pyflink.multimodal.operators.image_sharpness#

image_sharpness(*columns, max_edge=None, allow_upscale=False, concurrency=None)[source]#

Compute a raw sharpness metric using Laplacian variance.

The result is in [0, +inf) and larger values usually indicate sharper images. This is a low-cost focus measure, not an absolute quality score: thresholds depend on image scale, texture, contrast, and noise.

Parameters

*columns – Optional image columns. When provided, the UDF is applied directly (syntactic sugar).
max_edge – Cap the longest edge to this many pixels before computing sharpness. None (default) preserves the original scale.
allow_upscale – When max_edge is set, allow upscaling smaller images. Default False.
concurrency – UDF concurrency. None uses the framework default.

Returns

A UDF producing float64.

Example::

>>> # As a reusable variable
>>> sharpness = image_sharpness(max_edge=512)
>>> df = df.with_columns("sharpness", sharpness(col("img")))
>>>
>>> # Inline
>>> df = df.with_columns("sharpness", image_sharpness(col("img"), max_edge=512))