glue_certainty_function.py

import Image
import math
# import matplotlib.pyplot as plt
# import time
# import numpy
# from scipy.optimize import curve_fit

# initializing variables
resolution = 1 # every pixel
i = 0
j = 0
k = 0

# load image with under exposure
f_under = Image.open('under_ex_lowres.JPG')
x_max = f_under.size[0]
y_max = f_under.size[1]
image_under = f_under.load()

# load image with over_exposure
f_over = Image.open('over_ex_lowres.JPG')
x_max = f_over.size[0]
y_max = f_over.size[1]
image_over = f_over.load()

# # loop through to try different ratios
# for under_ratio in range(40,200,20):

#certainty function
def certainty(pixelvalue):
	x = 6*(float(pixelvalue)/255) - 3
	return math.exp(-x**2)

# create a new image
f_combine = Image.new('RGB', (x_max, y_max))
image_combine = f_combine.load()

while i < x_max:
	while j < y_max:

		temp_arr = [0.0,0.0,0.0]

		for k in range(3):
			a = float(image_under[i,j][k])
			b = float(image_over[i,j][k])

			temp_arr[k] = (certainty(a)*a + certainty(b)*b) / (certainty(a) + certainty(b))
			temp_arr[k] = int(temp_arr[k])

			k += 1
		image_combine[i,j] = tuple(temp_arr)
		k = 0
		temp_arr = []

		j += resolution
	i += resolution
	j = 0 

f_combine.show()