---
title: More Informed RANSAC (MI-RANSAC)
url: https://www.emergentmind.com/topics/more-informed-ransac-mi-ransac
type: topic
---

# More Informed RANSAC (MI-RANSAC)

More-Informed RANSAC (MI-RANSAC) is a methodological enhancement to the conventional Random Sample Consensus (RANSAC) framework applied in minimal solver problems for multiview geometry, specifically for the estimation of essential and fundamental matrices in Structure-from-Motion (SfM) pipelines. MI-RANSAC leverages a mathematically principled pre-test to filter out ill-conditioned minimal samples, thereby improving both computational efficiency and hypothesis robustness without measurably compromising accuracy [2112.14651].

## 1. Riemannian-Manifold Conditioning and Minimal Problems

The framework situates the minimal problems—namely, the 5-point (essential matrix, E) and 7-point (fundamental matrix, F)—within Riemannian manifolds:

- $\W$: world‐scene manifold
- $\X$: image‐data manifold, being either $(\mathbb{R}^2 \times \mathbb{R}^2)^5$ (5-point) or $(\mathbb{R}^2 \times \mathbb{R}^2)^7$ (7-point)
- $\Y$: epipolar manifold (the set of E or F matrices as appropriate, each a submanifold of $\mathbb{P}(\mathbb{R}^{3\times3})$)

The forward map $\Phi : \W \dashrightarrow \X$ and the epipolar map $\Psi : \W \dashrightarrow \Y$ formalize the generative process. The solution map
\[
\S = \Psi \circ \Theta : \; \X \supset U \;\to\; \Y, \quad \Theta = \Phi^{-1}
\]
is defined wherever $D\Phi(w)$ is invertible. The local conditioning of the solution is characterized by the operator norm:
\[
\kappa(\S, x) = \| D\Psi(w) \circ D\Phi(w)^{-1} \|, \quad w = \Theta(x)
\]
where $\|\cdot\|$ is induced by the Riemannian metrics on $\X$ and $\Y$.

For the 5-point problem, $D\Phi$ is a $20\times20$ Jacobian; the corresponding condition number is
\[
\kappa_E(x) = \sigma_{\max}\left(
G^{1/2} \begin{bmatrix} 0_{5\times15} & I_{5\times5} \end{bmatrix}
(D\Phi(x))^{-1}
\right)
\]
with $G^{1/2}$ the $5\times5$ Gram root. In the 7-point variant, an analogous formula involves a $28\times28$ Jacobian and a $7\times7$ Gram root $H^{1/2}$ [2112.14651, Prop. 3.1–3.2].

## 2. Geometric Characterization of Ill-Posed Loci

Configurations are classified as ill-posed if $D\Phi(w)$ is singular for some corresponding world scene $w$.

- **World Scene Description**:
    - 5-point: Ill-posed if there exists a rectangular ruled quadric $\mathcal{Q} \subset \mathbb{R}^3$ through the $X_i$, containing the baseline $\ell = \operatorname{Span}(-R^\top t)$, and such that all planes orthogonal to $\ell$ intersect $\mathcal{Q}$ in circles.
    - 7-point: Ill-posed if a quadric $\mathcal{Q}$ exists through the seven 3D points, containing the corresponding baseline $\ell$ [Thms 3.3, 3.4].

- **Image Data Description & X.5-Point Curves**:
    - 5-point: The last point $y_5$ lies on a degree-30 real algebraic 4.5-point curve in the image space if and only if the configuration is ill-posed. The defining polynomial $\mathcal{P}$ is bidegree 30 in $(x_1,y_1,\dots,x_5,y_5)$.
    - 7-point: The 7th point $y_7$ lies on a degree-6 algebraic 6.5-point curve if and only if ill-posed; this can be computed by homotopy or an explicit 1668-term Plücker polynomial [Thms 3.5, 3.6].

The distance from the last correspondence to its respective X.5-point curve, $d(y_k, \text{curve})$, serves as a robust numerical indicator of proximity to the ill-posed locus.

## 3. MI-RANSAC Sampling and Pre-Testing Procedure

Rather than sampling minimal subsets indiscriminately, MI-RANSAC incorporates a conditioning pre-test: only subsets sufficiently far from the ill-posed locus are admitted for solver evaluation. This process is summarized for the calibrated (5-point) case as follows:

- **Inputs**: Set of correspondences $C = \{(x_i, y_i)\}_{i=1}^N$, inlier threshold $\tau_\mathrm{in}$, pre-test distance threshold $\delta_\mathrm{min}$, max hypotheses $M$.
- For each random 5-subset $S$:
  1. Split $S$ into first $4.5$ correspondences and the last point.
  2. Compute the 4.5-point curve given the first $4.5$ points.
  3. Evaluate distance $d(y_5, \text{curve}_{4.5})$.
  4. If $d < \delta_\mathrm{min}$, reject $S$; otherwise, solve the 5-point problem.
  5. For each real root, score the model and retain the hypothesis with maximal inlier count.

The uncalibrated (7-point) procedure analogously uses the 6.5-point curve and a 7th correspondence. Pseudocode for the full process is provided in [2112.14651], including explicit handling of the curve computation and distance evaluation.

## 4. Empirical Analysis: Efficacy and Speed-up

Synthetic experiments demonstrate that:

- For 7-point (fundamental matrix), with $\delta_{\min} \approx 5$ px, approximately 90% of ill-conditioned samples are rejected based on the X.5-point distance pre-test, with only ≈10% of stable samples inadvertently rejected.
- For 5-point (essential matrix), a threshold of 2 px similarly distinguishes almost all ill-posed samples from stable ones.

Overall, MI-RANSAC requires roughly half as many minimal-problem solves to generate the same count of well-conditioned hypotheses, yielding computation speed-ups of approximately $1.5 \times$ to $2 \times$ for the hypothesis-generation stage. The computational load of evaluating the pre-test polynomial ($\sim0.01\,\mathrm{ms}$) is negligible compared to the roots-solving step ($0.4$–$1\,\mathrm{ms}$) [2112.14651, Fig. 9].

## 5. Practical Implementation Details

Key considerations for integration include:

- **Curve Stability**: The X.5-point curves exhibit sub-pixel stability ($\ll 1$ px movement) when the fixed correspondences are perturbed with realistic noise levels ($\sigma \leq 0.5$ px), ensuring pre-test robustness [Fig. 10].
- **Polynomial Evaluation**: For fundamental matrix estimation, evaluating the 1668-term Plücker polynomial requires at most $2\,\mathrm{kFLOPs}$, with integer coefficients $< 100$. For the essential matrix, homotopy computation of the high-degree curve is an offline cost; runtime evaluation is reduced to evaluating a univariate degree-30 polynomial.
- **Threshold Selection**: Empirically, $\delta_\mathrm{min} \approx 5$ px (uncalibrated) or $2$ px (calibrated) divides degenerate from stable samples with classification error $< 10\%$.
- **Alternative Condition Number Estimation**: Approximations of $\kappa$ via local linear matrix inequalities (LMIs) are feasible, but the X.5-point distance is faster and more direct in practice.
- **Integration**: MI-RANSAC is orthogonal to other RANSAC accelerations such as PROSAC or guided sampling and can be integrated into any two-view step within an SfM pipeline.

## 6. Theoretical Context and Implications

The rigorous condition-number formulation draws on Riemannian manifold techniques and aligns with the geometric characterization of minimal solvers’ stability. The investigation of ill-posed loci both in the world scene and in image data, as made explicit by X.5-point algebraic curves, provides a foundational justification for MI-RANSAC’s early rejection mechanism. The distinct separation—quantified in pixels—between stable and unstable configurations underpins the reliability of the pre-test.

A plausible implication is that this methodology could be ported to other minimal-problem domains or serve as a pre-processing layer in non-minimal solvers, provided comparable conditioning criteria can be formulated.

## 7. Significance and Integration in Structure-from-Motion Pipelines

By precluding near-degenerate samples before minimal solver invocation, MI-RANSAC maintains standard RANSAC’s robustness while delivering a $1.5$–$2\times$ speed-up in hypothesis generation at negligible accuracy loss. This renders it a practical, theory-backed augmentation for SfM and other multiview geometry pipelines reliant on minimal solvers, with immediate empirical and algorithmic impact [2112.14651].

Source: https://www.emergentmind.com/topics/more-informed-ransac-mi-ransac