The result is:
plt.plot and plt.scatter is used in this page as an example.
You can plot by mapping function that convert the point of the plotting data to that of the image.
In [1]:
%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
from PIL import Image
In [2]:
img = Image.open("./lena_color.gif")
The original image is:
In [3]:
img
Out[3]:
and the type and the shape of the image is follows:
In [4]:
type(img)
Out[4]:
In [5]:
img.size # returns (x, y) of the image
Out[5]:
if plt.imshow() is used as drawing function:
In [6]:
plt.imshow(img)
Out[6]:
you can hide the axis by adding six lines:
In [7]:
fig, ax = plt.subplots()
ax.imshow(img)
ax.spines['top'].set_visible(False)
ax.spines['left'].set_visible(False)
ax.spines['bottom'].set_visible(False)
ax.spines['right'].set_visible(False)
ax.set_xticks([])
ax.set_yticks([])
plt.show()
You can also rewrite the ticklabels by using the mapping function. For example:
In [8]:
xy2imgxy = lambda x,y: (img.size[0] * x / np.max(ticklx),\
img.size[1] * (np.max(tickly) - y) / np.max(tickly))
In [9]:
ticklx = np.linspace(0,100,6)
tickly = np.linspace(0,100,6)
tickpx,tickpy = xy2imgxy(ticklx,tickly)
In [10]:
fig,ax = plt.subplots()
ax.imshow(img)
# Rewrite x,y ticks
ax.set_xticks(tickpx)
ax.set_yticks(tickpy)
ax.set_xticklabels(ticklx.astype('int'))
ax.set_yticklabels(tickly.astype('int'))
plt.show()
You can draw plot or scatter plot on the image by converting the point of the data using mapping function. For example:
In [11]:
fig,ax = plt.subplots()
ax.imshow(img)
# Rewrite x,y ticks
ax.set_xticks(tickpx)
ax.set_yticks(tickpy)
ax.set_xticklabels(ticklx.astype('int'))
ax.set_yticklabels(tickly.astype('int'))
# Add scatter point on the image.
px,py = 20,20
imgx,imgy = xy2imgxy(px,py)
ax.scatter(imgx,imgy,s=100,lw=5,facecolor="none",edgecolor="yellow")
# Add plot on the image.
px = np.linspace(0,100,500)
py = 10*np.abs(np.sin(2*np.pi*0.02*px))
imgx,imgy = xy2imgxy(px,py)
ax.plot(imgx,imgy,color="yellow",lw=3)
# Adjust the axis.
ax.set_xlim(0,tickpx.max())
ax.set_ylim(tickpy.max(),0)
# Save figure.
plt.savefig("lena_color_mod.png",bbox_inches="tight",pad_inches=0.02,dpi=250)
plt.show()
