使用Python-opencv对图像原始数据进行操作
在ipython下进行调试，结果如下
以下是一些很基础的图像操作，虽然现有的图像处理工具很强大，封装的很高级，但正是因为太高级，也导致使用者不了解一些底层的东西，当然也包括我。
"""
参考：《OpenCV3计算机视觉 Python语言实现 第二版》，该书使用的是Python2
"""
In [1]: import cv2
In [2]: import numpy as np
In [4]: img = np.zeros([640,480,3],dtype = np.uint8)
In [5]: img
Out[5]:
array([[[0, 0, 0],
[0, 0, 0],
[0, 0, 0],
...,
[0, 0, 0],
[0, 0, 0],
[0, 0, 0]],
[[0, 0, 0],
[0, 0, 0],
[0, 0, 0],
...,
[0, 0, 0],
[0, 0, 0],
[0, 0, 0]],
[[0, 0, 0],
[0, 0, 0],
[0, 0, 0],
...,
[0, 0, 0],
[0, 0, 0],
[0, 0, 0]],
...,
[[0, 0, 0],
[0, 0, 0],
[0, 0, 0],
...,
[0, 0, 0],
[0, 0, 0],
[0, 0, 0]],
[[0, 0, 0],
[0, 0, 0],
[0, 0, 0],
...,
[0, 0, 0],
[0, 0, 0],
[0, 0, 0]],
[[0, 0, 0],
[0, 0, 0],
[0, 0, 0],
...,
[0, 0, 0],
[0, 0, 0],
[0, 0, 0]]], dtype=uint8)
In [6]: cv2.imwrite('black.png',img)#RGB888数据要保存成png格式
Out[6]: True
该图即为输出结果。
"""
In [7]: img[:,:,0]
Out[7]:
array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0, ..., 0, 0, 0],
...,
[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0, ..., 0, 0, 0]], dtype=uint8)
在opencv中，读写图片通道顺序位均BGR
"""
In [8]: img[:,:,0] = 255 # 通道1 blue
In [9]: cv2.imwrite('blu.png',img)
Out[9]: True
In [11]: img[:,:,1] = 255 # 通道2 green
In [12]: img[:,:,0] = 0
In [13]: cv2.imwrite('green.png',img)
Out[13]: True
In [14]: img[:,:,1] = 0
In [15]: img[:,:,2] = 255
In [16]: cv2.imwrite('red.png',img)
Out[16]: True
In [17]: image = cv2.imread('red.png')
In [18]: image
Out[18]:
array([[[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255],
...,
[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255]],
[[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255],
...,
[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255]],
[[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255],
...,
[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255]],
...,
[[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255],
...,
[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255]],
[[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255],
...,
[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255]],
[[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255],
...,
[ 0, 0, 255],
[ 0, 0, 255],
[ 0, 0, 255]]], dtype=uint8)
In [23]: imgcvt = cv2.cvtColor(img,cv2.COLOR_BGR2RGB)
In [25]: imgcvt
Out[25]:
array([[[255, 0, 0],
[255, 0, 0],
[255, 0, 0],
...,
[255, 0, 0],
[255, 0, 0],
[255, 0, 0]],
[[255, 0, 0],
[255, 0, 0],
[255, 0, 0],
...,
[255, 0, 0],
[255, 0, 0],
[255, 0, 0]],
[[255, 0, 0],
[255, 0, 0],
[255, 0, 0],
...,
[255, 0, 0],
[255, 0, 0],
[255, 0, 0]],
...,
[[255, 0, 0],
[255, 0, 0],
[255, 0, 0],
...,
[255, 0, 0],
[255, 0, 0],
[255, 0, 0]],
[[255, 0, 0],
[255, 0, 0],
[255, 0, 0],
...,
[255, 0, 0],
[255, 0, 0],
[255, 0, 0]],
[[255, 0, 0],
[255, 0, 0],
[255, 0, 0],
...,
[255, 0, 0],
[255, 0, 0],
[255, 0, 0]]], dtype=uint8)
In [26]: cv2.imwrite('BLUE.png',imgcvt)
Out[26]: True
图像数据与原始字节转换演示：
"""
In [37]: A = np.zeros([5,5,3],dtype = np.uint8)
In [44]: for i in range(5):
...: for j in range(5):
...: A[:,:,0][i,j] = i+j
...:
In [45]: A[:,:,0]
Out[45]:
array([[0, 1, 2, 3, 4],
[1, 2, 3, 4, 5],
[2, 3, 4, 5, 6],
[3, 4, 5, 6, 7],
[4, 5, 6, 7, 8]], dtype=uint8)
In [46]: A[:,:,1] = A[:,:,0] +1
In [47]: A[:,:,2] = A[:,:,0] +2
In [48]: A[:,:,1]
Out[48]:
array([[1, 2, 3, 4, 5],
[2, 3, 4, 5, 6],
[3, 4, 5, 6, 7],
[4, 5, 6, 7, 8],
[5, 6, 7, 8, 9]], dtype=uint8)
In [49]: A[:,:,2]
Out[49]:
array([[ 2, 3, 4, 5, 6],
[ 3, 4, 5, 6, 7],
[ 4, 5, 6, 7, 8],
[ 5, 6, 7, 8, 9],
[ 6, 7, 8, 9, 10]], dtype=uint8)
#图像数据与原始字节转换
In [51]: AbyteArray = bytearray(A) #
In [52]: AbyteArray
bytearray(b'\x00\x01\x02\x01\x02\x03\x02\x03\x04\x03\x04\x05\x04\x05\x06\x01\x02\x03\x02\x03\x04\x03\x04\x05\x04\x05\x06\x05\x06\x07\x02\x03\x04\x03\x04\x05\x04\x05\x06\x05\x06\x07\x06\x07\x08\x03\x04\x05\x04\x05\x06\x05\x06\x07\x06\x07\x08\x07\x08\t\x04\x05\x06\x05\x06\x07\x06\x07\x08\x07\x08\t\x08\t\n')
#由此可以看出，使用bytearray进行数据与原始字节转换时，是按照先处理像素后处理通道的原则进行的。
In [64]: A.tostring() == bytearray(A)
Out[64]: True
In [65]: A.tostring() == A.tobytes()
Out[65]: True
In [84]: np.fromstring(bytearray(A),dtype='uint8')
---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
<ipython-input-84-c84d62f1d047> in <module>()
----> 1 np.fromstring(bytearray(A),dtype='uint8')
TypeError: fromstring() argument 1 must be read-only bytes-like object, not bytearray
In [86]: B = np.frombuffer(A.tobytes(),dtype='uint8')
In [87]: B
Out[87]:
array([ 0, 1, 2, 1, 2, 3, 2, 3, 4, 3, 4, 5, 4, 5, 6, 1, 2,
3, 2, 3, 4, 3, 4, 5, 4, 5, 6, 5, 6, 7, 2, 3, 4, 3,
4, 5, 4, 5, 6, 5, 6, 7, 6, 7, 8, 3, 4, 5, 4, 5, 6,
5, 6, 7, 6, 7, 8, 7, 8, 9, 4, 5, 6, 5, 6, 7, 6, 7,
8, 7, 8, 9, 8, 9, 10], dtype=uint8)
In [91]: C = B.reshape(5,5,3)
In [94]: C == A
Out[94]:
array([[[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True]],
[[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True]],
[[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True]],
[[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True]],
[[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True],
[ True, True, True]]])
版权声明：本文为python_java0C原创文章，遵循CC 4.0 BY-SA版权协议，转载请附上原文出处链接和本声明。