|View source on GitHub|
size using the specified
tf.image.resize( images, size, method=ResizeMethod.BILINEAR, preserve_aspect_ratio=False, antialias=False, name=None )
Used in the guide:
Used in the tutorials:
- Adversarial example using FGSM
- Image captioning with visual attention
- Image segmentation
- Load images
- Neural style transfer
- Transfer learning with a pretrained ConvNet
Resized images will be distorted if their original aspect ratio is not
the same as
size. To avoid distortions see
When 'antialias' is true, the sampling filter will anti-alias the input image as well as interpolate. When downsampling an image with anti-aliasing the sampling filter kernel is scaled in order to properly anti-alias the input image signal. 'antialias' has no effect when upsampling an image.
bilinear: Bilinear interpolation. If 'antialias' is true, becomes a hat/tent filter function with radius 1 when downsampling.
lanczos3: Lanczos kernel with radius 3. High-quality practical filter but may have some ringing especially on synthetic images.
lanczos5: Lanczos kernel with radius 5. Very-high-quality filter but may have stronger ringing.
bicubic: Cubic interpolant of Keys. Equivalent to Catmull-Rom kernel. Reasonably good quality and faster than Lanczos3Kernel, particularly when upsampling.
gaussian: Gaussian kernel with radius 3, sigma = 1.5 / 3.0.
nearest: Nearest neighbor interpolation. 'antialias' has no effect when used with nearest neighbor interpolation.
area: Anti-aliased resampling with area interpolation. 'antialias' has no effect when used with area interpolation; it always anti-aliases.
mitchellcubic: Mitchell-Netravali Cubic non-interpolating filter. For synthetic images (especially those lacking proper prefiltering), less ringing than Keys cubic kernel but less sharp.
Note that near image edges the filtering kernel may be partially outside the image boundaries. For these pixels, only input pixels inside the image will be included in the filter sum, and the output value will be appropriately normalized.
The return value has the same type as
ResizeMethod.NEAREST_NEIGHBOR. Otherwise, the return value has type
images: 4-D Tensor of shape
[batch, height, width, channels]or 3-D Tensor of shape
[height, width, channels].
size: A 1-D int32 Tensor of 2 elements:
new_height, new_width. The new size for the images.
method: ResizeMethod. Defaults to
preserve_aspect_ratio: Whether to preserve the aspect ratio. If this is set, then
imageswill be resized to a size that fits in
sizewhile preserving the aspect ratio of the original image. Scales up the image if
sizeis bigger than the current size of the
image. Defaults to False.
antialias: Whether to use an anti-aliasing filter when downsampling an image.
name: A name for this operation (optional).
ValueError: if the shape of
imagesis incompatible with the shape arguments to this function
sizehas invalid shape or type.
ValueError: if an unsupported resize method is specified.
images was 4-D, a 4-D float Tensor of shape
[batch, new_height, new_width, channels].
images was 3-D, a 3-D float Tensor of shape
[new_height, new_width, channels].