Abstract
The line geometric model of 3D projective geometry offers a powerful correspondence where the Lie algebra $sl(4)$ of projective transformations is isomorphic to the bivector algebra of $Cl(3,3)$. However, the standard canonical homomorphism from $SL(4)$ to $Spin(3,3)$ is not surjective, and transformations with negative determinants—such as polarities—do not naturally induce orthogonal transformations in Plücker coordinate space. This paper establishes a rigorous algebraic framework using Clifford algebra to resolve these defects. We introduce the group $Pin^{sp}(3,3)$ to provide a quadruple-cover of 3D projective transformations and polarities. Furthermore, we develop factored spinors for generating rigid-body motions and extend classical screw theory into a generalized "superscrew" algebra for the 15-dimensional Lie algebra $sl(4)$, defining associated cross products and virtual work principles.
Introduction
In the study of 3D projective geometry, $SL(4)$—the group of $4 \times 4$ matrices with determinant 1—is the standard tool for describing transformations. While the isomorphism between its Lie algebra and the bivectors of $Cl(3,3)$ is well-established, the practical application of this mapping via the Plücker correspondence suffers from significant theoretical gaps. Specifically, the map from $SL(4)$ to $Spin(3,3)$ fails to capture the full range of projective transformations, and the framework struggles to accommodate polarities (which swap points and planes) because they involve negative determinants.
This research addresses these limitations by shifting the foundational focus from points in $\mathbb{R}^4$ to lines in $\mathbb{R}^{3,3}$. By utilizing Geometric Algebra (GA), we provide a unified framework that treats reflections, rigid-body motions, and polarities with consistent algebraic elegance.
The $Pin^{sp}(3,3)$ Framework
The core contribution of this work is the proposal of the $Pin^{sp}(3,3)$ group. By defining this group with $Pin(3,3)$ as its normal subgroup, we achieve a quadruple-cover of the group of 3D projective transformations. This construction effectively fixes the surjectivity issues inherent in the $Spin(3,3)$ model.
Within this framework, we utilize Witt decompositions to define points and planes as null 3-spaces (trivectors) within the 6D space of lines. This allows for the explicit recovery of point and plane coordinates from their trivector representations. To handle transformations, we construct spinors in factored form—products of invertible vectors—to generate both simple reflections and complex rigid-body motions.
From Screw Theory to Superscrew Algebra
Classical screw theory is a staple of robotics and mechanics, typically operating within the 6D Lie algebra $se(3)$. We extend this concept to the entire 15-dimensional projective Lie algebra $sl(4)$. By decomposing $sl(4)$ into five 3D vector spaces, we create a "superscrew" representation.
This extension allows us to generalize the concepts of "cross products" and "virtual work" beyond Euclidean rigid-body motions to all projective motions. We demonstrate that the "Wrench" (force and torque) and "Twist" (velocity and angular velocity) models from classical mechanics are specific instances of this broader projective geometric structure. The resulting generalized cross products for various 6D Lie subalgebras show that the virtual work of a wrench is intrinsically linked to the algebraic structure of $Cl(3,3)$.
Key Results and Implications
Our framework provides a complete classification of projective transformations induced by one or two reflections in line space. By treating polarities as reflections within the $Pin$ group, we unify the treatment of duality in projective geometry.
Key achievements include:
- Complete Coverage: The $Pin^{sp}(3,3)$ group successfully maps to all 3D projective transformations, including those previously unreachable.
- Algebraic Duality: Points and planes are handled symmetrically as trivectors, simplifying the geometry of polarities.
- Mechanical Unification: The framework reveals that the principles of virtual work and screw theory are fundamental properties of the projective Lie algebra.
Future Directions
While this work provides a rigorous theoretical foundation, several avenues for future research remain. The connection between this $Cl(3,3)$ model and the $Cl(4,1)$ Conformal Geometric Algebra (CGA) commonly used in computer vision warrants a detailed mapping. Additionally, the development of computational libraries for "superscrew" operations will be essential for applying this theory to practical engineering challenges, such as modeling pin-hole cameras or exploring non-Newtonian mechanics.
Summary updated