如何将列表（或 numpy 数组或矩阵）转换为图像？

Question

I have a huge table (or matrix, if you want). I would like to visualize this table by making an image in which each pixel corresponds to a cell in the table (so, the row and columns in the table corresponds to the "rows" and "columns" of the pixels in the image).

The values in the cells corresponds already to the colors (for example [255, 0, 0] is the value that corresponds to red). In other words, each cell in the table contains a 3 dimensional vector representing a color.

The table can be represented as a list of lists of lists or as a 3D numpy array.

Answer 1

You can use matplotlib's imshow function to do this. You just need to normalise your array from 0 - 255 to 0 - 1 , then you can feed imshow your MxNx3 array.

Here's a simple example:

import matplotlib.pyplot as plt
import numpy as np

fig,ax=plt.subplots(1,subplot_kw={'aspect':'equal'})

# A small 3x3 sample array
imarray=np.array([
    [[255,0,0],[0,255,0],[0,0,255]],
    [[255,255,0],[0,255,255],[255,0,255]],
    [[255,255,255],[127,127,127],[0,0,0]]
    ])

# Normalise the data to between 0 and 1
imarray = imarray.astype('float')/255.

# plot the array
ax.imshow(imarray,interpolation='nearest')

plt.show()