Resizing an image using a resampling filter can increase the quality of the output image at the cost of performance. Lower quality filters may be necessary if speed is a priority.

Comparing Resampling Filters (Pillow)

There are multiple resampling filters supported by Pillow.
- PIL.Image.NEAREST: Pick the nearest pixel from the input image for the output pixel
- PIL.Image.BOX: Each pixel of the input image contributes to one pixel of the output image equally
- PIL.Image.BILINEAR: Output pixels calculated with linear interpolation
- PIL.Image.HAMMING: Produces a sharper image
- PIL.Image.BICUBIC: Output pixels calculated with cubic interpolation
- PIL.Image.LANCZOS: Output pixels calculated with a high-quality Lanczos filter

Saint Olaf College

Image instances have a resize() function which accepts a tuple that contains the desired width and height. Optionally, a resample filter to be applied to the image and a 4-tuple bounding box that provides a specific region of the image to scale can be passed as arguments.
```python
size = (newwidth, newheight)
image.resize(size, resample=None, box=None)
```

Resizing Images (Pillow)

https://pillow.readthedocs.io/en/stable/index.html

Learn Before

Related

Learn After