The Fréchet Distance Unleashed: Approximating a Dog with a Frog

Sariel Har-Peled,Benjamin Raichel,Eliot W. Robson
2024-07-10
Abstract:We show that a minor variant of the continuous Fréchet distance between polygonal curves can be computed using essentially the same algorithm used to solve the discrete version, thus dramatically simplifying the algorithm for computing it. The new variant is not necessarily monotone, but this shortcoming can be easily handled via refinement. Combined with a Dijkstra/Prim type algorithm, this leads to a realization of the Fréchet distance (i.e., a morphing) that is locally optimal (aka locally correct), that is both easy to compute, and in practice, takes near linear time on many inputs. The new morphing has the property that the leash is always as short-as-possible. We implemented the new algorithm, and developed various strategies to get a fast execution in practice. Among our new contributions is a new simplification strategy that is distance-sensitive, and enables us to compute the exact continuous Fréchet distance in near linear time in practice. We preformed extensive experiments on our new algorithm, and released Julia and Python packages with these new implementations.
Computational Geometry
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to simplify the calculation of the continuous Fréchet distance and improve its efficiency and effectiveness in practical applications. Specifically: 1. **Simplified algorithm**: The paper shows that a small variant of the continuous Fréchet distance can be calculated by an algorithm almost identical to the discrete version, thus greatly simplifying the calculation process. Although this new variant is not necessarily monotonic, this shortcoming can be refined. 2. **Locally optimal matching**: Combined with Dijkstra/Prim - type algorithms, this method can achieve locally optimal (i.e., locally correct) Fréchet distance calculations. This kind of matching is not only easy to calculate but also close to linear time on many inputs. 3. **Natural matching**: The matching generated by the new method is more natural and better than that generated by the standard algorithm. This makes the Fréchet distance more attractive in practical applications, especially when dealing with noise and outliers. 4. **Fast execution strategy**: The paper also proposes a series of strategies to achieve fast execution in practice, including a new simplification strategy that is sensitive to distance and can calculate the exact continuous Fréchet distance in near - linear time in practice. 5. **Practical applications**: The authors implemented the new algorithm and developed various strategies to achieve fast execution. They conducted extensive experiments and released Julia and Python packages that contain the new implementation. Through these improvements, the paper aims to make the calculation of the Fréchet distance more efficient and practical, suitable for large - scale practical application scenarios.