tb_lf_synthetic_aperture — 2D typed op

Data kinds: lightfieldimage

Call: fullseye.apply(img, "tb_lf_synthetic_aperture", a=0.5, b=0.5) (the 2-D model is one image plus two scalar knobs a,b∈[0,1])

*No figure: this op takes lightfield as input. A Studio program starting from an image cannot reach that type — see the runnable examples below for how it is used.*

Usage

Refocus through a shaped aperture — and, with `reduce="median"`, through occluders.

Same shift-and-add geometry as :func:lf_refocus, with two additions:

• *mask* — a `(V, U) weight array from :func:lf_aperture_mask` (or your

own). A small mask is a stopped-down aperture: less defocus blur, less

light. `None` weights every view equally.

• *reduce* — how the aligned views are combined. `mean` is the classical

(and linear) synthetic aperture. `median` is the interesting one: when

a foreground occluder covers a *minority* of the views at a pixel, the

median rejects it and the background behind it is reconstructed —

looking through a fence, or through a rack of parts. `max / min`

are the order-statistic extremes, useful for specular / shadow work.

`median, max and min` use the mask only to *select* views

(weight > 0), because an order statistic has no meaningful weighting;

that is stated here rather than silently ignoring the weights.

The see-through result is not a metaphor. When fewer than half the views

are blocked at a hidden pixel, more than half of them carry the identical

background sample and the median is that sample exactly. Measured

2026-09-01 on a 9x9x64x64 field with an occluder covering 25% of the centre

view at slope 3.0 (blocking at most 46% of the views at any hidden pixel):

RMS against the true, hidden background was 0.0 for `median` and

0.159 for `mean`, with the centre view itself at 0.280. Push the coverage

to 35% (up to 60% of views blocked) and the guarantee is gone — the median

lands at 0.133, worse than nothing.

Returns a `(H, W)` 2-D image.

Raises `ValueError`: *lf* not a valid light field, a *mask* whose

shape is not `(V, U)` or which is non-finite / negative / selects no view,

a non-finite or over-large *slope*, unknown *reduce* / *interp* / *edge*.

Typed bridge of the lightfield op `lf_synthetic_aperture into the 2-D evolution registry: the same implementation, called under the op(v, a, b) convention. This op has no tunable parameter; a and b` are unused.

References (sample data, literature)

• 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.

Runnable examples (verified samples that actually call this op)

• (none yet)

Ops the type connects to (they accept image as input)

identity · gaussian · mean_box · bilateral · unsharp · median · min_filter · max_filter

Same category (typed)

tb_points_to_voxel · tb_estimate_point_normals · tb_iss_keypoints · tb_angle_3points · tb_project_points · tb_render_point_depth · tb_statistical_outlier_removal · tb_radius_outlier_removal


*Provenance: ops.py — 2D 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.