MPImfp 0.0.1

Python library to measure the image mean free path (IMFP)

MPImfp

Python library to measure the image mean free path (IMFP).

Example

import MPImfp
import cv2
import numpy as np
import matplotlib.pyplot as plt

img = cv2.imread('sample.png', 0)
size = 1000
barrier = 0
freq_single = np.zeros(size, dtype=np.uint32)
freq_double = np.zeros(size, dtype=np.uint32)
dpix = 1.0
nsample = 1000000
seed = 12345
# single mode
MPImfp.measure(img, barrier, freq_single, dpix, nsample, seed, 0)
ave_single = np.sum(np.arange(size)*np.array(freq_single, dtype=np.float64)/np.sum(freq_single))
print('Average Single :', ave_single)
# double mode
MPImfp.measure(img, barrier, freq_double, dpix, nsample, seed, 1)
ave_double = np.sum(np.arange(size)*np.array(freq_double, dtype=np.float64)/np.sum(freq_double))
print('Average Double :', ave_double)
# plot
plt.plot(np.arange(size), freq_single)
plt.plot(np.arange(size), freq_double)
plt.xlabel('Pixel'), plt.ylabel('Frequency')
plt.show()

References

Methods :

measure(img, barrier, f, dpix, nsample, seed, dflag)

measure image mean free path
return seed

• img : input image
• barrier : pixel value of barrier, 0 - 255
• f : numpy dimension for result, dtype=np.uint32
• dpix : length of a pixel
• nsample : number of sample
• seed : seed for random number
• dflag : measure mode, 0 : single mode, 1 : double mode