Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis

Xupeng Wang, Ferdous Sohel, Mohammed Bennamoun, Hang Lei

    Research output: Chapter in Book/Conference paperConference paper

    2 Citations (Scopus)

    Abstract

    This paper presents an efficient feature point descriptor for non-rigid shape analysis. The descriptor is developed based on the properties of the heat diffusion process on a shape. We use, for the first time, the Heat Kernel Signature of a particular time scale to define the scalar field on a manifold. Then, motivated by the successful use of a local reference frame for rigid shape analysis, we construct a repetitive local polar coordinate system, which is invariant under isometric deformations. Finally, a binary descriptor is derived by comparing the intensities of the neighboring points for each feature point. We show that the descriptor is highly discriminative and can be computed simply using ‘intensity comparisons’ on a shape. Furthermore, its similarity can be evaluated using the Hamming distance, which is very efficient to compute compared with the commonly used L2L2 norm. Our experiments demonstrate a superior performance compared to existing techniques on the standard benchmark TOSCA.
    Original languageEnglish
    Title of host publicationPacific-Rim Symposium on Image and Video Technology
    EditorsThomas Bräunl, Brendan McCane, Mariano Rivera, Xinguo Yu
    Place of PublicationUSA
    PublisherSpringer
    Pages751-761
    Number of pages11
    ISBN (Print)9783319294513
    DOIs
    Publication statusPublished - 2016
    Event7th Pacific-Rim Symposium on Image and Video Technology: PSIVT 2015 - Auckland, New Zealand
    Duration: 23 Nov 201527 Nov 2015

    Publication series

    NameLecture Notes in Computer Science
    Volume9431

    Conference

    Conference7th Pacific-Rim Symposium on Image and Video Technology
    Abbreviated titlePSIVT 2015
    CountryNew Zealand
    CityAuckland
    Period23/11/1527/11/15

    Fingerprint

    Hamming distance
    Experiments
    Hot Temperature

    Cite this

    Wang, X., Sohel, F., Bennamoun, M., & Lei, H. (2016). Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis. In T. Bräunl, B. McCane, M. Rivera, & X. Yu (Eds.), Pacific-Rim Symposium on Image and Video Technology (pp. 751-761). (Lecture Notes in Computer Science ; Vol. 9431). USA: Springer. https://doi.org/10.1007/978-3-319-29451-3_59
    Wang, Xupeng ; Sohel, Ferdous ; Bennamoun, Mohammed ; Lei, Hang. / Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis. Pacific-Rim Symposium on Image and Video Technology. editor / Thomas Bräunl ; Brendan McCane ; Mariano Rivera ; Xinguo Yu. USA : Springer, 2016. pp. 751-761 (Lecture Notes in Computer Science ).
    @inproceedings{b05b68e4886f4eb3a6bf7fe3c94166b3,
    title = "Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis",
    abstract = "This paper presents an efficient feature point descriptor for non-rigid shape analysis. The descriptor is developed based on the properties of the heat diffusion process on a shape. We use, for the first time, the Heat Kernel Signature of a particular time scale to define the scalar field on a manifold. Then, motivated by the successful use of a local reference frame for rigid shape analysis, we construct a repetitive local polar coordinate system, which is invariant under isometric deformations. Finally, a binary descriptor is derived by comparing the intensities of the neighboring points for each feature point. We show that the descriptor is highly discriminative and can be computed simply using ‘intensity comparisons’ on a shape. Furthermore, its similarity can be evaluated using the Hamming distance, which is very efficient to compute compared with the commonly used L2L2 norm. Our experiments demonstrate a superior performance compared to existing techniques on the standard benchmark TOSCA.",
    author = "Xupeng Wang and Ferdous Sohel and Mohammed Bennamoun and Hang Lei",
    year = "2016",
    doi = "10.1007/978-3-319-29451-3_59",
    language = "English",
    isbn = "9783319294513",
    series = "Lecture Notes in Computer Science",
    publisher = "Springer",
    pages = "751--761",
    editor = "Br{\"a}unl, {Thomas } and McCane, {Brendan } and Rivera, {Mariano } and Yu, {Xinguo }",
    booktitle = "Pacific-Rim Symposium on Image and Video Technology",
    address = "Netherlands",

    }

    Wang, X, Sohel, F, Bennamoun, M & Lei, H 2016, Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis. in T Bräunl, B McCane, M Rivera & X Yu (eds), Pacific-Rim Symposium on Image and Video Technology. Lecture Notes in Computer Science , vol. 9431, Springer, USA, pp. 751-761, 7th Pacific-Rim Symposium on Image and Video Technology, Auckland, New Zealand, 23/11/15. https://doi.org/10.1007/978-3-319-29451-3_59

    Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis. / Wang, Xupeng; Sohel, Ferdous; Bennamoun, Mohammed; Lei, Hang.

    Pacific-Rim Symposium on Image and Video Technology. ed. / Thomas Bräunl; Brendan McCane; Mariano Rivera; Xinguo Yu. USA : Springer, 2016. p. 751-761 (Lecture Notes in Computer Science ; Vol. 9431).

    Research output: Chapter in Book/Conference paperConference paper

    TY - GEN

    T1 - Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis

    AU - Wang, Xupeng

    AU - Sohel, Ferdous

    AU - Bennamoun, Mohammed

    AU - Lei, Hang

    PY - 2016

    Y1 - 2016

    N2 - This paper presents an efficient feature point descriptor for non-rigid shape analysis. The descriptor is developed based on the properties of the heat diffusion process on a shape. We use, for the first time, the Heat Kernel Signature of a particular time scale to define the scalar field on a manifold. Then, motivated by the successful use of a local reference frame for rigid shape analysis, we construct a repetitive local polar coordinate system, which is invariant under isometric deformations. Finally, a binary descriptor is derived by comparing the intensities of the neighboring points for each feature point. We show that the descriptor is highly discriminative and can be computed simply using ‘intensity comparisons’ on a shape. Furthermore, its similarity can be evaluated using the Hamming distance, which is very efficient to compute compared with the commonly used L2L2 norm. Our experiments demonstrate a superior performance compared to existing techniques on the standard benchmark TOSCA.

    AB - This paper presents an efficient feature point descriptor for non-rigid shape analysis. The descriptor is developed based on the properties of the heat diffusion process on a shape. We use, for the first time, the Heat Kernel Signature of a particular time scale to define the scalar field on a manifold. Then, motivated by the successful use of a local reference frame for rigid shape analysis, we construct a repetitive local polar coordinate system, which is invariant under isometric deformations. Finally, a binary descriptor is derived by comparing the intensities of the neighboring points for each feature point. We show that the descriptor is highly discriminative and can be computed simply using ‘intensity comparisons’ on a shape. Furthermore, its similarity can be evaluated using the Hamming distance, which is very efficient to compute compared with the commonly used L2L2 norm. Our experiments demonstrate a superior performance compared to existing techniques on the standard benchmark TOSCA.

    U2 - 10.1007/978-3-319-29451-3_59

    DO - 10.1007/978-3-319-29451-3_59

    M3 - Conference paper

    SN - 9783319294513

    T3 - Lecture Notes in Computer Science

    SP - 751

    EP - 761

    BT - Pacific-Rim Symposium on Image and Video Technology

    A2 - Bräunl, Thomas

    A2 - McCane, Brendan

    A2 - Rivera, Mariano

    A2 - Yu, Xinguo

    PB - Springer

    CY - USA

    ER -

    Wang X, Sohel F, Bennamoun M, Lei H. Binary Descriptor Based on Heat Diffusion for Non-rigid Shape Analysis. In Bräunl T, McCane B, Rivera M, Yu X, editors, Pacific-Rim Symposium on Image and Video Technology. USA: Springer. 2016. p. 751-761. (Lecture Notes in Computer Science ). https://doi.org/10.1007/978-3-319-29451-3_59