OpenCV的轮廓检测和阈值处理综合运用

1.轮廓查找和绘制轮廓

https://mydreamambitious.blog.csdn.net/article/details/125339604

---

2.Opencv中的多变形的逼近和凸包

https://mydreamambitious.blog.csdn.net/article/details/125340595

---

3.Opencv全局二值化和局部二值化

https://mydreamambitious.blog.csdn.net/article/details/125249121

---

4.什么是图像Mask

https://www.cnblogs.com/skyfsm/p/6894685.html

---

5.图像的基本转转（翻转等）

https://mydreamambitious.blog.csdn.net/article/details/125428402

---

6.Opencv中计算轮廓的周长和面积

https://mydreamambitious.blog.csdn.net/article/details/125340081

---

7.代码实战

（1）对图像进行进行翻转和缩放处理

#对图片进行翻转，Opencv读取的图片和我们正常的显示方式相反
    img=cv2.flip(src=frame,flipCode=2)
    img=cv2.resize(src=img,dsize=(600,600))
1
2
3

（2）去除背景

#去除背景
bgsegment=cv2.createBackgroundSubtractorMOG2()
#转换为灰度图
    img=cv2.cvtColor(src=img,code=cv2.COLOR_BGR2GRAY)
    img = bgsegment.apply(img)
1
2
3
4
5

（3）高斯处理和二值化处理

#对图像进行高斯噪声处理
    gauss=cv2.GaussianBlur(src=img,ksize=(3,3),sigmaX=0)

    #图像二值化
    ok, dst = cv2.threshold(src=gauss, thresh=70, maxval=255, type=cv2.THRESH_BINARY)

1
2
3
4
5
6

（4）对图像腐蚀和膨胀操作

# 首先获取卷积核
    kernel = cv2.getStructuringElement(shape=cv2.MORPH_RECT, ksize=(3, 3))
    #对图像进行腐蚀
    erode=cv2.erode(src=dst,kernel=kernel,iterations=1)
    # canny=cv2.Canny(image=erode,threshold1=50,threshold2=100)
    #对图像进行膨胀
    dilate=cv2.dilate(src=erode,kernel=kernel,iterations=1)
1
2
3
4
5
6
7

（5）轮廓查找

 #查找轮廓
    contours,hierarchy=cv2.findContours(image=dilate,mode=cv2.RETR_TREE,
                                        method=cv2.CHAIN_APPROX_SIMPLE)
1
2
3

（6）找出面积最大的轮廓并进行轮廓的绘制

#找到其中最大的轮廓
    index=0
    maxArea=0
    length=len(contours)
    if length>0:
        for i in range(length):
            #计算轮廓的面积
            area=cv2.contourArea(contours[i])
            if area>maxArea:
                maxArea=area
                index=i
        # print('len(contours)= {} ; index={}'.format(len(contours),index))
        #首先绘制图像的形状
        image=np.zeros(shape=frame.shape,dtype=np.uint8)
        res=contours[index]
        cv2.drawContours(image=image,contours=[res],
                         contourIdx=0,color=(0,255,0),thickness=2)
        #绘制凸包
        hull=cv2.convexHull(points=res)
        cv2.drawContours(image=image,contours=[hull],contourIdx=0,
                         color=(0,0,255),thickness=3)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

（7）代码综合

import os
import cv2
import numpy as np

cap=cv2.VideoCapture(0)
#去除背景
bgsegment=cv2.createBackgroundSubtractorMOG2()
while cap.isOpened():
    OK,frame=cap.read()
    if OK==False:
        break
    #对图片进行翻转，Opencv读取的图片和我们正常的显示方式相反
    img=cv2.flip(src=frame,flipCode=2)
    img=cv2.resize(src=img,dsize=(600,600))
    #转换为灰度图
    img=cv2.cvtColor(src=img,code=cv2.COLOR_BGR2GRAY)
    img = bgsegment.apply(img)

    #对图像进行高斯噪声处理
    gauss=cv2.GaussianBlur(src=img,ksize=(3,3),sigmaX=0)

    #图像二值化
    ok, dst = cv2.threshold(src=gauss, thresh=70, maxval=255, type=cv2.THRESH_BINARY)

    # 首先获取卷积核
    kernel = cv2.getStructuringElement(shape=cv2.MORPH_RECT, ksize=(3, 3))
    #对图像进行腐蚀
    erode=cv2.erode(src=dst,kernel=kernel,iterations=1)
    # canny=cv2.Canny(image=erode,threshold1=50,threshold2=100)
    #对图像进行膨胀
    dilate=cv2.dilate(src=erode,kernel=kernel,iterations=1)

    #查找轮廓
    contours,hierarchy=cv2.findContours(image=dilate,mode=cv2.RETR_TREE,
                                        method=cv2.CHAIN_APPROX_SIMPLE)
    #找到其中最大的轮廓
    index=0
    maxArea=0
    length=len(contours)
    if length>0:
        for i in range(length):
            #计算轮廓的面积
            area=cv2.contourArea(contours[i])
            if area>maxArea:
                maxArea=area
                index=i
        # print('len(contours)= {} ; index={}'.format(len(contours),index))
        #首先绘制图像的形状
        image=np.zeros(shape=frame.shape,dtype=np.uint8)
        res=contours[index]
        cv2.drawContours(image=image,contours=[res],
                         contourIdx=0,color=(0,255,0),thickness=2)
        #绘制凸包
        hull=cv2.convexHull(points=res)
        cv2.drawContours(image=image,contours=[hull],contourIdx=0,
                         color=(0,0,255),thickness=3)

    cv2.imshow('image',image)
    if cv2.waitKey(1)&0xFF==27:
        break
cv2.destroyAllWindows()


if __name__ == '__main__':
    print('Pycharm')

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66