2. OpenCV-Python——阈值平滑处理形态学操作

Posted 2020-12-23 红叶楠木

一、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

• ret：  返回阈值的数值

• 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；cv2.THRESH_BINARY_INV； cv2.THRESH_TRUNC； cv2.THRESH_TOZERO；cv2.THRESH_TOZERO_INV

• cv2.THRESH_BINARY 超过阈值部分取maxval（最大值），否则取0

• cv2.THRESH_BINARY_INV THRESH_BINARY的反转

• 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 THRESH_TOZERO的反转

 1 # *******************阈值**********************开始
 2 import cv2
 3 import matplotlib.pyplot as plt
 4 
 5 img = cv2.imread(\'cat.jpg\')                        # 读取图像
 6 img_gray = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)    # 灰度化
 7 
 8 ret, thresh1 = cv2.threshold(img_gray, 127, 255, cv2.THRESH_BINARY)
 9 ret, thresh2 = cv2.threshold(img_gray, 127, 255, cv2.THRESH_BINARY_INV)
10 ret, thresh3 = cv2.threshold(img_gray, 127, 255, cv2.THRESH_TRUNC)
11 ret, thresh4 = cv2.threshold(img_gray, 127, 255, cv2.THRESH_TOZERO)
12 ret, thresh5 = cv2.threshold(img_gray, 127, 255, cv2.THRESH_TOZERO_INV)
13 
14 titles = [\'Original Image\', \'BINARY\', \'BINARY_INV\', \'TRUNC\', \'TOZERO\', \'TOZERO_INV\']
15 images = [img, thresh1, thresh2, thresh3, thresh4, thresh5]
16 
17 for i in range(6):
18     plt.subplot(2, 3, i + 1), plt.imshow(images[i], \'gray\')
19     plt.title(titles[i])
20     plt.xticks([]), plt.yticks([])
21 plt.show()
22 # *******************阈值**********************结束

二、平滑处理

 1 # *******************平滑处理**********************开始
 2 import cv2
 3 import numpy as np
 4 # import matplotlib.pyplot as plt
 5 
 6 img = cv2.imread(\'lenaNoise.png\')
 7 
 8 # cv2.imshow(\'img\', img)
 9 # cv2.waitKey(0)
10 # cv2.destroyAllWindows()
11 
12 # 均值滤波
13 # 简单的平均卷积操作
14 blur = cv2.blur(img, (3, 3))  # 3*3的卷积核
15 
16 # cv2.imshow(\'blur\', blur)
17 # cv2.waitKey(0)
18 # cv2.destroyAllWindows()
19 
20 # 方框滤波
21 # 基本和均值一样，可以选择归一化，不做归一化容易越界溢出
22 box = cv2.boxFilter(img,-1,(3,3), normalize=True)  # -1表示在颜色通道上保持一致
23 
24 # cv2.imshow(\'box\', box)
25 # cv2.waitKey(0)
26 # cv2.destroyAllWindows()
27 
28 # 高斯滤波
29 # 高斯模糊的卷积核里的数值是满足高斯分布，相当于更重视中间的
30 aussian = cv2.GaussianBlur(img, (5, 5), 1)
31 
32 # cv2.imshow(\'aussian\', aussian)
33 # cv2.waitKey(0)
34 # cv2.destroyAllWindows()
35 
36 # 中值滤波
37 # 相当于用中值代替
38 median = cv2.medianBlur(img, 5)  # 中值滤波
39 
40 # cv2.imshow(\'median\', median)
41 # cv2.waitKey(0)
42 # cv2.destroyAllWindows()
43 
44 # 展示所有的滤波结果
45 res = np.hstack((blur,aussian,median))  # vstack为垂直方向
46 #print (res)
47 cv2.imshow(\'median vs average\', res)
48 cv2.waitKey(0)
49 cv2.destroyAllWindows()
50 # *******************平滑处理**********************结束

三、形态学操作

1、腐蚀操作

 1 # *******************形态学-腐蚀**********************开始
 2 import cv2
 3 import numpy as np
 4 
 5 img = cv2.imread(\'dige.png\')
 6 
 7 cv2.imshow(\'img\', img)
 8 cv2.waitKey(0)
 9 cv2.destroyAllWindows()
10 
11 kernel = np.ones((3,3),np.uint8)
12 erosion = cv2.erode(img,kernel,iterations = 1) # irerations迭代次数
13 
14 cv2.imshow(\'erosion\', erosion)
15 cv2.waitKey(0)
16 cv2.destroyAllWindows()
17 # *******************形态学-腐蚀**********************结束

2、膨胀操作

 1 # *******************形态学-膨胀**********************开始
 2 import cv2
 3 import numpy as np
 4 
 5 img = cv2.imread(\'dige.png\')
 6 cv2.imshow(\'img\', img)
 7 cv2.waitKey(0)
 8 cv2.destroyAllWindows()
 9 
10 # 先腐蚀
11 kernel = np.ones((3,3),np.uint8)
12 erosion = cv2.erode(img,kernel,iterations = 1)
13 # 后膨胀
14 kernel = np.ones((3,3),np.uint8)
15 dige_dilate = cv2.dilate(erosion,kernel,iterations = 1)
16 
17 cv2.imshow(\'dilate\', dige_dilate)
18 cv2.waitKey(0)
19 cv2.destroyAllWindows()
20 # *******************形态学-膨胀**********************结束

3、开运算与闭运算

 1 # *******************形态学-开运算与闭运算**********************开始
 2 import cv2
 3 import numpy as np
 4 
 5 # 开：先腐蚀，再膨胀
 6 img = cv2.imread(\'dige.png\')
 7 
 8 kernel = np.ones((5,5),np.uint8)
 9 opening = cv2.morphologyEx(img, cv2.MORPH_OPEN, kernel)
10 
11 cv2.imshow(\'opening\', opening)
12 cv2.waitKey(0)
13 cv2.destroyAllWindows()
14 
15 # 闭：先膨胀，再腐蚀
16 img = cv2.imread(\'dige.png\')
17 
18 kernel = np.ones((5,5),np.uint8)
19 closing = cv2.morphologyEx(img, cv2.MORPH_CLOSE, kernel)
20 
21 cv2.imshow(\'closing\', closing)
22 cv2.waitKey(0)
23 cv2.destroyAllWindows()
24 # *******************形态学-开运算与闭运算**********************结束

4、梯度运算

 1 # *******************形态学-梯度运算**********************开始
 2 import cv2
 3 import numpy as np
 4 
 5 # 梯度=膨胀-腐蚀
 6 pie = cv2.imread(\'pie.png\')
 7 kernel = np.ones((7,7),np.uint8)
 8 #---------------------------------------------------------
 9 dilate = cv2.dilate(pie,kernel,iterations = 5)
10 erosion = cv2.erode(pie,kernel,iterations = 5)
11 
12 res = np.hstack((dilate,erosion)) # 显示分别膨胀和腐蚀的结果
13 cv2.imshow(\'res\', res)
14 cv2.waitKey(0)
15 cv2.destroyAllWindows()
16 
17 #---------------------------------------------------------
18 gradient = cv2.morphologyEx(pie, cv2.MORPH_GRADIENT, kernel) # 梯度运算
19 cv2.imshow(\'gradient\', gradient)
20 cv2.waitKey(0)
21 cv2.destroyAllWindows()
22 # *******************形态学-梯度运算**********************结束

5、礼帽与黑帽

• 礼帽 = 原始输入-开运算结果
• 黑帽 = 闭运算-原始输入

 1 # *******************形态学-礼帽与黑帽**********************开始
 2 import cv2
 3 import numpy as np
 4 
 5 #-------------------------------------------------------
 6 #礼帽
 7 img = cv2.imread(\'dige.png\')
 8 cv2.imshow(\'image\',img)
 9 kernel = np.ones((7,7),np.uint8)
10 tophat = cv2.morphologyEx(img, cv2.MORPH_TOPHAT, kernel)
11 cv2.imshow(\'tophat\', tophat)
12 cv2.waitKey(0)
13 cv2.destroyAllWindows()
14 
15 #-------------------------------------------------------
16 #黑帽
17 img = cv2.imread(\'dige.png\')
18 kernel = np.ones((7,7),np.uint8)
19 blackhat  = cv2.morphologyEx(img,cv2.MORPH_BLACKHAT, kernel)
20 cv2.imshow(\'blackhat \', blackhat )
21 cv2.waitKey(0)
22 cv2.destroyAllWindows()
23 # *******************形态学-礼帽与黑帽**********************结束