正文

数字图像处理的基本步骤

  • 图像获取
  • 图像增强：对图像进行某种操作， 使结果在特定应用中比原图像更为合适的过程
  • 图像复原：改善图像外观的一个领域， 倾向于以图像退化的数学或概率模型为基础
  • 彩色图像处理
  • 小波：以不同分辨率来表示图像的基础
  • 形态学处理：提取图像中用于表示和描述形状的成分的处理工具
  • 分割：将一幅图像划分为各个组成部分和目标
  • 特征提取
  • 图像模式分类：根据目标特征描述子对目标赋予标记的过程

Tutorial 22 - Reading images into Python

Reading Images into Python

  • scikit-image: pip install scikit-image
  • opencv: pip install opencv-python
  • Pillow: PIL - does not import images as numpy array. You can convert using numpy.asarray(img)

使用 skimage 读取图像

from skimage import io

img = io.imread('images/Osteosarcoma_01.tif')
img.shape  # y, x, c, dtype=uint8
(1104, 1376, 3)
  • 转成 float 格式
from skimage import io, img_as_float

img2 = img_as_float(img)

不要使用 .astype(np.float)

import numpy as np

img3 = img.astype(np.float64)

转换成uint8格式的图像

from skimage import io, img_as_ubyte

img_8bit = img_as_ubyte(img2)

png

使用 OpenCV 读取图像

import cv2

img_cv2 =  cv2.imread('images/Osteosarcoma_01.tif')  # 彩色图像
grey_img = cv2.imread('images/Osteosarcoma_01.tif', 0)  # 灰度图像
color_img = cv2.imread('images/Osteosarcoma_01.tif', 1)  # 彩色图像
img_opencv = cv2.cvtColor(color_img, cv2.COLOR_BGR2RGB)  # 将图像转为 RGB 格式

OpenCV 默认使用 BGR 颜色空间, skimage 默认使用 RGB 颜色空间

png

Tutorial 23 - Reading proprietary images in python

  • 使用 Python 读取专有图像文件
import tifffile

img = tifffile.imread('images/Osteosarcoma_01.tif')
from skimage import io

img2 = io.imread('images/Osteosarcoma_01.tif', as_gray=False)

  • tif 还有 3D image 格式 和 Time series 格式

  • 还可以读取 czi(显微镜图像格式) 格式 (Time series, scenes, channels, y, x, z, RGB)

好专业的图像格式...

# Let us extract only relevant pixels, all channels in x and y
img1 = img[0, 0, :, :, :, 0]  # 分离出高, 宽, 通道

# Next, let us extract each channel image.
img2 = img1[0, :, :]  # 第一个通道 Red
img3 = img1[0, :, :]  # 第二个通道 Green
img4 = img1[2, :, :]  # 第三个通道 Blue DAPI

Tutorial 24 - Saving images from python to your local drive

skimage

from skimage import io

img = io.imread('images/Osteosarcoma_01.tif')
from skimage import filters

gaussian_img = filters.gaussian(img, sigma=3)  # 此时图像会变成 float 形式
io.imsave('images/exported/saved_using_skimage.jpg', gaussian_img)
Lossy conversion from float64 to uint8. Range [0, 1]. Convert image to uint8 prior to saving to suppress this warning.

io.imsave('images/exported/saved_using_skimage.tif', gaussian_img)  # 打不开, 因为识别不出float格式的颜色
from skimage import img_as_ubyte

gaussian_img_8bit = img_as_ubyte(gaussian_img)
io.imsave('images/exported/saved_using_skimage_8bit.tif', gaussian_img_8bit)  # 此时可以打开

OpenCV

import cv2

# OpenCV并不会将浮点数映射到0-255之间, 而是直接取整, 所以会显示黑色的图片
cv2.imwrite('images/exported/saved_using_opencv.jpg', gaussian_img)
True

# OpenCV 默认输出 BGR 颜色模型, 如果直接输出, 颜色会有变化
cv2.imwrite('images/exported/saved_using_opencv_8bit.jpg', gaussian_img_8bit)
True

# 预处理后再输出
gaussian_img_8bit_RGB = cv2.cvtColor(gaussian_img_8bit, cv2.COLOR_BGR2RGB)
cv2.imwrite('images/exported/saved_using_opencv_8bit_RGB.jpg', gaussian_img_8bit_RGB)
True

matplotlib

from matplotlib import pyplot as plt

plt.imsave('images/exported/saved_using_pyplot.jpg', gaussian_img)

tifffile

import tifffile

tifffile.imwrite('images/exported/saved_using_tifffile.tiff', gaussian_img)

png

Tutorial 25 - Viewing 2d images in python

matplotlib

from skimage import io

img = io.imread('images/Osteosarcoma_01.tif')

io.imshow(img)
<matplotlib.image.AxesImage at 0x1d668f31a08>

png

import matplotlib.pyplot as plt

plt.imshow(img)
<matplotlib.image.AxesImage at 0x1d66951bbc8>

png

cmap: 用多种风格显示灰度图像:

pyplot cmap颜色 - 知乎 (zhihu.com)

png

img_gray = io.imread('images/Osteosarcoma_01.tif', as_gray=True)
plt.imshow(img_gray, cmap='hot')
<matplotlib.image.AxesImage at 0x1d6698518c8>

png

plt.imshow(img_gray, cmap='jet')
<matplotlib.image.AxesImage at 0x1d6698c9948>

png

plt.imshow(img_gray, cmap='Blues')
<matplotlib.image.AxesImage at 0x1d66b5018c8>

png

fig = plt.figure(figsize=(10, 10))

ax1 = fig.add_subplot(2,2,1)
ax1.imshow(img_gray, cmap='hot')
ax1.title.set_text('1st')

ax2 = fig.add_subplot(2,2,2)
ax2.imshow(img_gray, cmap='jet')
ax2.title.set_text('2nd')

ax3 = fig.add_subplot(2,2,3)
ax3.imshow(img_gray, cmap='gray')
ax3.title.set_text('3rd')

ax4 = fig.add_subplot(2,2,4)
ax4.imshow(img_gray, cmap='nipy_spectral')
ax4.title.set_text('4th')
plt.show()

png

OpenCV

import cv2

gray_img = cv2.imread("images/Osteosarcoma_01.tif", 0)
color_img = cv2.imread("images/Osteosarcoma_01.tif", 1)

# Use the function cv2.imshow() to display an image in a window. 
# First argument is the window name which is a string. second argument is our image. 

cv2.imshow("pic from skimage import", img)  # Shows weird colors as R and B channels are swapped
cv2.imshow("color pic from opencv", color_img)
cv2.imshow("gray pic from opencv", gray_img)

# Maintain output window until 
# user presses a key or 1000 ms (1s)
cv2.waitKey(0)          

# destroys all windows created
cv2.destroyAllWindows()

png

Tutorial 26 - Basic plotting in python using matplot.pyplot

教你怎么用 matplotlib 库

  • 画折线图
from matplotlib import pyplot as plt

x = [1, 2, 3, 4, 5]
y = [1, 4, 9, 16, 25]

plt.plot(x, y)
[<matplotlib.lines.Line2D at 0x22168454048>]

png

import numpy as np

a = np.array(x)
b = np.array(y)
plt.plot(a, b)
[<matplotlib.lines.Line2D at 0x22168e11148>]

png

  • 显示图片
import cv2

gray_img = cv2.imread('images/sandstone.tif', 0)

plt.imshow(gray_img, cmap='gray')
<matplotlib.image.AxesImage at 0x2216b0b5908>

png

  • 显示直方图
plt.hist(gray_img.flat, bins=100, range=(0, 150))
(array([  544.,    30.,    99.,    71.,   122.,    74.,   183.,    88.,
          274.,   170.,   394.,   253.,   594.,   341.,   806.,   507.,
         1145.,   626.,  1431.,   794.,  1808.,  1001.,  2091.,  1062.,
         2303.,  1184.,  2352.,  1178.,  2271.,  1113.,  2083.,  1021.,
         1830.,   803.,  1568.,   662.,  1367.,   680.,  1320.,   621.,
         1218.,   613.,  1211.,   608.,  1206.,   634.,  1245.,   614.,
         1325.,   686.,  1397.,   683.,  1371.,   715.,  1390.,   754.,
         1493.,   756.,  1630.,   828.,  1763.,   916.,  1849.,   975.,
         2028.,  1202.,  2575.,  1480.,  3536.,  2095.,  5313.,  3461.,
         8307.,  4992., 12007.,  6800., 15600.,  8629., 18104.,  9210.,
        18742.,  9243., 17429.,  8156., 14734.,  6413., 11317.,  4760.,
         8432.,  3439.,  5933.,  2508.,  4391.,  1920.,  3222.,  1417.,
         2677.,  1213.,  2205.,  1962.]),
 array([  0. ,   1.5,   3. ,   4.5,   6. ,   7.5,   9. ,  10.5,  12. ,
         13.5,  15. ,  16.5,  18. ,  19.5,  21. ,  22.5,  24. ,  25.5,
         27. ,  28.5,  30. ,  31.5,  33. ,  34.5,  36. ,  37.5,  39. ,
         40.5,  42. ,  43.5,  45. ,  46.5,  48. ,  49.5,  51. ,  52.5,
         54. ,  55.5,  57. ,  58.5,  60. ,  61.5,  63. ,  64.5,  66. ,
         67.5,  69. ,  70.5,  72. ,  73.5,  75. ,  76.5,  78. ,  79.5,
         81. ,  82.5,  84. ,  85.5,  87. ,  88.5,  90. ,  91.5,  93. ,
         94.5,  96. ,  97.5,  99. , 100.5, 102. , 103.5, 105. , 106.5,
        108. , 109.5, 111. , 112.5, 114. , 115.5, 117. , 118.5, 120. ,
        121.5, 123. , 124.5, 126. , 127.5, 129. , 130.5, 132. , 133.5,
        135. , 136.5, 138. , 139.5, 141. , 142.5, 144. , 145.5, 147. ,
        148.5, 150. ]),
 <BarContainer object of 100 artists>)

png

  • 显示图片样式
from matplotlib import pyplot as plt
import numpy as np

a = np.array([1, 2, 3, 4, 5])
b = np.array([1, 4, 9, 16, 25])

plt.plot(a, b, 'r--')
plt.axis([0, 6, 0, 50])
(0.0, 6.0, 0.0, 50.0)

png

  • 多种图形
from matplotlib import pyplot as plt

wells = ['well1', 'well2', 'well3', 'well4', 'well5']
cells = [80, 62, 88, 110, 90]

plt.bar(wells, cells)  # 条形图
plt.scatter(wells, cells)  # 散点图
plt.plot(wells, cells)  # 折线图
[<matplotlib.lines.Line2D at 0x2216c346dc8>]

png

from matplotlib import pyplot as plt

# Adding labels and annotations
wells = [1,2,3,4,5]
cells = [80, 62, 88, 110, 90]

plt.figure(figsize=(8, 8))
plt.bar(wells, cells)
plt.xlabel('Well #', fontsize=18, color='red')
plt.ylabel('# dead cells')
plt.title('Dead cells in each well')
plt.axis([1, 6, 60, 120])   #xmin, xmax, ymin, ymax
plt.grid(True)  # 显示网格
plt.show()

png

  • 更改坐标轴单位
from matplotlib import pyplot as plt

x = [1,2,3,4,5]
y = [10, 125, 1350, 11250, 100500]

plt.figure(figsize=(12, 6))

# linear
plt.subplot(121)
plt.plot(x, y)
plt.yscale('linear')
plt.title('linear')
plt.grid(True)

#Log
plt.subplot(122)
plt.plot(x, y)
plt.yscale('log')
plt.title('log')
plt.grid(True)

png

  • 显示多幅图片
 from matplotlib import pyplot as plt

wells = ['well1', 'well2', 'well3', 'well4', 'well5']
cells = [80, 62, 88, 110, 90]


#Initialize the plot and sublots
# Initialize the plot
fig = plt.figure(figsize=(16,6))
ax1 = fig.add_subplot(131)
ax1.set(title='vertical bar', xlabel='Well #', ylabel='# cells')
        
ax2 = fig.add_subplot(132)
ax1.set(title='horizontal bar', xlabel='Well #', ylabel='# cells')

ax3 = fig.add_subplot(133)

# Plot the data
ax1.bar(wells, cells)
ax2.barh(wells, cells)
ax3.plot(wells, cells)

plt.savefig("images/my_plot.jpg")  # Save plot
# Show the plot
plt.show()

png

Tutorial 27 - Using glob to read multiple files in python

Python模块之glob模块 - 知乎

使用 glob 遍历某个文件夹, 得到这个文件夹下面所有文件的名称

  • 一次性得到整个列表
import cv2
import glob

file_list = glob.glob('images/*.*')
file_list
['images\\Alloy_gradient.jpg',
 'images\\BSE.tif',
 'images\\bubbles.tif',
 'images\\cast_iron1.tif',
 'images\\cast_iron2.jpg',
 'images\\monalisa.jpg',
 'images\\Osteosarcoma_01.tif',
 'images\\Osteosarcoma_01_1sigma_blur.tif',
 'images\\Osteosarcoma_01_25Sigma_noise.tif',
 'images\\Osteosarcoma_01_2sigma_blur.tif',
 'images\\Osteosarcoma_01_8bit.ome.tiff',
 'images\\Osteosarcoma_01_8bit_salt_pepper.tif',
 'images\\Osteosarcoma_01_8bit_salt_pepper_cropped.tif',
 'images\\Osteosarcoma_01_small.tif',
 'images\\Osteosarcoma_01_transl.tif',
 'images\\Osteosarcoma_01_transl_rot.tif',
 'images\\sandstone.tif',
 'images\\sandstone_blur_2sigma.tif',
 'images\\sandstone_low_contrast.tif',
 'images\\scratch_time_series.tif',
 'images\\synthetic.jpg',
 'images\\Ti_powder.tif',
 'images\\Ti_powder_single.tif']
  • 通过 for 循环得到列表
my_list = []
path = 'images/*.*'
for file in glob.glob(path):
    print(file)
    a = cv2.imread(file)
    my_list.append(a)
images\Alloy_gradient.jpg
images\BSE.tif
images\bubbles.tif
images\cast_iron1.tif
images\cast_iron2.jpg
images\monalisa.jpg
images\Osteosarcoma_01.tif
images\Osteosarcoma_01_1sigma_blur.tif
images\Osteosarcoma_01_25Sigma_noise.tif
images\Osteosarcoma_01_2sigma_blur.tif
images\Osteosarcoma_01_8bit.ome.tiff
images\Osteosarcoma_01_8bit_salt_pepper.tif
images\Osteosarcoma_01_8bit_salt_pepper_cropped.tif
images\Osteosarcoma_01_small.tif
images\Osteosarcoma_01_transl.tif
images\Osteosarcoma_01_transl_rot.tif
images\sandstone.tif
images\sandstone_blur_2sigma.tif
images\sandstone_low_contrast.tif
images\scratch_time_series.tif
images\synthetic.jpg
images\Ti_powder.tif
images\Ti_powder_single.tif

Tutorial 28 - Using os.listdir to read multiple files

  • 使用 os.listdir
import os

path = 'images/'
print(os.listdir(path))

for image in os.listdir(path):
    print(image)
['Alloy_gradient.jpg', 'BSE.tif', 'bubbles.tif', 'cast_iron1.tif', 'cast_iron2.jpg', 'exported', 'monalisa.jpg', 'Osteosarcoma_01.tif', 'Osteosarcoma_01_1sigma_blur.tif', 'Osteosarcoma_01_25Sigma_noise.tif', 'Osteosarcoma_01_2sigma_blur.tif', 'Osteosarcoma_01_8bit.ome.tiff', 'Osteosarcoma_01_8bit_salt_pepper.tif', 'Osteosarcoma_01_8bit_salt_pepper_cropped.tif', 'Osteosarcoma_01_small.tif', 'Osteosarcoma_01_transl.tif', 'Osteosarcoma_01_transl_rot.tif', 'sandstone.tif', 'sandstone_blur_2sigma.tif', 'sandstone_low_contrast.tif', 'scratch_time_series.tif', 'synthetic.jpg', 'Ti_powder.tif', 'Ti_powder_single.tif']
Alloy_gradient.jpg
BSE.tif
bubbles.tif
cast_iron1.tif
cast_iron2.jpg
exported
monalisa.jpg
Osteosarcoma_01.tif
Osteosarcoma_01_1sigma_blur.tif
Osteosarcoma_01_25Sigma_noise.tif
Osteosarcoma_01_2sigma_blur.tif
Osteosarcoma_01_8bit.ome.tiff
Osteosarcoma_01_8bit_salt_pepper.tif
Osteosarcoma_01_8bit_salt_pepper_cropped.tif
Osteosarcoma_01_small.tif
Osteosarcoma_01_transl.tif
Osteosarcoma_01_transl_rot.tif
sandstone.tif
sandstone_blur_2sigma.tif
sandstone_low_contrast.tif
scratch_time_series.tif
synthetic.jpg
Ti_powder.tif
Ti_powder_single.tif
  • 使用 os.walk 遍历文件
import os

print(os.walk('.'))  # Nothing to see here as this is just a generator

for root, dirs, files in os.walk('.'):
    # root 表示当前正在访问的文件夹路径
    # dirs 表示该文件夹下的子目录名list
    # files 表示该文件夹下的文件list
    # 遍历文件
    for f in files:
         print(os.path.join(root, f))
<generator object walk at 0x000001AAF06408C8>
.\4.ipynb
.\Untitled.ipynb
.\Untitled1.ipynb
.\Untitled2.ipynb
.\Untitled3.ipynb
.\Untitled4.ipynb
.\Untitled5.ipynb
.\Untitled6.ipynb
.\Untitled7.ipynb
.\Untitled8.ipynb
.\Untitled9.ipynb
.\.ipynb_checkpoints\4-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled1-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled2-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled3-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled4-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled5-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled6-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled7-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled8-checkpoint.ipynb
.\.ipynb_checkpoints\Untitled9-checkpoint.ipynb
.\images\Alloy_gradient.jpg
.\images\BSE.tif
.\images\bubbles.tif
.\images\cast_iron1.tif
.\images\cast_iron2.jpg
.\images\monalisa.jpg
.\images\Osteosarcoma_01.tif
.\images\Osteosarcoma_01_1sigma_blur.tif
.\images\Osteosarcoma_01_25Sigma_noise.tif
.\images\Osteosarcoma_01_2sigma_blur.tif
.\images\Osteosarcoma_01_8bit.ome.tiff
.\images\Osteosarcoma_01_8bit_salt_pepper.tif
.\images\Osteosarcoma_01_8bit_salt_pepper_cropped.tif
.\images\Osteosarcoma_01_small.tif
.\images\Osteosarcoma_01_transl.tif
.\images\Osteosarcoma_01_transl_rot.tif
.\images\sandstone.tif
.\images\sandstone_blur_2sigma.tif
.\images\sandstone_low_contrast.tif
.\images\scratch_time_series.tif
.\images\synthetic.jpg
.\images\Ti_powder.tif
.\images\Ti_powder_single.tif
.\images\exported\saved_using_opencv.jpg
.\images\exported\saved_using_opencv.tif
.\images\exported\saved_using_opencv_8bit.jpg
.\images\exported\saved_using_opencv_8bit_RGB.jpg
.\images\exported\saved_using_pyplot.jpg
.\images\exported\saved_using_skimage.jpg
.\images\exported\saved_using_skimage.tif
.\images\exported\saved_using_skimage_8bit.tif
.\images\exported\saved_using_tifffile.tiff
  • 使用 os.walk 遍历文件夹
import os

print(os.walk('.'))  # Nothing to see here as this is just a generator

for root, dirs, files in os.walk('.'):
    # root 表示当前正在访问的文件夹路径
    # dirs 表示该文件夹下的子目录名list
    # files 表示该文件夹下的文件list
    # 遍历所有的文件夹
    for d in dirs:
        print(os.path.join(root, d))
<generator object walk at 0x000001AAF06406C8>
.\.ipynb_checkpoints
.\images
.\images\exported
上一篇：DIP-Introductory python tutorials for image processing(29-41)-Image Filtering
下一篇：DIP-Introductory python tutorials for image processing(1-21)-Python Basics