Getting the pixel coordinates of text or ticks in matplotlib

The exact pixel coordinates of title, labels, legends or ticks are important information for the training data of deep learning. 

 

import numpy as np
import matplotlib.pyplot as plt

plt.plot([1,2],label="first_image")
plt.plot([2,1],label="second_image")
leg = plt.legend()

tit = plt.title("sss")
xla = plt.xlabel("xxx")

plt.gcf().canvas.draw()
ticks = [t for t in plt.gca().get_xticklabels()]

print(tit.get_text(),str(tit.get_window_extent()))    # return the context and pixel coordinates of 'tit'
print(np.array(tit.get_window_extent()))
print(xla.get_text(),str(xla.get_window_extent()))

for i, t in enumerate(ticks):
    print ("Label "+str(i)+", data: "+t.get_text()+str(t.get_window_extent()))

print ("Legend location: "+str(leg.get_window_extent()))
for i, l in enumerate(leg.texts):
    print ("Label "+str(i)+", data: "+l.get_text()+str(l.get_window_extent()))
    pass
#plt.savefig("test.png")
plt.show()

sss Bbox(x0=211.8375, y0=256.44, x1=230.9625, y1=269.44)
[[ 211.8375  256.44  ]
 [ 230.9625  269.44  ]]
xxx Bbox(x0=212.5875, y0=5.0, x1=230.2125, y1=15.0)
Label 0, data: Bbox(x0=69.21818181818182, y0=29.0, x1=69.21818181818182, y1=29.0)
Label 1, data: 0.0Bbox(x0=61.21818181818182, y0=19.0, x1=77.21818181818182, y1=29.0)
Label 2, data: 0.2Bbox(x0=122.0909090909091, y0=19.0, x1=138.0909090909091, y1=29.0)
Label 3, data: 0.4Bbox(x0=182.9636363636364, y0=19.0, x1=198.9636363636364, y1=29.0)
Label 4, data: 0.6Bbox(x0=243.83636363636367, y0=19.0, x1=259.83636363636367, y1=29.0)
Label 5, data: 0.8Bbox(x0=304.70909090909095, y0=19.0, x1=320.70909090909095, y1=29.0)
Label 6, data: 1.0Bbox(x0=365.5818181818182, y0=19.0, x1=381.5818181818182, y1=29.0)
Label 7, data: Bbox(x0=69.21818181818182, y0=29.0, x1=69.21818181818182, y1=29.0)
Legend location: Bbox(x0=276.05, y0=127.31375, x1=383.8, y1=162.12625)
Label 0, data: first_imageBbox(x0=308.05, y0=147.22, x1=363.925, y1=158.12625)
Label 1, data: second_imageBbox(x0=308.05, y0=131.31375, x1=379.8, y1=142.22)

 

 

reference:

1. https://matplotlib.org/api/text_api.html#matplotlib.text.Annotation.get_window_extent

2. https://matplotlib.org/api/transformations.html#matplotlib.transforms.Bbox