/usr/lib/python2.7/dist-packages/ssim/ssimlib.py is in python-pyssim 0.2-1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 | """Contains SSIM library functions and classes."""
from __future__ import absolute_import
import argparse
import glob
import sys
import numpy as np
from scipy import signal
from ssim import compat
from ssim.compat import Image, ImageOps
from ssim.utils import convolve_gaussian_2d
from ssim.utils import get_gaussian_kernel
from ssim.utils import to_grayscale
class SSIMImage(object):
"""Wraps a PIL Image object with SSIM state.
Attributes:
img: Original PIL Image.
img_gray: grayscale Image.
img_gray_squared: squared img_gray.
img_gray_mu: img_gray convolved with gaussian kernel.
img_gray_mu_squared: squared img_gray_mu.
img_gray_sigma_squared: img_gray convolved with gaussian kernel -
img_gray_mu_squared.
"""
def __init__(self, img, gaussian_kernel_1d=None, size=None):
"""Create an SSIMImage.
Args:
img (str or PIL.Image): PIL Image object or file name.
gaussian_kernel_1d (np.ndarray, optional): Gaussian kernel
that was generated with utils.get_gaussian_kernel is used
to precompute common objects for SSIM computation
size (tuple, optional): New image size to resize image to.
"""
# Use existing or create a new PIL.Image
self.img = img if not isinstance(img, compat.basestring) \
else compat.Image.open(img)
# Resize image if size is defined and different
# from original image
if size and size != self.img.size:
self.img = self.img.resize(size, Image.ANTIALIAS)
# Set the size of the image
self.size = self.img.size
# If gaussian kernel is defined we create
# common SSIM objects
if gaussian_kernel_1d is not None:
self.gaussian_kernel_1d = gaussian_kernel_1d
# np.array of grayscale and alpha image
self.img_gray, self.img_alpha = to_grayscale(self.img)
if self.img_alpha is not None:
self.img_gray[self.img_alpha == 255] = 0
# Squared grayscale
self.img_gray_squared = self.img_gray ** 2
# Convolve grayscale image with gaussian
self.img_gray_mu = convolve_gaussian_2d(
self.img_gray, self.gaussian_kernel_1d)
# Squared mu
self.img_gray_mu_squared = self.img_gray_mu ** 2
# Convolve squared grayscale with gaussian
self.img_gray_sigma_squared = convolve_gaussian_2d(
self.img_gray_squared, self.gaussian_kernel_1d)
# Substract squared mu
self.img_gray_sigma_squared -= self.img_gray_mu_squared
# If we don't define gaussian kernel, we create
# common CW-SSIM objects
else:
# Grayscale PIL.Image
self.img_gray = ImageOps.grayscale(self.img)
class SSIM(object):
"""Computes SSIM between two images."""
def __init__(self, img, gaussian_kernel_1d=None, size=None,
l=255, k_1=0.01, k_2=0.03, k=0.01):
"""Create an SSIM object.
Args:
img (str or PIL.Image): Reference image to compare other images to.
l, k_1, k_2 (float): SSIM configuration variables.
k (float): CW-SSIM configuration variable (default 0.01)
gaussian_kernel_1d (np.ndarray, optional): Gaussian kernel
that was generated with utils.get_gaussian_kernel is used
to precompute common objects for SSIM computation
size (tuple, optional): resize the image to the tuple size
"""
self.k = k
# Set k1,k2 & c1,c2 to depend on L (width of color map).
self.c_1 = (k_1 * l) ** 2
self.c_2 = (k_2 * l) ** 2
self.gaussian_kernel_1d = gaussian_kernel_1d
self.img = SSIMImage(img, gaussian_kernel_1d, size)
def ssim_value(self, target):
"""Compute the SSIM value from the reference image to the target image.
Args:
target (str or PIL.Image): Input image to compare the reference image
to. This may be a PIL Image object or, to save time, an SSIMImage
object (e.g. the img member of another SSIM object).
Returns:
Computed SSIM float value.
"""
# Performance boost if handed a compatible SSIMImage object.
if not isinstance(target, SSIMImage) \
or not np.array_equal(self.gaussian_kernel_1d,
target.gaussian_kernel_1d):
target = SSIMImage(target, self.gaussian_kernel_1d, self.img.size)
img_mat_12 = self.img.img_gray * target.img_gray
img_mat_sigma_12 = convolve_gaussian_2d(
img_mat_12, self.gaussian_kernel_1d)
img_mat_mu_12 = self.img.img_gray_mu * target.img_gray_mu
img_mat_sigma_12 = img_mat_sigma_12 - img_mat_mu_12
# Numerator of SSIM
num_ssim = ((2 * img_mat_mu_12 + self.c_1) *
(2 * img_mat_sigma_12 + self.c_2))
# Denominator of SSIM
den_ssim = (
(self.img.img_gray_mu_squared + target.img_gray_mu_squared +
self.c_1) *
(self.img.img_gray_sigma_squared +
target.img_gray_sigma_squared + self.c_2))
ssim_map = num_ssim / den_ssim
index = np.average(ssim_map)
return index
def cw_ssim_value(self, target, width=30):
"""Compute the complex wavelet SSIM (CW-SSIM) value from the reference
image to the target image.
Args:
target (str or PIL.Image): Input image to compare the reference image
to. This may be a PIL Image object or, to save time, an SSIMImage
object (e.g. the img member of another SSIM object).
width: width for the wavelet convolution (default: 30)
Returns:
Computed CW-SSIM float value.
"""
if not isinstance(target, SSIMImage):
target = SSIMImage(target, size=self.img.size)
# Define a width for the wavelet convolution
widths = np.arange(1, width+1)
# Use the image data as arrays
sig1 = np.asarray(self.img.img_gray.getdata())
sig2 = np.asarray(target.img_gray.getdata())
# Convolution
cwtmatr1 = signal.cwt(sig1, signal.ricker, widths)
cwtmatr2 = signal.cwt(sig2, signal.ricker, widths)
# Compute the first term
c1c2 = np.multiply(abs(cwtmatr1), abs(cwtmatr2))
c1_2 = np.square(abs(cwtmatr1))
c2_2 = np.square(abs(cwtmatr2))
num_ssim_1 = 2 * np.sum(c1c2, axis=0) + self.k
den_ssim_1 = np.sum(c1_2, axis=0) + np.sum(c2_2, axis=0) + self.k
# Compute the second term
c1c2_conj = np.multiply(cwtmatr1, np.conjugate(cwtmatr2))
num_ssim_2 = 2 * np.abs(np.sum(c1c2_conj, axis=0)) + self.k
den_ssim_2 = 2 * np.sum(np.abs(c1c2_conj), axis=0) + self.k
# Construct the result
ssim_map = (num_ssim_1 / den_ssim_1) * (num_ssim_2 / den_ssim_2)
# Average the per pixel results
index = np.average(ssim_map)
return index
def main():
"""Main function for pyssim."""
description = '\n'.join([
'Compares an image with a list of images using the SSIM metric.',
' Example:',
' pyssim test-images/test1-1.png "test-images/*"'
])
parser = argparse.ArgumentParser(
prog='pyssim', formatter_class=argparse.RawTextHelpFormatter,
description=description)
parser.add_argument('--cw', help='compute the complex wavelet SSIM',
action='store_true')
parser.add_argument(
'base_image', metavar='image1.png', type=argparse.FileType('r'))
parser.add_argument(
'comparison_images', metavar='image path with* or image2.png')
parser.add_argument('--width', type=int, default=None,
help='scales the image before computing SSIM')
parser.add_argument('--height', type=int, default=None,
help='scales the image before computing SSIM')
args = parser.parse_args()
if args.width and args.height:
size = (args.width, args.height)
else:
size = None
if not args.cw:
gaussian_kernel_sigma = 1.5
gaussian_kernel_width = 11
gaussian_kernel_1d = get_gaussian_kernel(
gaussian_kernel_width, gaussian_kernel_sigma)
comparison_images = glob.glob(args.comparison_images)
is_a_single_image = len(comparison_images) == 1
for comparison_image in comparison_images:
if args.cw:
ssim = SSIM(args.base_image.name, size=size)
ssim_value = ssim.cw_ssim_value(comparison_image)
else:
ssim = SSIM(args.base_image.name, gaussian_kernel_1d, size=size)
ssim_value = ssim.ssim_value(comparison_image)
if is_a_single_image:
sys.stdout.write('%.7g' % ssim_value)
else:
sys.stdout.write('%s - %s: %.7g' % (
args.base_image.name, comparison_image, ssim_value))
sys.stdout.write('\n')
if __name__ == '__main__':
main()
|