In [2]:
from skimage import data
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=data.checkerboard()
用边长为5的正方形滤波器进行膨胀滤波：
In [3]:
dst1=sm.dilation(img,sm.square(5))
/tmp/ipykernel_463/1982442659.py:1: FutureWarning: `square` is deprecated since version 0.25 and will be removed in version 0.27. Use `skimage.morphology.footprint_rectangle` instead. dst1=sm.dilation(img,sm.square(5))
用边长为15的正方形滤波器进行膨胀滤波：
In [4]:
dst2=sm.dilation(img,sm.square(15))
/tmp/ipykernel_463/2403410989.py:1: FutureWarning: `square` is deprecated since version 0.25 and will be removed in version 0.27. Use `skimage.morphology.footprint_rectangle` instead. dst2=sm.dilation(img,sm.square(15))
In [5]:
plt.figure('morphology',figsize=(8,8))
plt.subplot(131)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
Out[5]:
<matplotlib.image.AxesImage at 0x7f92ac8af0e0>
In [6]:
plt.subplot(132)
plt.title('morphological image')
plt.imshow(dst1,plt.cm.gray)
Out[6]:
<matplotlib.image.AxesImage at 0x7f92a4433fe0>
In [7]:
plt.subplot(133)
plt.title('morphological image')
plt.imshow(dst2,plt.cm.gray)
Out[7]:
<matplotlib.image.AxesImage at 0x7f92a4493830>
分别用边长为5或15的正方形滤波器对棋盘图片进行膨胀操作，结果如上图。
可见滤波器的大小，对操作结果的影响非常大。一般设置为奇数。
除了正方形的滤波器外，滤波器的形状还有一些，现列举如下：
- morphology.square: 正方形
- morphology.disk: 平面圆形
- morphology.ball: 球形
- morphology.cube: 立方体形
- morphology.diamond: 钻石形
- morphology.rectangle: 矩形
- morphology.star: 星形
- morphology.octagon: 八角形
- morphology.octahedron： 八面体
注意，如果处理图像为二值图像（只有0和1两个值），则可以调用：
skimage.morphology.binary_dilation(image, selem=None）用此函数比处理灰度图像要快。
腐蚀（erosion)
函数：
skimage.morphology.erosion(image, selem=None)selem 表示结构元素，用于设定局部区域的形状和大小。
腐蚀是和膨胀相反的操作，将0值扩充到邻近像素。扩大黑色部分，减小白色部分。 可用来提取骨干信息，去掉毛刺，去掉孤立的像素。
In [8]:
from skimage import data
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=data.checkerboard()
用边长为5的正方形滤波器进行膨胀滤波：
In [9]:
dst1=sm.erosion(img,sm.square(5))
/tmp/ipykernel_463/1850116054.py:1: FutureWarning: `square` is deprecated since version 0.25 and will be removed in version 0.27. Use `skimage.morphology.footprint_rectangle` instead. dst1=sm.erosion(img,sm.square(5))
用边长为25的正方形滤波器进行膨胀滤波：
In [10]:
dst2=sm.erosion(img,sm.square(25))
/tmp/ipykernel_463/108573180.py:1: FutureWarning: `square` is deprecated since version 0.25 and will be removed in version 0.27. Use `skimage.morphology.footprint_rectangle` instead. dst2=sm.erosion(img,sm.square(25))
In [11]:
plt.figure('morphology',figsize=(8,8))
plt.subplot(131)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
Out[11]:
<matplotlib.image.AxesImage at 0x7f92a44e16a0>
In [12]:
plt.subplot(132)
plt.title('morphological image')
plt.imshow(dst1,plt.cm.gray)
Out[12]:
<matplotlib.image.AxesImage at 0x7f92a46b0050>
In [13]:
plt.subplot(133)
plt.title('morphological image')
plt.imshow(dst2,plt.cm.gray)
Out[13]:
<matplotlib.image.AxesImage at 0x7f92a4713f50>
In [14]:
from skimage import io,color
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(io.imread('/data/demo/sec09_xzv8.png'))
用边长为9的圆形滤波器进行膨胀滤波：
In [15]:
dst=sm.opening(img,sm.disk(9))
In [16]:
plt.figure('morphology',figsize=(8,8))
plt.subplot(121)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.axis('off')
Out[16]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [17]:
plt.subplot(122)
plt.title('morphological image')
plt.imshow(dst,plt.cm.gray)
plt.axis('off')
Out[17]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [18]:
from skimage import io,color
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(io.imread('/data/demo/sec09_xzv8.png'))
用边长为5的圆形滤波器进行膨胀滤波：
In [19]:
dst=sm.closing(img,sm.disk(9))
In [20]:
plt.figure('morphology',figsize=(8,8))
plt.subplot(121)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.axis('off')
Out[20]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [21]:
plt.subplot(122)
plt.title('morphological image')
plt.imshow(dst,plt.cm.gray)
plt.axis('off')
Out[21]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [22]:
from skimage import io,color
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(io.imread('/data/demo/sec09_xzv8.png'))
dst=sm.white_tophat(img,sm.square(21))
/tmp/ipykernel_463/4003695167.py:5: FutureWarning: `square` is deprecated since version 0.25 and will be removed in version 0.27. Use `skimage.morphology.footprint_rectangle` instead. dst=sm.white_tophat(img,sm.square(21))
In [23]:
plt.figure('morphology',figsize=(8,8))
plt.subplot(121)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.axis('off')
Out[23]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [24]:
plt.subplot(122)
plt.title('morphological image')
plt.imshow(dst,plt.cm.gray)
plt.axis('off')
Out[24]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [25]:
from skimage import io,color
import skimage.morphology as sm
import matplotlib.pyplot as plt
img=color.rgb2gray(io.imread('/data/demo/sec09_xzv8.png'))
dst=sm.black_tophat(img,sm.square(21))
/tmp/ipykernel_463/4184403036.py:5: FutureWarning: `square` is deprecated since version 0.25 and will be removed in version 0.27. Use `skimage.morphology.footprint_rectangle` instead. dst=sm.black_tophat(img,sm.square(21))
In [26]:
plt.figure('morphology',figsize=(8,8))
plt.subplot(121)
plt.title('origin image')
plt.imshow(img,plt.cm.gray)
plt.axis('off')
Out[26]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
In [27]:
plt.subplot(122)
plt.title('morphological image')
plt.imshow(dst,plt.cm.gray)
plt.axis('off')
Out[27]:
(np.float64(-0.5), np.float64(511.5), np.float64(511.5), np.float64(-0.5))
