P
Physics World Model
← All principles

Multi-Focus / Panorama Fusion (all-in-focus composite imaging) L1-076

Computational PhotographyImage fusion and stitchingδ=3 · standardL_DAG = 3📋 Stub — not mineable
📋

Unclaimed Principle — open for contribution

This Principle is declared in the catalog but has no reference solver, no pinned dataset, and is not registered on-chain. There is no reward pool. Submitting a cert against this Principle today will record the cert for reproducibility but pay zero PWM.

To claim it as a Bounty #7 contribution: open a PR adding (1) a reference solver, (2) ≥1 dataset pinned to IPFS, (3) updates to the L3 manifest with dataset CIDs. After verifier-agent triple-review, the founders' 3-of-5 multisig signs PWMRegistry.register() and the Principle becomes mineable.

Forward model E

A sequence of K images {y_k} of the same scene is captured with varying focus distance z_k (multi-focus) or camera pose (panorama). Each y_k blurs the 3D scene x(u,v,z) through a depth-dependent PSF_k and an affine warp W_k. Fusion recovers an all-in-focus, globally aligned image X.

L-DAG

S.scan.pose -> K.psf.defocus -> L.warp.affine -> int.spatial
S.scan.poseK.psf.defocusL.warp.affineint.spatial

Well-posedness W

Existence:
true
Uniqueness:
true
Stability:
conditional
κ:
800

Well-conditioned when scenes are Lambertian, inter-frame parallax is small (pure rotation or small translation), and focus distances span the scene depth range. Degrades catastrophically for moving subjects (ghosting) and unconstrained parallax (depth ambiguity).

Solvability C

Solver class:
Laplacian-pyramid fusion, gradient-based focus measure (SML, ML, tenengrad), Poisson-blend stitching, learned fusion (MFIF-GAN, U2Fusion), graph-cut seam selection
Convergence rate q:
2
Complexity:
O(K * H * W * log(max(H,W))) for pyramid; O(K * H * W) for gradient fusion

Specs (0)

No L2 specs registered yet for this principle.