opencv
基于https://www.bilibili.com/video/BV1PV411774y/ 学习笔记
感觉学完了一半还是不会高级的用法,只会简单的图像读取展示,其中有很多概念根本听不懂。。
只会跟着敲代码,看结果展示。。
基本用法
数据读取-图像
- cv2.IMREAD_COLOR : 彩色图像
- cv2.IMREAD_GRAYSCALE : 灰度图像
注意:OpenCV读取的格式是BGR
图像的显示
import cv2
img = cv2.imread("money.png")
def cv_show(name,img):
cv2.imshow(name,img)#展示图像
cv2.waitKey(0)#一直显示
cv2.destroyAllWindows()#按下任意键关闭窗口
cv_show("money",img)
获取图像 高 宽
print(img.shape)
#(120, 400, 3)#h w 3表示BGR
读取灰度图像
img = cv2.imread("money.png",cv2.IMREAD_GRAYSCALE)
print(img.shap)
#(120,400)#只有一个颜色通道
图像保存
cv2.imwrite("money_gray.png",img)
读取视频
#coding:utf-8
import cv2
vc = cv2.VideoCapture("kungfu.mp4")
#检查是否打开正确
if vc.isOpened():
open,frame = vc.read()
else:
open = False
while open:
ret,frame = vc.read()
if frame is None:
break;
if ret == True:
gray = cv2.cvtColor(frame,cv2.COLOR_BGR2GRAY)
cv2.imshow("result",gray)
if cv2.waitKey(10) & 0xFF == ord('q'):
break;
vc.release()
cv2.destroyAllWindows()
转换成灰色视频播放,检测到q
按键退出
截取部分图像数据
#coding:utf-8
import cv2
img = cv2.imread("money.png",cv2.IMREAD_GRAYSCALE)
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
clip = img[0:200,0:200]
cv_show("clip",clip)
颜色通道提取
#coding:utf-8
import cv2
img = cv2.imread("money.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
b,g,r= cv2.split(img)
print(b)
组合
img = cv2.merge([r,g,b])
只保留R
cur_img = img.copy()
cur_img[:,:,0] = 0
cur_img[:,:,1] = 0
cv_show("cur_img",cur_img)
只保留G
cur_img = img.copy()
cur_img[:,:,0] = 0
cur_img[:,:,2] = 0
cv_show("cur_img",cur_img)
边界填充
#coding:utf-8
import cv2
img = cv2.imread("money.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
top_size,bottom_size,left_size,right_size = (50,50,50,50)
replicate = cv2.copyMakeBorder(img,top_size,bottom_size,left_size,right_size,borderType=cv2.BORDER_REPLICATE)
reflect = cv2.copyMakeBorder(img,top_size,bottom_size,left_size,right_size,borderType=cv2.BORDER_REFLECT)
reflect101 = cv2.copyMakeBorder(img,top_size,bottom_size,left_size,right_size,borderType=cv2.BORDER_REFLECT_101)
wrap = cv2.copyMakeBorder(img,top_size,bottom_size,left_size,right_size,borderType=cv2.BORDER_WRAP)
constant = cv2.copyMakeBorder(img,top_size,bottom_size,left_size,right_size,borderType=cv2.BORDER_CONSTANT,value=[0,0,0])
cv_show("0",img)
cv_show("1",replicate)
cv_show("2",reflect)
cv_show("3",reflect101)
cv_show("4",wrap)
cv_show("5",constant)
import matplotlib.pyplot as plt
plt.subplot(231),plt.imshow(img,'gray'),plt.title('ORIGINAL')
plt.subplot(232),plt.imshow(replicate,'gray'),plt.title('REPLICATE')
plt.subplot(233),plt.imshow(reflect,'gray'),plt.title('REFLECT')
plt.subplot(234),plt.imshow(reflect101,'gray'),plt.title('REFLECT_101')
plt.subplot(235),plt.imshow(wrap,'gray'),plt.title('WRAP')
plt.subplot(236),plt.imshow(constant,'gray'),plt.title('CONSTANT')
plt.show()
- BORDER_REPLICATE: 复制法,也就是复制边缘像素
- BORDER_REFLECT: 反射法,对感兴趣的图像中的像素在两边进行复制例如:fedcba|abcdefgh|hgfedcb
- BORDER_REFLECT_101: 反射法,也就是以最为边缘像素为轴,对称,gfedcb|abcdefgh|gfedcba
- BORDER_WRAP: 外包装法cdefgh|abcdefgh|abcdefg
- BORDER_CONSTANT: 常量法,常量值填充
改变图像长宽
img = cv2.resize(img,(50,50))#长宽改为50x50
img = cv2.resize(img,(0,0),fx=3,fy=1)#长x3,宽x1
数值计算
img = cv2.imread("money.png")
print(img[:5,:,0])
img += 10
print(img[:5,:,0])
所有像素点都加,再%256(不会越界)
img3 = cv2.add(img,img2)#像素点越界时,取255
图像融合
如果两个图像的size不一样,相加会报错
img2 = cv2.resize(img2,(500,500))#改变长宽
#res = A*img1 + B*img2 + l
res = cv2.addWeighted(img1,0.4,img2,0.6,0)
img1占比A,img2占比B,亮度l
图像阈值
ret,dst = cv2.threshold(src,thresh,maxval,type)
- src: 输入图,只能输入单通道图像,通常来说为灰度图
- dst:输出图
- thresh:阈值
- maxval:当像素值超过了阈值(或者小于阈值,根据type来决定),所赋予的值
- type:二值化操作的类型,包含以下5种类型:
- cv2.THRESH_BINARY; 超过阈值部分取maxval(最大值),否则取0
- cv2.THRESH_BINARY_INV; 上面相反
- cv2.THRESH_TRUNC; 大于阈值部分设为阈值,否则不变
- cv2.THRESH_TOZERO; 大于阈值部分不改变,否则设为0
- cv2.THRESH_TOZERO_INV;上面相反
#coding:utf-8
import cv2
import matplotlib.pyplot as plt
img = cv2.imread("money.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
ret,thresh1 = cv2.threshold(img,127,255,cv2.THRESH_BINARY)
ret,thresh2 = cv2.threshold(img,127,255,cv2.THRESH_BINARY_INV)
ret,thresh3 = cv2.threshold(img,127,255,cv2.THRESH_TRUNC)
ret,thresh4 = cv2.threshold(img,127,255,cv2.THRESH_TOZERO)
ret,thresh5 = cv2.threshold(img,127,255,cv2.THRESH_TOZERO_INV)
titles = ['Original Image','BINARY','BINARY_INV','TRUNC','TOZERO','TOZERO_INV']
images = [img, thresh1, thresh2, thresh3, thresh4, thresh5]
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()
图像平滑
均值滤波
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("lady.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#均值滤波
#简单的平均卷积操作
blur = cv2.blur(img,(3,3))
#双图一起展示
show_img = np.hstack((img,blur))
cv_show("show",show_img)
方框滤波
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("lady.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#方框滤波
#基本和均值一样,可以选择归一化
box = cv2.boxFilter(img,-1,(3,3),normalize=True)
#双图一起展示
show_img = np.hstack((img,box))
cv_show("show",show_img)
效果和均值滤波一样
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("lady.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#方框滤波
#基本和均值一样,可以选择归一化,容易越界
box = cv2.boxFilter(img,-1,(3,3),normalize=False)
#双图一起展示
show_img = np.hstack((img,box))
cv_show("show",show_img)
高斯滤波
靠的近的权重大
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("lady.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#高斯滤波
#高斯模糊的卷积核里的数值是满足高斯分布的,相当于更重视中间的
aussian = cv2.GaussianBlur(img,(5,5),1)
#双图一起展示
show_img = np.hstack((img,aussian))
cv_show("show",show_img)
效果和均值滤波差不多
中值滤波
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("lady.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#中值滤波
#相当于中值代替
median = cv2.medianBlur(img,5)
#双图一起展示
show_img = np.hstack((img,median))
cv_show("show",show_img)
形态学-腐蚀操作
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("dige.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
kernel = np.ones((5,5),np.uint8)
erosion = cv2.erode(img,kernel,iterations = 1)#腐蚀一次
erosion2 = cv2.erode(img,kernel,iterations = 2)#腐蚀二次
#双图一起展示
show_img = np.hstack((img,erosion,erosion2))
cv_show("show",show_img)
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("pie.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
kernel = np.ones((30,30),np.uint8)
erosion1 = cv2.erode(img,kernel,iterations = 1)#腐蚀一次
erosion2 = cv2.erode(img,kernel,iterations = 2)#腐蚀二次
erosion3 = cv2.erode(img,kernel,iterations = 3)#腐蚀三次
#双图一起展示
show_img = np.hstack((erosion1,erosion2,erosion3))
cv_show("show",show_img)
形态学-膨胀操作
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("dige.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
kernel = np.ones((3,3),np.uint8)
erosion = cv2.erode(img,kernel,iterations = 1)#先腐蚀去线
dilate = cv2.dilate(erosion,kernel,iterations = 1)#再膨胀修复
#双图一起展示
show_img = np.hstack((img,erosion,dilate))
cv_show("show",show_img)
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("pie.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
kernel = np.ones((30,30),np.uint8)
dilate1 = cv2.dilate(img,kernel,iterations = 1)
dilate2 = cv2.dilate(img,kernel,iterations = 2)
dilate3 = cv2.dilate(img,kernel,iterations = 3)
#双图一起展示
show_img = np.hstack((dilate1,dilate2,dilate3))
cv_show("show",show_img)
开运算与闭运算
#开:先腐蚀,再膨胀
#闭:先膨胀,再腐蚀
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("dige.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#开:先腐蚀,再膨胀
kernel = np.ones((5,5),np.uint8)
opening = cv2.morphologyEx(img,cv2.MORPH_OPEN,kernel)
#双图一起展示
show_img = np.hstack((img,opening))
cv_show("show",show_img)
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("dige.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#闭:先膨胀,再腐蚀
kernel = np.ones((5,5),np.uint8)
opening = cv2.morphologyEx(img,cv2.MORPH_CLOSE,kernel)
#双图一起展示
show_img = np.hstack((img,opening))
cv_show("show",show_img)
梯度运算
梯度 = 膨胀 - 腐蚀
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("pie.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#梯度 = 膨胀 - 腐蚀
kernel = np.ones((7,7),np.uint8)
dilate = cv2.dilate(img,kernel,iterations = 5)
erosion = cv2.erode(img,kernel,iterations = 5)
#双图一起展示
show_img = np.hstack((dilate,erosion))
cv_show("show",show_img)
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("pie.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#梯度 = 膨胀 - 腐蚀
kernel = np.ones((7,7),np.uint8)
gradient = cv2.morphologyEx(img,cv2.MORPH_GRADIENT,kernel)
#双图一起展示
# show_img = np.hstack((gradient))
cv_show("show",gradient)
礼帽与黑帽
- 礼帽 = 原始输入 - 开运算
- 黑帽 = 闭运算 - 原始输入
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("dige.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
kernel = np.ones((5,5),np.uint8)
tophat = cv2.morphologyEx(img,cv2.MORPH_TOPHAT,kernel)
hlackhat = cv2.morphologyEx(img,cv2.MORPH_BLACKHAT,kernel)
#双图一起展示
show_img = np.hstack((tophat,hlackhat))
cv_show("show",show_img)
图像梯度-Sobel算子
作用:找图像边缘。
(右 - 左 ; 下 - 上)
dst = cv2.Sobel(src, ddepth, dx, dy, ksize)
- ddepth: 图像深度
- dx 和 dy 分别表示水平和竖直方向
- ksize 是Sobel算子的大小
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("pie.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
# kernel = np.ones((5,5),np.uint8)
sobelx = cv2.Sobel(img,cv2.CV_64F,1,0,ksize=3)
#右 - 左,白到黑是整数,黑到白就是负数了,
# 所有的负数会被截断成 0 ,
# 所以要取绝对值
# sobelx = np.absolute(sobelx)
sobelx = cv2.convertScaleAbs(sobelx)
#双图一起展示
# show_img = np.hstack((tophat,hlackhat))
cv_show("show",sobelx)
上面只计算了 Gx 的值,还要算 Gy 的值,再求和
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
img = cv2.imread("pie.png")
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
# Gx
sobelx = cv2.Sobel(img,cv2.CV_64F,1,0,ksize=3)
#右 - 左,白到黑是整数,黑到白就是负数了,
# 所有的负数会被截断成 0 ,
# 所以要取绝对值
# sobelx = np.absolute(sobelx)
sobelx = cv2.convertScaleAbs(sobelx)
#Gy
sobely = cv2.Sobel(img,cv2.CV_64F,0,1,ksize=3)
sobely = cv2.convertScaleAbs(sobely)
#sum
sobelxy = cv2.addWeighted(sobelx,0.5,sobely,0.5,0)
#双图一起展示
# show_img = np.hstack((tophat,hlackhat))
cv_show("show",sobelxy)
不建议直接计算,效果不是很好
sobelxy = cv2.Sobel(img,cv2.CV_64F,1,1,ksize=3)
sobelxy = cv2.convertScaleAbs(sobelxy)
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("lena.png",cv2.IMREAD_GRAYSCALE)
sobelx = cv2.Sobel(img,cv2.CV_64F,1,0,ksize=3)
sobelx = cv2.convertScaleAbs(sobelx)
sobely = cv2.Sobel(img,cv2.CV_64F,0,1,ksize=3)
sobely = cv2.convertScaleAbs(sobely)
sobelxy = cv2.addWeighted(sobelx,0.5,sobely,0.5,0)
sobelxy2 = cv2.Sobel(img,cv2.CV_64F,1,1,ksize=3)
sobelxy2 = cv2.convertScaleAbs(sobelxy2)
#双图一起展示
show_img = np.hstack((img,sobelxy,sobelxy2))
cv_show("show",show_img)
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-eyIQidV6-1689045274409)(C:\Users\Administrator\Desktop\photo\image-20230704195817494.png)]
其他算子
laplacian算子对噪点比较敏感
不同算子的差异
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#不同算子的差异
img = cv2.imread("lena.png",cv2.IMREAD_GRAYSCALE)
sobelx = cv2.Sobel(img,cv2.CV_64F,1,0,ksize=3)
sobelx = cv2.convertScaleAbs(sobelx)
sobely = cv2.Sobel(img,cv2.CV_64F,0,1,ksize=3)
sobely = cv2.convertScaleAbs(sobely)
sobelxy = cv2.addWeighted(sobelx,0.5,sobely,0.5,0)
scharrx = cv2.Scharr(img,cv2.CV_64F,1,0)
scharrx = cv2.convertScaleAbs(scharrx)
scharry = cv2.Scharr(img,cv2.CV_64F,0,1)
scharry = cv2.convertScaleAbs(scharry)
scharrxy = cv2.addWeighted(scharrx,0.5,scharry,0.5,0)
laplacian = cv2.Laplacian(img,cv2.CV_64F)
laplacian = cv2.convertScaleAbs(laplacian)
#双图一起展示
show_img = np.hstack((img,sobelxy,scharrxy,laplacian))
cv_show("show",show_img)
Canny边缘检测
- 使用高斯滤波器,以平滑图像,滤除噪声
- 计算图像中每个像素点的梯度强度和方向
- 应用非极大值(Non-Maximun Suppression)抑制,以消除边缘检测带来的杂散响应、
- 应用双阈值(Double-Threshold)检测来确定真实的和潜在的边缘
- 通过抑制孤立的弱边缘最终完成边缘检测
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("lena.png",cv2.IMREAD_GRAYSCALE)
v1 = cv2.Canny(img,80,150)#minVal maxVal
v2 = cv2.Canny(img,50,100)
#双图一起展示
show_img = np.hstack((v1,v2))
cv_show("show",show_img)
图像金字塔
- 将图像在每个方向扩大为原来的两倍,新增的行和列用 0 填充
- 使用先前同样的内核(乘以4)与放大后的图像卷积,获得 近似值
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("am.png")
print(img.shape)
up = cv2.pyrUp(img)
print(up.shape)
down = cv2.pyrDown(img)
print(down.shape)
(380, 289, 3)
(760, 578, 3)
(190, 145, 3)
拉普拉斯金字塔
Gi是原始图像
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("am.png")
down = cv2.pyrDown(img)
down_up = cv2.pyrUp(down)
down_up = cv2.resize(down_up,(img.shape[1],img.shape[0]))
l = img - down_up
#双图一起展示
show_img = np.hstack((img,l))
cv_show("show",show_img)
图像轮廓
cv2.findContours(img,mode,method)
mode:轮廓检索模式
- RETR_EXTERNAL : 只检索最外面的轮廓
- RETR_LIST:检索所有的轮廓,并将其保存到一条链表中
- RETR_CCOMP:检索所有轮廓,并将他们组织为两层:顶层是各部分的外部边界,第二层是空洞的边界
- RETR_TREE : 检索所有的轮廓,并重构嵌套轮廓的整个层次(常用)
method:轮廓逼近方法
- CHAIN_APPROX_NONE:以 Freeman 链码的方式轮廓,所有其他方法输出多边形(顶点的序列)
- CHAIN_APPROX_SIMPLE: 压缩水平的、垂直的和斜的部分,也就是,函数只保留他们的终点部分
为了更高的准确率,使用二值图图像
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#二值图图像
img = cv2.imread("car.png")
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
cv_show("img",img)
cv_show("thresh",thresh)
[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-7rYswLDv-1689045274416)(C:\Users\Administrator\Desktop\photo\image-20230705202358871.png)]
contours,hierarchy = cv2.findContours(thresh,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
contours :轮廓集合
绘制轮廓:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#二值图图像
img = cv2.imread("car.png")
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
# cv_show("img",img)
# cv_show("thresh",thresh)
#轮廓集合
contours,hierarchy = cv2.findContours(thresh,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
#传入 (绘制图像,轮廓,轮廓索引,颜色模式(b,g,r),线条厚度)
#注意需要copy,要不原图会变
draw_img = img.copy()
res = cv2.drawContours(draw_img,contours,-1,(0,255,0),1)#-1 表示所有
cv_show("res",res)
draw_img = img.copy()
res = cv2.drawContours(draw_img,contours,0,(0,255,0),1)
cv_show("res",res)#原图
用下面的图像好找轮廓一点
轮廓特征:
计算面积、周长:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("contours.png")
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
# cv_show("img",img)
# cv_show("thresh",thresh)
contours,hierarchy = cv2.findContours(thresh,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
#计算第一个轮廓的面积
cnt = contours[0]
print(cv2.contourArea(cnt))#6332.5
#计算周长,True表示闭合的
print(cv2.arcLength(cnt,True))#379.32084918022156
轮廓近似:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("contours2.png")
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
# cv_show("img",img)
# cv_show("thresh",thresh)
contours,hierarchy = cv2.findContours(thresh,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
#轮廓
cnt = contours[0]
draw_img = img.copy()
res = cv2.drawContours(draw_img,[cnt],-1,(0,0,255),2)
# cv_show("res",res)
#轮廓近似
epsilon = 0.1*cv2.arcLength(cnt,True)
approx = cv2.approxPolyDP(cnt,epsilon,True)
draw_img = img.copy()
res2 = cv2.drawContours(draw_img,[approx],-1,(0,0,255),2)
# cv_show("res",res)
#双图一起展示
show_img = np.hstack((img,res,res2))
cv_show("show",show_img)
边界矩形:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("contours.png")
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
# cv_show("img",img)
# cv_show("thresh",thresh)
contours,hierarchy = cv2.findContours(thresh,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
#轮廓
cnt = contours[0]
draw_img = img.copy()
res = cv2.drawContours(draw_img,[cnt],-1,(0,0,255),2)
# cv_show("res",res)
#边界矩形
x,y,w,h = cv2.boundingRect(cnt)
img = cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,0),2)
#轮廓面积与比边界矩形比
area = cv2.contourArea(cnt)
rect_area = w * h
extent = float(area)/rect_area
print("轮廓面积与比边界矩形比",extent)#0.5098631239935588
#双图一起展示
show_img = np.hstack((res,img))
cv_show("show",show_img)
外接圆:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("contours.png")
gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(gray,127,255,cv2.THRESH_BINARY)
# cv_show("img",img)
# cv_show("thresh",thresh)
contours,hierarchy = cv2.findContours(thresh,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
#轮廓
cnt = contours[0]
draw_img = img.copy()
res = cv2.drawContours(draw_img,[cnt],-1,(0,0,255),2)
# cv_show("res",res)
#外接圆
(x,y),radius = cv2.minEnclosingCircle(cnt)
center = (int(x),int(y))
radius = int(radius)
img = cv2.circle(img,center,radius,(0,255,0),2)
#双图一起展示
show_img = np.hstack((res,img))
cv_show("show",show_img)
模版匹配
cv::TemplateMatchModes
模版匹配和卷积原理很像,模版在原图像上从原点开始滑动,计算模版与(图像被模版覆盖的地方)的差别程度,这个差别程度的计算方法在opencv里有6种,然后将每次计算的结果放入一个矩阵里,作为结果输出。假如原图形是AxB大小,而模版是axb大小,则输出结果的矩阵是(A-a+1)x(B-b+1)
- TM_SQDIFF:计算平方不同,计算出来的值越小,越相关
- TM_CCORR:计算相关性,计算出来的值越大,越相关
- TM_CCOEFF:计算相关系数,计算出来的值越大,越相关
- TM_SQDIFF_NORMED:计算归一化平方不同,计算出来的值越接近0,越相关
- TM_CCORR_NORMED:计算归一化相关性,计算出来的值越接近1,越相关
- TM_CCOEFF_NORMED:计算归一化相关系数,计算出来的值越接近1,越相关
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("lena.png")
template = cv2.imread("face.png")
h,w = template.shape[:2]
res = cv2.matchTemplate(img,template,cv2.TM_SQDIFF)
#返回最小值最大值,和其坐标位置
min_val,max_val,min_loc,max_loc = cv2.minMaxLoc(res)
print(min_val,max_val,min_loc,max_loc)
#0.0 0.39683234691619873 (187, 138) (74, 253)
bottom = (min_loc[0]+w,min_loc[1]+h)
cv2.rectangle(img,min_loc,bottom,255,2)
plt.subplot(121),plt.imshow(res,cmap = 'gray')
plt.xticks([]),plt.yticks([])#隐藏坐标轴
plt.subplot(122),plt.imshow(img,cmap = 'gray')
plt.xticks([]),plt.yticks([])
plt.suptitle("hh")
plt.show()
# cv_show("img",img)
匹配多个对象:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("mario.png")
img_gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
template = cv2.imread("coin.png",0)
h,w = template.shape[:2]
res = cv2.matchTemplate(img_gray,template,cv2.TM_CCOEFF_NORMED)
threshold = 0.9
#取匹配程度大于90%的坐标
loc = np.where(res>=threshold)
for pt in zip(*loc[::-1]):# *号表示可选参数
cv2.rectangle(img,pt,(pt[0]+w,pt[1]+h),(0,0,255),1)
cv_show("img",img)
直方图
cv2.calcHist ( images, channels, mask, histSize,ranges )
- images : 原图像图像格式为uint8或float32 。当传入函数时应该用中括号 [] 括来,例如 [img]
- channels : 同样用中括号括来它会告函数我们统幅图像的直方图。如果入图像是灰度图它的值就是 [0] ,如果是彩色图像的传入的参数可以是 [0][1][2] 它们分别对应着BGR
- mask : 掩模图像。统整幅图像的直方图就把它为 None。但是如果你想统图像某一分的直方图的你就制作一个掩膜图像并使用它
- hisSize:BIN的树木。也应该用中括号括来
- ranges:像素值范围常为 [0256]
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("lena.png",0)
hist = cv2.calcHist([img],[0],None,[256],[0,256])
print(hist.shape)#(256, 1)
plt.hist(img.ravel(),256)
plt.show()
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#
img = cv2.imread("lena.png")
color = ('b','g','r')
for i,col in enumerate(color):
histr = cv2.calcHist([img],[i],None,[256],[0,256])
plt.plot(histr,color = col)
plt.xlim([0,256])
plt.show()
mash操作:
#coding:utf-8
import cv2
import numpy as np
import matplotlib.pyplot as plt
def cv_show(name,img):
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
#原图
img = cv2.imread("lena.png",0)
#创建mast
mask = np.zeros(img.shape[:2],np.uint8)
mask[150:350,150:350] = 255
#原图结合mast
mask_img = cv2.bitwise_and(img,img,mask = mask)#与操作
hist_full = cv2.calcHist([img],[0],None,[256],[0,256])
hist_mask = cv2.calcHist([img],[0],mask,[256],[0,256])
plt.subplot(221),plt.imshow(img,'gray')
plt.subplot(222),plt.imshow(mask,'gray')
plt.subplot(223),plt.imshow(mask_img,'gray')
plt.subplot(224),plt.plot(hist_full),plt.plot(hist_mask)
plt.xlim([0,256])
plt.show()
直方图均衡化:
img = cv2.imread("lena.png",0)
plt.hist(img.ravel(),256)
plt.show()
#原图
img = cv2.imread("lena.png",0)
equ = cv2.equalizeHist(img)
plt.hist(equ.ravel(),256)
plt.show()
#双图一起展示
show_img = np.hstack((img,equ))
cv_show("show",show_img)
自适应直方图均衡化
img = cv2.imread("lena.png",0)
clahe = cv2.createCLAHE(clipLimit=2.0,tileGridSize=(8,8))
res_clahe = clahe.apply(img)
res = np.hstack((img,res_clahe))
cv_show("res",res)
可以多保留一些细节。
傅里叶变换
傅里叶变换的作用:
- 高频:变换剧烈的灰度分量,例如边界
- 低频:变化缓慢的灰度分量,例如一片大海
滤波:
-
低通滤波器:只保留低频,会使得图像模糊
-
高通滤波器:只保留高频,会使得图像细节增强
-
opencv 中主要就是 cv2.dft() 和 cv2.idft() ,输入图像需要先转换成 np.float32 格式。
-
得到的结果中频率为0的部分会在左上角,通常要转换到中心位置,可以通过shift变换来实现。
-
cv2.dit() 返回的结果是双通道的(实部,虚部),通常还需要转换成图像格式才能展示(0,255)。
#原图
img = cv2.imread("lena.png",0)
img_float32 = np.float32(img)
dft = cv2.dft(img_float32,flags = cv2.DFT_COMPLEX_OUTPUT)
dft_shift = np.fft.fftshift(dft)
#得到灰度图能表示的形式
magnitude_spectrum = 20*np.log(cv2.magnitude(dft_shift[:,:,0],dft_shift[:,:,1]))
plt.subplot(121),plt.imshow(img,cmap = 'gray')
plt.title('Input Image'), plt.xticks([]), plt.yticks([])
plt.subplot(122),plt.imshow(magnitude_spectrum,cmap = 'gray')
plt.title('Magnitude Spectrum'), plt.xticks([]), plt.yticks([])
plt.show()
低通滤波:只保留低频,会使得图像模糊
#原图
img = cv2.imread("lena.png",0)
img_float32 = np.float32(img)
dft = cv2.dft(img_float32,flags = cv2.DFT_COMPLEX_OUTPUT)
dft_shift = np.fft.fftshift(dft)
rows,cols = img.shape
crow,ccol = int(rows/2),int(cols/2)#中心位置
#低通滤波
mask = np.zeros((rows,cols,2),np.uint8)
mask[crow-30:crow+30,ccol-30:ccol+30] = 1
#IDFT
fshift = dft_shift*mask
f_ishift = np.fft.ifftshift(fshift)
img_back = cv2.idft(f_ishift)
img_back = cv2.magnitude(img_back[:,:,0],img_back[:,:,1])
plt.subplot(121),plt.imshow(img,cmap = 'gray')
plt.title('Input Image'), plt.xticks([]), plt.yticks([])
plt.subplot(122),plt.imshow(img_back,cmap = 'gray')
plt.title('Magnitude Spectrum'), plt.xticks([]), plt.yticks([])
plt.show()
高通滤波 : 只保留高频,会使得图像细节增强
#原图
img = cv2.imread("lena.png",0)
img_float32 = np.float32(img)
dft = cv2.dft(img_float32,flags = cv2.DFT_COMPLEX_OUTPUT)
dft_shift = np.fft.fftshift(dft)
rows,cols = img.shape
crow,ccol = int(rows/2),int(cols/2)#中心位置
#高通滤波
mask = np.ones((rows,cols,2),np.uint8)
mask[crow-30:crow+30,ccol-30:ccol+30] = 0
#IDFT
fshift = dft_shift*mask
f_ishift = np.fft.ifftshift(fshift)
img_back = cv2.idft(f_ishift)
img_back = cv2.magnitude(img_back[:,:,0],img_back[:,:,1])
plt.subplot(121),plt.imshow(img,cmap = 'gray')
plt.title('Input Image'), plt.xticks([]), plt.yticks([])
plt.subplot(122),plt.imshow(img_back,cmap = 'gray')
plt.title('Magnitude Spectrum'), plt.xticks([]), plt.yticks([])
plt.show()