A library that helps you split image into small, overlappable patches, and merge patches back into the original image.
Project description
patchify
patchfy can split images into small overlappable patches by given patch cell size, and merge patches into original image.
This library provides two functions: patchify
, unpatchify
.
Installation
pip install patchify
Usage
Split image to patches
patchify(image_to_patch, patch_shape, step=1)
2D image:
#This will split the image into small images of shape [3,3]
patches = patchify(image, (3, 3), step=1)
3D image:
#This will split the image into small images of shape [3,3,3]
patches = patchify(image, (3, 3, 3), step=1)
Merge patches into original image
unpatchify(patches_to_merge, merged_image_size)
reconstructed_image = unpatchify(patches, image.shape)
This will reconstruct the original image that was patchified in previous code.
Caveat: in order for unpatchify
to work, you need to create patchies with equal step size. e.g. if the original image has width 3 and the patch has width 2, you cannot really create equal step size patches with step size 2. (first patch [elem0, elem1] and second patch [elem2, elem3], which is out of bound).
The required condition for unpatchify to success is to have (width - patch_width) mod step_size = 0.
Full running examples
2D image patchify and merge
import numpy as np
from patchify import patchify, unpatchify
image = np.array([[1,2,3,4], [5,6,7,8], [9,10,11,12]])
patches = patchify(image, (2,2), step=1) # split image into 2*3 small 2*2 patches.
assert patches.shape == (2, 3, 2, 2)
reconstructed_image = unpatchify(patches, image.shape)
assert (reconstructed_image == image).all()
3D image patchify and merge
import numpy as np
from patchify import patchify, unpatchify
image = np.random.rand(512,512,3)
patches = patchify(image, (2,2,3), step=1) # patch shape [2,2,3]
print(patches.shape) # (511, 511, 1, 2, 2, 3). Total patches created: 511x511x1
assert patches.shape == (511, 511, 1, 2, 2, 3)
reconstructed_image = unpatchify(patches, image.shape)
print(reconstructed_image.shape) # (512, 512, 3)
assert (reconstructed_image == image).all()
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file patchify-0.2.3.tar.gz
.
File metadata
- Download URL: patchify-0.2.3.tar.gz
- Upload date:
- Size: 5.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.1.5 CPython/3.8.6 Linux/5.4.0-65-generic
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 6cc409124f34ceee672f1931d818923f88f5116f323ac7bb9be7e6c5d0845502 |
|
MD5 | 97b7e7a0d900446c6cbb0ea610e21596 |
|
BLAKE2b-256 | 79b71b281a31d8fdbdfc44af2fcb7a0750158e160f3a8c608fcb602e62be24a8 |
File details
Details for the file patchify-0.2.3-py3-none-any.whl
.
File metadata
- Download URL: patchify-0.2.3-py3-none-any.whl
- Upload date:
- Size: 6.6 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.1.5 CPython/3.8.6 Linux/5.4.0-65-generic
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 4bd4f80c83280b36c6968cb4d802bde28cd11446cc8ace94e0aa14f573fcf41b |
|
MD5 | 8628c93c912938a9a38bcf6b5bc66bb3 |
|
BLAKE2b-256 | 00ced76647d95aad98d5532b5d25801dedda788c6b5d4c7457d30ae3612132af |