Litcius/Paper detail

Point Cloud Sampling via Graph Balancing and Gershgorin Disc Alignment

Chinthaka Dinesh, Gene Cheung, Ivan V. Bajić

2022IEEE Transactions on Pattern Analysis and Machine Intelligence16 citationsDOI

Abstract

Point cloud (PC)—a collection of discrete geometric samples of a 3D object’s surface—is typically large, which entails expensive subsequent operations. Thus, PC sub-sampling is of practical importance. Previous model-based sub-sampling schemes are ad-hoc in design and do not preserve the overall shape sufficiently well, while previous data-driven schemes are trained for specific pre-determined input PC sizes and sub-sampling rates and thus do not generalize well. Leveraging advances in graph sampling, we propose a fast PC sub-sampling algorithm of linear time complexity that chooses a 3D point subset while minimizing a global reconstruction error. Specifically, to articulate a sampling objective, we first assume a super-resolution (SR) method based on feature graph Laplacian regularization (FGLR) that reconstructs the original high-res PC, given points chosen by a sampling matrix <inline-formula><tex-math notation="LaTeX">${\mathbf H}$</tex-math></inline-formula> . We prove that minimizing a worst-case SR reconstruction error is equivalent to maximizing the smallest eigenvalue <inline-formula><tex-math notation="LaTeX">$\lambda _{\min }$</tex-math></inline-formula> of matrix <inline-formula><tex-math notation="LaTeX">${\mathbf H}^{\top } {\mathbf H}+ \mu {\boldsymbol{\mathcal{L}}}$</tex-math></inline-formula> , where <inline-formula><tex-math notation="LaTeX">${\boldsymbol{\mathcal{L}}}$</tex-math></inline-formula> is a symmetric, positive semi-definite matrix derived from a neighborhood graph connecting the 3D points. To arrive at a fast algorithm, instead of maximizing <inline-formula><tex-math notation="LaTeX">$\lambda _{\min }$</tex-math></inline-formula> , we maximize a lower bound <inline-formula><tex-math notation="LaTeX">$\lambda ^-_{\min }({\mathbf H}^{\top } {\mathbf H}+ \mu {\boldsymbol{\mathcal{L}}})$</tex-math></inline-formula> via selection of <inline-formula><tex-math notation="LaTeX">${\mathbf H}$</tex-math></inline-formula> —this translates to a graph sampling problem for a signed graph <inline-formula><tex-math notation="LaTeX">${\mathcal G}$</tex-math></inline-formula> with self-loops specified by graph Laplacian <inline-formula><tex-math notation="LaTeX">${\boldsymbol{\mathcal{L}}}$</tex-math></inline-formula> . We tackle this general graph sampling problem in three steps. First, we approximate <inline-formula><tex-math notation="LaTeX">${\mathcal G}$</tex-math></inline-formula> with a balanced graph <inline-formula><tex-math notation="LaTeX">${\mathcal G}_B$</tex-math></inline-formula> specified by Laplacian <inline-formula><tex-math notation="LaTeX">${\boldsymbol{\mathcal{L}}}_B$</tex-math></inline-formula> . Second, leveraging a recent linear algebraic theorem called Gershgorin disc perfect alignment (GDPA), we perform a similarity transform <inline-formula><tex-math notation="LaTeX">${\boldsymbol{\mathcal{L}}}_p~=~{\mathbf S}{\boldsymbol{\mathcal{L}}}_B {\mathbf S}^{-1}$</tex-math></inline-formula> , so that all Gershgorin disc left-ends of <inline-formula><tex-math notation="LaTeX">${\boldsymbol{\mathcal{L}}}_p$</tex-math></inline-formula> are aligned exactly at <inline-formula><tex-math notation="LaTeX">$\lambda _{\min }({\boldsymbol{\mathcal{L}}}_B)$</tex-math></inline-formula> . Finally, we choose samples on <inline-formula><tex-math notation="LaTeX">${\mathcal G}_B$</tex-math></inline-formula> using a previous graph sampling algorithm to maximize <inline-formula><tex-math notation="LaTeX">$\lambda ^-_{\min }({\mathbf H}^{\top } {\mathbf H}+ \mu {\boldsymbol{\mathcal{L}}}_p)$</tex-math></inline-formula> in linear time. Experimental results show that 3D points chosen by our algorithm outperformed competing schemes both numerically and visually in reconstruction quality.

Topics & Concepts

Computer scienceCloud computingPoint cloudGraphArtificial intelligenceGraph theoryComputer visionAlgorithmTheoretical computer scienceMathematicsCombinatoricsOperating system3D Shape Modeling and AnalysisRemote Sensing and LiDAR ApplicationsRobotics and Sensor-Based Localization