isp op• Data kinds: image2d → image2d
• Call: import fullseye as fs; fs.ledger.lens_shading_gain(flat, pattern=None, smooth_sigma=0.0) (to call the implementation directly, import gfx2d; gfx2d.lens_shading_gain(flat, pattern=None, smooth_sigma=0.0); from the registry, opsgfx2d.get("lens_shading_gain"))
The gain map that flattens a flat-field frame: `gain = mean / flat per colour channel (or for the whole frame when *pattern* is None`).
Shoot a uniform white target; vignetting and the micro-lens fall-off make
the corners darker. Multiplying any later frame by this map undoes that.
*smooth_sigma* > 0 Gaussian-smooths the flat first (per channel) so sensor
noise and dust do not become gain speckle. The gain is normalised so the
channel mean stays 1 — the map corrects shape, not exposure.
Ground truth: `lens_shading_correct(flat, lens_shading_gain(flat))` is a
constant frame (each channel equal to its own mean) to 1e-12.
Raises `ValueError`: non-finite input, a zero or negative pixel in
the flat (no gain can be defined there), or a negative *smooth_sigma*.
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• raw_to_display_isp — py -3.11 examples/raw_to_display_isp.py
image2d as input)shadow_cast_2d · dither · raw_black_level · raw_dead_pixel_mask · raw_dead_pixel_correct · lens_shading_correct · raw_apply_gains · raw_demosaic_bilinear
isp)raw_black_level · raw_dead_pixel_mask · raw_dead_pixel_correct · lens_shading_correct · awb_gains · rgb_apply_gains · raw_apply_gains · raw_demosaic_bilinear
*Provenance: gfx2d.py — GFX2D operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.