如何修改带有保留假色的灰色cmap?

时间:2017-06-02 06:37:47

标签: python image matplotlib

假设256灰度图像,

如何修改颜色贴图plt.cm.gray,以便给定灰度值的像素以给定颜色(红色,蓝色......)出现。例如,如何将value = 1的像素设置为红色,将value = 2的像素设置为绿色?

我了解masked array example。但在该示例中,似乎只能设置一种颜色。

我尝试生成自定义cmap" agrey" (失败了):

## try to make a custom cmap
Ngrey = 256
a = np.linspace(0,1,num=Ngrey, endpoint=True)
A = np.array((a,a,a)).transpose()

#Set pixel with greylevel=1 to red
A[1,1:3]=0

col_dict = {'red':A,'green':A, 'blue':A}
print col_dict['blue'].shape
agrey = LinearSegmentedColormap('mygray', col_dict)

1 个答案:

答案 0 :(得分:3)

当您处理离散灰度时,不是使用LinearSegmentedColormap,而是使用ListedColormap,您可以在其中定义256个灰度值,然后覆盖您想要着色的值。下面是一个随机图片的最小例子:

from matplotlib import pyplot as plt
from matplotlib.colors import ListedColormap
import numpy as np

pic = np.random.randint(256, size=(100,100))

Ngrey = 256
greys = np.linspace(0,1,Ngrey)

colors = [[g,g,g] for g in greys]

red = [1,0,0]
green = [0,1,0]
blue = [0,0,1]

colors[5] = red
colors[100] = blue
colors[200] = green

mymap=ListedColormap(colors)

plt.matshow(pic, cmap=mymap)
plt.show()

结果看起来像这样:random grey scale image with certain, discrete values coloured in red, green, and blue

Python 3.5

上进行测试