻域内进行非最大值抑制，局部最大值点即为图像中的角点。cv2.cornerHarris()参数如下：img-数据类型为float32的输入图像blocksize-角点检测中的领域大小ksize-Sobel求导中使用的窗口大小k-Harris角点

HOG实现过程：1灰度化（将图像看做一个x,y,z(灰度）的三维图像）2采用Gamma校正法对输入图像进行颜色空间的标准化（归一化）3计算图像每个像素的梯度（包括大小和方向）4将图像划分为cells5统计每个cell的梯度直方图（不同梯度的个数），得到cell的描述子6将每几个cell组成一个block，得到block的描述子7将图像image内的所有block的HOG特征descriptor串联起来就可以得到HOG特征，该特征向量就是用来目标检测或分类的特征。...

分水岭算法步骤：1加载原始图像2阈值分割，将图像分割为黑白两个部分3对图像进行开运算，即先腐蚀再膨胀4对开运算的结果再进行膨胀，得到大部分是背景的区域5通过距离变换Distance Transform获取前景区域6背景区域sure_bg和前景区域sure_fg相减，得到既有前景又有背景的重合区域7连通区域处理8最后使用分水岭算法代码有问题，不知道错哪里了。。import cv2import numpy as npimg = cv2.imread('test4

区域生长原理： 1 2 1 1 1 2 2 2 1 2 1 1 2 2 基本思想：将具有相似性质的像素集合起来构成区域。步骤;1.对图像顺序扫描，找到第1个还没有归属的像素，...

梯度算子：是一阶导数算子，是水平G(x),G(y)方向对应模板的组合，也有对角线方向。常见的一阶算子：Roberts交叉算子，Rrewitt算子，Sobel算子0 1 -1 0 =1 0 0 -1 Roberts交叉算子是对角线方向的梯度算子，对应的水平方向和竖直方向的梯度分别为：Robert算子优缺点：优点：边缘定位比较准，适用于边缘明显且噪声较少的图像。缺点：没有描述水平和竖直方向的灰度变化，只是关注了对角线方向，鲁棒性差，..

#自适应阈值与固定阈值对比import cv2import matplotlib.pyplot as pltimg = cv2.imread('test.jpg',0)ret,th1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)th2 = cv2.adaptiveThreshold(img,255,cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,55,4)th3 = cv2.adaptiveThresho.

THRESH_BINARY:THRESH_BINARY_INV:THRESH_TRUNC:THRESH_TOZERO:THRESH_TOZERO_INV:import cv2import matplotlib.pyplot as pltimg = cv2.imread('test3.jpg',0)ret,thresh1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)ret,thr...

#黑帽：输入图像和图像闭运算之差import cv2 as cvimport numpy as npimport matplotlib.pyplot as pltimg = cv.imread('test9.jpg')img = cv.cvtColor(img,cv.COLOR_BGR2RGB)kernel = np.ones((9,9),np.uint8)tophat = cv.morphologyEx(img,cv.MORPH_BLACKHAT,kernel)plt.subplot(1.

#顶帽：输入图像和图像开运算之差，下面的示例针对9*9内核完成。import cv2 as cvimport numpy as npimport matplotlib.pyplot as pltimg = cv.imread('test9.jpg')img = cv.cvtColor(img,cv.COLOR_BGR2RGB)kernel = np.ones((5,5),np.uint8)tophat = cv.morphologyEx(img,cv.MORPH_TOPHAT,kernel).

#形态学梯度，这是图像膨胀和腐蚀之间的区别，结果将看起来像对象的轮廓import cv2 as cvimport numpy as npimport matplotlib.pyplot as pltimg = cv.imread('test9.jpg')img = cv.cvtColor(img,cv.COLOR_BGR2RGB)kernel = np.ones((3,3),np.uint8)gradient = cv.morphologyEx(img,cv.COLOR_BGR2RGB,ke.

#闭运算，闭运算与开运算相反，先扩张然后再侵蚀。在关闭前景对象内部的小孔或对象上的小黑点时很有用。import cv2import numpy as npimport matplotlib.pyplot as pltimg = cv2.imread('test9.jpg')kernel = np.ones((3,3),np.uint8)closing = cv2.morphologyEx(img,cv2.MORPH_CLOSE,kernel)plt.subplot(121),plt.imsh.

#开运算，腐蚀然后扩张的另一个名字，它对于消除噪音很有用import cv2import numpy as npimport matplotlib.pyplot as pltimg = cv2.imread('test9.jpg')img = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)kernel = cv2.getStructuringElement(cv2.MORPH_RECT,(5,5))opening = cv2.morphologyEx(img,cv2..

import matplotlib.pyplot as pltimport cv2img = cv2.imread('test9.jpg')img = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)kernel = cv2.getStructuringElement(cv2.MORPH_RECT,(3,3))dilation = cv2.dilate(img,kernel,iterations =1)plt.subplot(121),plt.imshow(img),pl.

import cv2import numpy as npimport matplotlib.pyplot as pltimg = cv2.imread('test9.jpg')img = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)kernel = np.ones((3,3),np.uint8)erosion = cv2.erode(img,kernel,iterations = 1)plt.subplot(121),plt.imshow(img),plt.titl.

#Gamma变化import cv2import numpy as npimg = cv2.imread('test.jpg')def adjust_gamma(image,gamma = 1.0): invGamma = 1.0/gamma table = [] for i in range(256): table.append(((i/255.0)**invGamma)*255) table = np.array(table).astype...

import cv2import numpy as npimg = cv2.imread("test.jpg",1)cv2.imshow("src",img)#彩色图像均衡化，需要分解通道，对每一个通道均衡化(b,g,r) = cv2.split(img)bH = cv2.equalizeHist(b)gH = cv2.equalizeHist(g)rH = cv2.equalizeHist(r)result = cv2.merge((bH,gH,rH))cv2.imshow("dst.

import cv2img = cv2.imread('test9.jpg',0)cv2.imshow("dark",img)#调用cv2.equalizeHist函数进行直方图均衡化img_equal = cv2.equalizeHist(img)cv2.imshow("img_equal",img_equal)cv2.waitKey(0)cv2.destoryAllWindows()

#双边滤波import cv2from matplotlib import pyplot as pltimg = cv2.imread('test.jpg')img = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)blur = cv2.bilateralFilter(img,-1,15,10)plt.subplot(121),plt.imshow(img),plt.title('Original')plt.xticks([]),plt.yticks([])plt..

#中值import cv2 as cvimport numpy as npfrom matplotlib import pyplot as pltimg = cv.imread('test.jpg')median = cv.medianBlur(img,9)plt.subplot(121),plt.imshow(img),plt.title('Original')plt.xticks([]),plt.yticks([])plt.subplot(122),plt.imshow(median).

import cv2 as cvimport numpy as npfrom matplotlib import pyplot as pltimg = cv.imread('test.jpg')img = cv.cvtColor(img,cv.COLOR_BGR2RGB)blur = cv.GaussianBlur(img,(7,7),7)plt.subplot(121),plt.imshow(img),plt.title('original')plt.xticks([]),plt.ytic.

#方框滤波import cv2import numpy as npimg = cv2.imread('test.jpg', cv2.IMREAD_UNCHANGED)r = cv2.boxFilter(img,-1,(3,3),normalize=1)d = cv2.boxFilter(img,-1,(3,3),normalize=0)cv2.namedWindow('img',cv2.WINDOW_AUTOSIZE)cv2.namedWindow('r',cv2.WINDOW_AUTOSI.

import cv2import numpy as npimport matplotlib.pyplot as pltsrc = cv2.imread('test.jpg')rows,cols = src.shape[:2]pos1 = np.float32([[114,82],[287,156],[8,100],[143,177]])pos2 = np.float32([[0,0],[188,0],[0,262],[188,262]])M = cv2.getPerspectiveTrans.

import cv2import numpy as npimg = cv2.imread('test.jpg',1)rows,cols = img.shape[:2]M = cv2.getRotationMatrix2D((cols/2,rows/2),45,2)print(M)dst = cv2.warpAffine(img,M,(4*cols,4*rows),borderValue=(25,255,255))while(1): cv2.imshow('img',img) .

import cv2import numpy as npimg = cv2.imread('test.jpg')res1 = cv2.resize(img,None,fx=2,fy=2,interpolation=cv2.INTER_LINEAR)height,width = img.shape[:2]res2 = cv2.resize(img,(int(0.8*width),int(0.8*height)),interpolation=cv2.INTER_AREA)cv2.imshow('o.

import numpy as npimport cv2img = np.zeros((512,512,3),np.uint8)font = cv2.FONT_HERSHEY_SIMPLEXcv2.putText(img,'OpenCV',(50,200),font,3,(0,255,255),cv2.LINE_AA)winname = 'example'cv2.namedWindow(winname)cv2.imshow(winname,img)cv2.waitKey(0)cv2.de.