Visual Curve Completion and Rotational Surfaces of Constant Negative Curvature

Alvaro Pampano

arxiv: 1907.05696 · v1 · pith:4ZKDNOVSnew · submitted 2019-06-27 · 🧮 math.DG

Visual Curve Completion and Rotational Surfaces of Constant Negative Curvature

Alvaro Pampano This is my paper

Pith reviewed 2026-05-25 14:27 UTC · model grok-4.3

classification 🧮 math.DG

keywords visual curve completionsub-Riemannian geodesicstotal curvature energyrotational surfacesconstant negative curvaturebinormal flowvariational problem

0 comments

The pith

Extremal curves of total curvature energy in the plane correspond one-to-one with rotational surfaces of constant negative curvature in three-space via sub-Riemannian lifts and binormal evolution.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper solves a variational problem for curves that minimize a total curvature energy, lifts those curves to geodesics in the unit tangent bundle of the plane, and shows that the same curves generate rotational surfaces of constant negative curvature when evolved by their binormal flow. It then proves that every such surface is locally foliated by these curves. The result establishes an exact match between the geometric objects used for visual contour completion and a classical family of surfaces in R3. If correct, this supplies a concrete differential-geometric model that unifies two previously separate problems.

Core claim

There exists a one-to-one correspondence between the sub-Riemannian geodesics used by the brain for visual curve completion and the rotational surfaces of constant negative curvature in R3, obtained by lifting extremal curves of a total curvature energy and evolving them under the binormal flow with prescribed velocity.

What carries the argument

Extremal curves of a total curvature energy, lifted to sub-Riemannian geodesics and evolved by binormal flow, which foliate the rotational surfaces.

If this is right

The variational problem of finding extremal curves for the total curvature energy is solved geometrically by reduction to the geometry of the unit tangent bundle.
Rotational surfaces of constant negative curvature are constructed explicitly by evolving the extremal curves under the associated binormal flow.
Locally, every rotational surface of constant negative curvature is foliated by these extremal curves.
The sub-Riemannian geodesics arising in visual completion are precisely the lifts of the curves that foliate the surfaces.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The correspondence supplies a geometric dictionary that could translate results about surface foliations into statements about perceptual completions and vice versa.
If the binormal flow preserves the extremal property, one could generate families of surfaces from known visual-completion curves without solving the surface equation directly.
The construction may extend to non-rotational surfaces of constant negative curvature if the same lifting and flow mechanism applies.

Load-bearing premise

The brain completes missing contours using sub-Riemannian geodesics of the unit tangent bundle that arise as lifts of extremal curves of a total curvature type energy.

What would settle it

A single rotational surface of constant negative curvature in R3 whose rulings are not extremal curves of the total curvature energy would break the claimed foliation and therefore the correspondence.

read the original abstract

If a piece of the contour of a picture is missing to the eye vision, then the brain tends to complete it using some kind of sub-Riemannian geodesics of the unit tangent bundle of the plane, R2xS1. These geodesics can be obtained by lifting extremal curves of a total curvature type energy in the plane. We completely solve this variational problem, geometrically. Moreover, we also show a way of constructing rotational surfaces of constant negative curvature in R3 by evolving these extremal curves under their associated binormal flow with prescribed velocity. Finally, we prove that, locally, all rotational constant negative curvature surfaces of R3 are foliated by extremal curves of these energies. Therefore, we conclude that there exists a one-to-one correspondence between the sub-Riemannian geodesics used by the brain for visual curve completion and these rotational surfaces of R3.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives a geometric solution to the variational problem and links the resulting curves to rotational constant-negative-curvature surfaces via binormal flow and local foliation, but the global one-to-one correspondence does not follow directly from the local statements.

read the letter

The paper solves the variational problem for extremal curves of a total-curvature-type energy geometrically. It then evolves those curves under the binormal flow with prescribed velocity to produce rotational surfaces of constant negative curvature in R3, and it proves that every such surface is locally foliated by the curves. Those two steps are the concrete new material: an explicit construction in one direction and a local foliation result in the other. Both are stated clearly enough to be checked against the cited literature on the energies and on constant-curvature surfaces. The connection to sub-Riemannian geodesics in the visual-completion model supplies the motivation but is not itself the technical contribution. The soft spot is the final claim of a one-to-one correspondence. The foliation is only local and the construction is presented as one method rather than a classification, so it is not immediate that the two maps are inverses on the level of complete geodesics versus complete surfaces. The abstract jumps to the global statement without additional arguments that would close the gap. Readers already working in sub-Riemannian geometry or in the classical theory of surfaces of constant curvature will find the constructions and the foliation useful. The vision link is a secondary motivation that narrows the audience further. The paper is coherent on its own terms and the claims are specific enough to be refereed. I would send it out for review so the details around the correspondence can be examined directly.

Referee Report

1 major / 0 minor

Summary. The manuscript solves geometrically the variational problem of extremal curves for a total-curvature-type energy in the plane; these curves lift to sub-Riemannian geodesics in R²×S¹ that model visual curve completion. It constructs rotational surfaces of constant negative curvature in R³ by binormal evolution of the extremal curves with prescribed velocity, proves that every such surface is locally foliated by the extremal curves, and concludes that a one-to-one correspondence therefore exists between the geodesics and the surfaces.

Significance. If the claimed global one-to-one correspondence can be established, the work would link sub-Riemannian models of visual perception with the classical differential geometry of surfaces of constant negative curvature, supplying explicit geometric constructions that could be useful in both areas. The geometric solution of the variational problem is a positive feature when the derivations are fully detailed.

major comments (1)

[Abstract] Abstract (final sentence): the local foliation result and the construction of surfaces 'by evolving these extremal curves under their associated binormal flow' are presented as sufficient to conclude a one-to-one correspondence between sub-Riemannian geodesics in R²×S¹ and complete rotational surfaces in R³. Because the foliation statement is explicitly local and the construction is described only as 'a way' rather than a classification, the passage from local data to the global bijective identification required by the visual-completion claim is not immediate and needs explicit justification.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading of the manuscript and the constructive feedback. We address the single major comment below.

read point-by-point responses

Referee: [Abstract] Abstract (final sentence): the local foliation result and the construction of surfaces 'by evolving these extremal curves under their associated binormal flow' are presented as sufficient to conclude a one-to-one correspondence between sub-Riemannian geodesics in R²×S¹ and complete rotational surfaces in R³. Because the foliation statement is explicitly local and the construction is described only as 'a way' rather than a classification, the passage from local data to the global bijective identification required by the visual-completion claim is not immediate and needs explicit justification.

Authors: We agree that the abstract's final sentence overstates the scope by concluding a one-to-one correspondence. The manuscript provides a construction of rotational surfaces of constant negative curvature via binormal evolution of the extremal curves and proves that every such surface is locally foliated by these curves. These results establish a local geometric link but do not include a global classification or bijection between all sub-Riemannian geodesics and all complete rotational surfaces. We will revise the abstract to state the results precisely: the local foliation property and the constructive method, without claiming a global identification. This revision will also better align the abstract with the local nature of the visual curve completion model. revision: yes

Circularity Check

0 steps flagged

Derivation chain self-contained; no circular reductions identified

full rationale

The paper solves the variational problem for extremal curves of the total-curvature energy, constructs rotational surfaces of constant negative curvature via binormal flow evolution of those curves, and establishes a local foliation result showing every such surface is foliated by the extremal curves. The one-to-one correspondence is presented as following from these geometric constructions and the local foliation theorem. No self-definitional steps, fitted parameters renamed as predictions, load-bearing self-citations, or ansatzes smuggled via citation are present. The local-vs-global distinction noted by the skeptic is a potential completeness issue but does not reduce any claimed derivation to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard differential geometry and variational calculus; no free parameters, ad-hoc axioms, or invented entities are introduced in the abstract.

axioms (1)

standard math Standard axioms of Riemannian and sub-Riemannian geometry together with the existence and regularity theory for variational problems on curves.
Invoked implicitly when lifting extremal curves to the unit tangent bundle and when applying the binormal flow.

pith-pipeline@v0.9.0 · 5673 in / 1186 out tokens · 28291 ms · 2026-05-25T14:27:05.776571+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

we prove that, locally, all rotational constant negative curvature surfaces of R3 are foliated by extremal curves of these energies. Therefore, we conclude that there exists a one-to-one correspondence...
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel echoes

?

echoes
ECHOES: this paper passage has the same mathematical shape or conceptual pattern as the Recognition theorem, but is not a direct formal dependency.

the curvature of γ is given by κ²(s)=a²(d−a²)f²(as)/(a²−(d−a²)f²(as)) where f(z)=sinh z, cosh z or e^z

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.