def resize(src):
height, width, _ = src.shape
dst = src
if height != width:
if a.pad:
size = max(height, width)
# pad to correct ratio
oh = (size - height) // 2
ow = (size - width) // 2
dst = im.pad(image=dst,
offset_height=oh,
offset_width=ow,
target_height=size,
target_width=size)
else:
# crop to correct ratio
size = min(height, width)
oh = (height - size) // 2
ow = (width - size) // 2
dst = im.crop(image=dst,
offset_height=oh,
offset_width=ow,
target_height=size,
target_width=size)
assert (dst.shape[0] == dst.shape[1])
size, _, _ = dst.shape
if size > a.size:
dst = im.downscale(images=dst, size=[a.size, a.size])
elif size < a.size:
dst = im.upscale(images=dst, size=[a.size, a.size])
return dst
def blur(src,scale=4): #通过缩小在放大的方式，对Image进行模糊处理；将image_blur和原image通过combine结合后，即可作为训练数据，输入pix2pix训练DCGAN，训练好的DCGAN中的G网络，当输入一个Image_blur后，可以输出一个image_clear；
height,width,_ = src.shape
height_down = height // scale
width_down = width // scale
dst = im.downscale(images=src,size=[height_down,width_down])
dst = im.upscale(images=dst,size=[height,width])
return dst
def blur(src, scale=4):
height, width, _ = src.shape
height_down = height // scale
width_down = height // scale
dst = im.downscale(image=src, size=[height_down, width_down])
dst = im.upscale(image=dst, size=[height, width])
return dst
def resize(src):
height, width, _ = src.shape
dst = src
if height != width:
if a.pad:
size = max(height, width)
# pad to correct ratio
oh = (size - height) // 2
ow = (size - width) // 2
dst = im.pad(image=dst, offset_height=oh, offset_width=ow, target_height=size, target_width=size)
else:
# crop to correct ratio
size = min(height, width)
oh = (height - size) // 2
ow = (width - size) // 2
dst = im.crop(image=dst, offset_height=oh, offset_width=ow, target_height=size, target_width=size)
assert(dst.shape[0] == dst.shape[1])
size, _, _ = dst.shape
if size > a.size:
dst = im.downscale(images=dst, size=[a.size, a.size])
elif size < a.size:
dst = im.upscale(images=dst, size=[a.size, a.size])
return dst
def blur(src, scale=4):
height, width, _ = src.shape
height_down = height // scale
width_down = width // scale
dst = im.downscale(images=src, size=[height_down, width_down])
dst = im.upscale(images=dst, size=[height, width])
return dst
def resize(src, tfUpscaleInterpolationType):
height, width, _ = src.shape
dst = src
if a.preserve_aspect_ratio:
if height != width:
if a.pad:
size = max(height, width)
# pad to correct ratio
oh = (size - height) // 2
ow = (size - width) // 2
dst = im.pad(image=dst,
offset_height=oh,
offset_width=ow,
target_height=size,
target_width=size)
else:
# crop to correct ratio
size = min(height, width)
oh = (height - size) // 2
ow = (width - size) // 2
dst = im.crop(image=dst,
offset_height=oh,
offset_width=ow,
target_height=size,
target_width=size)
assert (dst.shape[0] == dst.shape[1])
size, _, _ = dst.shape
if size > a.fm_input_size:
dst = im.downscale(images=dst,
size=[a.fm_input_size, a.fm_input_size])
elif size < a.fm_input_size:
if tfUpscaleInterpolationType == tf.image.ResizeMethod.BICUBIC:
dst = im.upscale(
images=dst,
size=[a.fm_input_size, a.fm_input_size])
elif tfUpscaleInterpolationType == tf.image.ResizeMethod.NEAREST_NEIGHBOR:
dst = im.upscaleWithNearestNeighborInterpolation(
images=dst,
size=[a.fm_input_size, a.fm_input_size])
else:
if tfUpscaleInterpolationType == tf.image.ResizeMethod.BICUBIC:
dst = im.upscale(images=dst,
size=[a.fm_input_size, a.fm_input_size])
elif tfUpscaleInterpolationType == tf.image.ResizeMethod.NEAREST_NEIGHBOR:
dst = im.upscaleWithNearestNeighborInterpolation(
images=dst, size=[a.fm_input_size, a.fm_input_size])
return dst
def resize(src):
height,width,_ =src.shape
dst = src
if height != width:
if a.pad: #如果需要padding，以较长边padding
size = max(height,width)
oh = (size - height) // 2
ow = (size - width) // 2
dst = im.pad(image = dst,offset_height = oh, offset_width = ow,target_height = size,target_width = size)
else: #否则，以较短边修剪image
size = min(height,width)
oh = (height - size) //2
ow = (width - size) //2
dst = im.crop(image=dst,offset_height=oh,offset_width=ow,target_height=size,target_width=size)
assert(dst.shape[0] == dst.shape[1])
size,_,_ = dst.shape
if size > a.size:#如果修剪后的image尺寸与a.size不符合，则进行相应的修改
dst = im.downscale(images=dst,size=[a.size,a.size])
elif size < a.size:
dst = im.upscale(images=dst,size=[a.size,a.size])
return dst
def downscaleToSize(src, scaleSize):
dst = im.downscale(images=src, size=scaleSize)
return dst