def process_layer(r):
if not isinstance(r, np.ndarray):
r = np.asarray(r).astype(float) / 255.0
if r.ndim < 3:
r2 = np.empty((r.shape[0], r.shape[1], 3), dtype=float)
r2[:, :, :] = r[:, :, np.newaxis]
r = r2
r /= np.percentile(r, 99.9)
return rgb_to_srgb(r)
def process_layer(r):
if not isinstance(r, np.ndarray):
r = np.asarray(r).astype(float) / 255.0
if r.ndim < 3:
r2 = np.empty((r.shape[0], r.shape[1], 3), dtype=float)
r2[:, :, :] = r[:, :, np.newaxis]
r = r2
r /= np.percentile(r, 99.9)
return rgb_to_srgb(r)
def open_multilayer_exr(filename, tonemap=False, thumb_size=0, show_progress=False):
"""
Load a multilayer OpenEXR file and return a dictionary mapping layers to
either numpy float32 arrays (if ``tonemap=False``) or to 8bit PIL images
(if ``tonemap=True``).
:param filename: string filename
:param tonemap: if ``True``, map to sRGB
:param thumb_size: if nonzero, resize images to have this as their max dimension
:param show_progress: if ``True``, print info about loading
"""
if show_progress:
print "Reading %s: %s layers..." % (filename, len(LAYER_CHANNELS))
# Open the input file
f = OpenEXR.InputFile(filename)
header = f.header()
# Compute the size
dw = header['dataWindow']
cols, rows = dw.max.x - dw.min.x + 1, dw.max.y - dw.min.y + 1
multilayer = {}
# load channels
FLOAT = Imath.PixelType(Imath.PixelType.FLOAT)
for key, channels in LAYER_CHANNELS.iteritems():
print "Loading layer %s..." % key
image = np.empty((rows, cols, 3), dtype=np.float32)
for (i, c) in enumerate(channels):
data = f.channel(c, FLOAT)
image[:, :, i] = np.fromstring(data, dtype=np.float32) \
.reshape((rows, cols))
multilayer[key] = image
#if denoise:
#for t in ["diff", "gloss", "trans"]:
#print "Denoising layer %s..." % t
#multilayer["%s_ind" % t] = denoise_indirect_image(multilayer["%s_ind" % t])
## recompute combined image using denoised layer
#multilayer["combined"] = multilayer["emit"] + multilayer["env"]
#for t in ["diff", "gloss", "trans"]:
#multilayer["combined"] += (
#multilayer["%s_col" % t] * (multilayer["%s_dir" % t] + multilayer["%s_ind" % t])
#)
# resize and tonemap
if tonemap:
for key, channels in LAYER_CHANNELS.iteritems():
print "Tonemapping layer %s..." % key
image = multilayer[key]
if key == "depth":
image /= np.max(image[np.isfinite(image)])
# convert to sRGB PIL
image = numpy_to_pil(rgb_to_srgb(np.clip(image, 0.0, 1.0)))
if thumb_size and key != "combined":
image = ResizeToFit(thumb_size, thumb_size).process(image)
multilayer[key] = image
return multilayer
def open_multilayer_exr(filename,
tonemap=False,
thumb_size=0,
show_progress=False):
"""
Load a multilayer OpenEXR file and return a dictionary mapping layers to
either numpy float32 arrays (if ``tonemap=False``) or to 8bit PIL images
(if ``tonemap=True``).
:param filename: string filename
:param tonemap: if ``True``, map to sRGB
:param thumb_size: if nonzero, resize images to have this as their max dimension
:param show_progress: if ``True``, print info about loading
"""
if show_progress:
print "Reading %s: %s layers..." % (filename, len(LAYER_CHANNELS))
# Open the input file
f = OpenEXR.InputFile(filename)
header = f.header()
# Compute the size
dw = header['dataWindow']
cols, rows = dw.max.x - dw.min.x + 1, dw.max.y - dw.min.y + 1
multilayer = {}
# load channels
FLOAT = Imath.PixelType(Imath.PixelType.FLOAT)
for key, channels in LAYER_CHANNELS.iteritems():
print "Loading layer %s..." % key
image = np.empty((rows, cols, 3), dtype=np.float32)
for (i, c) in enumerate(channels):
data = f.channel(c, FLOAT)
image[:, :, i] = np.fromstring(data, dtype=np.float32) \
.reshape((rows, cols))
multilayer[key] = image
#if denoise:
#for t in ["diff", "gloss", "trans"]:
#print "Denoising layer %s..." % t
#multilayer["%s_ind" % t] = denoise_indirect_image(multilayer["%s_ind" % t])
## recompute combined image using denoised layer
#multilayer["combined"] = multilayer["emit"] + multilayer["env"]
#for t in ["diff", "gloss", "trans"]:
#multilayer["combined"] += (
#multilayer["%s_col" % t] * (multilayer["%s_dir" % t] + multilayer["%s_ind" % t])
#)
# resize and tonemap
if tonemap:
for key, channels in LAYER_CHANNELS.iteritems():
print "Tonemapping layer %s..." % key
image = multilayer[key]
if key == "depth":
image /= np.max(image[np.isfinite(image)])
# convert to sRGB PIL
image = numpy_to_pil(rgb_to_srgb(np.clip(image, 0.0, 1.0)))
if thumb_size and key != "combined":
image = ResizeToFit(thumb_size, thumb_size).process(image)
multilayer[key] = image
return multilayer
