Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.pixeltable.com/llms.txt

Use this file to discover all available pages before exploring further.

Open in Kaggle  Open in Colab  Download Notebook
This documentation page is also available as an interactive notebook. You can launch the notebook in Kaggle or Colab, or download it for use with an IDE or local Jupyter installation, by clicking one of the above links.

​
Problem

You need to fix inconsistent lighting across hundreds of images—adjusting brightness, contrast, and color saturation.

​
Solution

What’s in this recipe:
  • Adjust brightness, contrast, and saturation
  • Test adjustments before applying
  • Process multiple images in batch
You adjust brightness, contrast, and saturation for images in your table using custom UDFs that wrap Pillow’s ImageEnhance module (relies on PIL/Pillow). This lets you control enhancement levels to match your needs. You can iterate on transformations before adding them to your table. Use .select() with .collect() to preview results on sample images—nothing is stored in your table. If you want to collect only the first few rows, use .head(n) instead of .collect(). Once you’re satisfied, use .add_computed_column() to apply the adjustments to all images in your table. For more on this workflow, see Get fast feedback on transformations.

​
Setup

%pip install -qU pixeltable

​
Load images

import pixeltable as pxt
from PIL import ImageEnhance

# Create a fresh directory (drop existing if present)
pxt.drop_dir('image_demo', force=True)
pxt.create_dir('image_demo')

Connected to Pixeltable database at: postgresql+psycopg://postgres:@/pixeltable?host=/Users/pjlb/.pixeltable/pgdata
Created directory ‘image_demo’.
<pixeltable.catalog.dir.Dir at 0x1441f2500>
t = pxt.create_table('image_demo/enhancements', {'image': pxt.Image})

Created table ‘enhancements’.
t.insert(
    [
        {
            'image': 'https://raw.githubusercontent.com/pixeltable/pixeltable/main/docs/resources/images/000000000016.jpg'
        },
        {
            'image': 'https://raw.githubusercontent.com/pixeltable/pixeltable/main/docs/resources/images/000000000049.jpg'
        },
        {
            'image': 'https://raw.githubusercontent.com/pixeltable/pixeltable/main/docs/resources/images/000000000036.jpg'
        },
    ]
)

Inserting rows into `enhancements`: 0 rows [00:00, ? rows/s]Inserting rows into `enhancements`: 3 rows [00:00, 601.16 rows/s]
Inserted 3 rows with 0 errors.
3 rows inserted, 6 values computed.

​
Iterate: adjust brightness and contrast for a few images first

@pxt.udf
def adjust_brightness(img: pxt.Image, factor: float) -> pxt.Image:
    """Adjust brightness. factor < 1 = darker, > 1 = brighter."""
    return ImageEnhance.Brightness(img).enhance(factor)


@pxt.udf
def adjust_contrast(img: pxt.Image, factor: float) -> pxt.Image:
    """Adjust contrast. factor < 1 = lower, > 1 = higher."""
    return ImageEnhance.Contrast(img).enhance(factor)


@pxt.udf
def adjust_saturation(img: pxt.Image, factor: float) -> pxt.Image:
    """Adjust saturation. factor < 1 = less saturated, > 1 = more saturated."""
    return ImageEnhance.Color(img).enhance(factor)

# Test brightness adjustments
t.select(
    t.image,
    adjust_brightness(t.image, 0.5),
    adjust_brightness(t.image, 1.5),
).head(1)

​
Add: adjust brightness and contrast for all images in your table

# Brightness adjustments (1.0 = original)
t.add_computed_column(darker=adjust_brightness(t.image, 0.5))
t.add_computed_column(brighter=adjust_brightness(t.image, 1.5))

# Contrast adjustments
t.add_computed_column(low_contrast=adjust_contrast(t.image, 0.5))
t.add_computed_column(high_contrast=adjust_contrast(t.image, 2.0))

# Color saturation
t.add_computed_column(desaturated=adjust_saturation(t.image, 0.3))
t.add_computed_column(saturated=adjust_saturation(t.image, 2.0))

Added 3 column values with 0 errors.
Added 3 column values with 0 errors.
Added 3 column values with 0 errors.
Added 3 column values with 0 errors.
Added 3 column values with 0 errors.
Added 3 column values with 0 errors.
3 rows updated, 3 values computed.

​
View results

Compare different enhancement levels side-by-side. 
# Compare brightness levels
t.select(t.image, t.darker, t.brighter).collect()

# Compare contrast levels
t.select(t.image, t.low_contrast, t.high_contrast).collect()

# Compare saturation levels
t.select(t.image, t.desaturated, t.saturated).collect()

​
Explanation

How the enhancement technique works: The UDFs wrap PIL’s ImageEnhance module to adjust visual properties of images. Each enhancement type creates an enhancer object for the image, then applies a multiplication factor. A factor of 1.0 leaves the image unchanged, values below 1.0 decrease the property (darker, less contrast, desaturated), and values above 1.0 increase it (brighter, more contrast, saturated). You can apply different factors to the same image to create multiple variations for comparison or different use cases. To customize the UDFs:
  • Brightness factors: Use 0.5 for darker images, 1.5 for brighter, or adjust to match your lighting needs
  • Contrast factors: Use 0.5 for lower contrast, 2.0 for higher contrast, or fine-tune for image clarity
  • Saturation factors: Use 0.3 for desaturated/muted colors, 2.0 for vibrant colors, or 0.0 for complete grayscale
  • Combine adjustments: Apply multiple enhancements to create complex transformations
The Pixeltable workflow: In traditional databases, .select() just picks which columns to view. In Pixeltable, .select() also lets you compute new transformations on the fly—define new columns without storing them. This makes .select() perfect for testing transformations before you commit them. When you use .select(), you’re creating a query that doesn’t execute until you call .collect(). You must use .collect() to execute the query and return results—nothing is stored in your table. If you want to collect only the first few rows, use .head(n) instead of .collect() to test on a subset before processing your full dataset. Once satisfied, use .add_computed_column() with the same expression to persist results permanently. For more on this workflow, see Get fast feedback on transformations.

​
See also

Last modified on May 1, 2026