Fast Iterative Region Inflation for Computing Large 2-D/3-D Convex Regions of Obstacle-Free Space
About
Convex polytopes have compact representations and exhibit convexity, which makes them suitable for abstracting obstacle-free spaces from various environments. Existing generation methods struggle with balancing high-quality output and efficiency. Moreover, another crucial requirement for convex polytopes to accurately contain certain seed point sets, such as a robot or a front-end path, is proposed in various tasks, which we refer to as manageability. In this paper, we propose Fast Iterative Regional Inflation (FIRI) to generate high-quality convex polytope while ensuring efficiency and manageability simultaneously. FIRI consists of two iteratively executed submodules: Restrictive Inflation (RsI) and Maximum Volume Inscribed Ellipsoid (MVIE) computation. By explicitly incorporating constraints that include the seed point set, RsI guarantees manageability. Meanwhile, iterative MVIE optimization ensures high-quality result through monotonic volume bound improvement.In terms of efficiency, we design methods tailored to the low-dimensional and multi-constrained nature of both modules, resulting in orders of magnitude improvement compared to generic solvers. Notably, in 2-D MVIE, we present the first linear-complexity analytical algorithm for maximum area inscribed ellipse, further enhancing the performance in 2-D cases. Extensive benchmarks conducted against state-of-the-art methods validate the superior performance of FIRI in terms of quality, manageability, and efficiency. Furthermore, various real-world applications showcase the generality and practicality of FIRI.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Convex Free-Space Approximation | Maze Point Seed | Time (ms)2.66 | 8 | |
| Convex Free-Space Approximation | Maze Line Seed | Time (ms)3.13 | 8 | |
| Convex Free-Space Approximation | Obstacle Point Seed | Time (ms)6.99 | 8 | |
| Convex Free-Space Approximation | Obstacle Line Seed | Time (ms)8.54 | 8 | |
| Convex Free-Space Approximation | Real Point Seed | Time (ms)2.26 | 4 | |
| Convex Free-Space Approximation | Real Line Seed | Execution Time (ms)3.45 | 4 |