Guo, Xiaohu

Permanent URI for this collection

Dr. Guo works in the areas of computer graphics, computer animation and simulation, geometric and physically-based modeling, computer animation, virtual reality and medical imaging. More information about Xiaohu Guo is available on his home and Research Explorer pages.


Recent Submissions

Now showing 1 - 5 of 5
  • Item
    Q-Mat+: An Error-Controllable and Feature-Sensitive Simplification Algorithm for Medial Axis Transform
    (Elsevier B.V.) Pan, Y.; Wang, B.; Guo, Xiaohu; Zeng, H.; Ma, Y.; Wang, W.; Guo, Xiaohu
    The medial axis transform (MAT), as an intrinsic shape representation, plays an important role in shape approximation, recognition and retrieval. Q-MAT is a state-of-the-art algorithm driven by quadratic error minimization to compute a geometrically precise, structurally concise, and compact representation of the MAT for 3D shapes. In this work we extend the technique to make it more robust, controllable, and name it Q-MAT+. Combining shape diameter function (SDF) and other mesh information, Q-MAT+ gets a more complete and accurate initial MAT than Q-MAT, even for extreme thin features, such as wires and sheets. Q-MAT+ could quickly remove insignificant branches while preserving significant ones to get a simple and faithful piecewise linear approximation of the MAT. Moreover, it performs the medial axis simplification with explicit maintenance and the control of Hausdorff error, which is not originally supported in Q-MAT. We further demonstrate the outstanding efficiency and accuracy of our method compared with other existing approaches for MAT generation and simplification. ©2019 Elsevier B.V.
  • Item
    Superpixel Generation by Agglomerative Clustering with Quadratic Error Minimization
    (Blackwell Publishing Ltd) Dong, Xiao; Chen, Z.; Yao, J.; Guo, Xiaohu; Dong, Xiao; Guo, Xiaohu
    Superpixel segmentation is a popular image pre-processing technique in many computer vision applications. In this paper, we present a novel superpixel generation algorithm by agglomerative clustering with quadratic error minimization. We use a quadratic error metric (QEM) to measure the difference of spatial compactness and colour homogeneity between superpixels. Based on the quadratic function, we propose a bottom-up greedy clustering algorithm to obtain higher quality superpixel segmentation. There are two steps in our algorithm: merging and swapping. First, we calculate the merging cost of two superpixels and iteratively merge the pair with the minimum cost until the termination condition is satisfied. Then, we optimize the boundary of superpixels by swapping pixels according to their swapping cost to improve the compactness. Due to the quadratic nature of the energy function, each of these atomic operations has only O(1) time complexity. We compare the new method with other state-of-the-art superpixel generation algorithms on two datasets, and our algorithm demonstrates superior performance. ©2018 The Authors Computer Graphics Forum ©2018 The Eurographics Association and John Wiley & Sons Ltd.
  • Item
    Field-Aligned and Lattice-Guided Tetrahedral Meshing
    (John Wiley & Sons Ltd.) Ni, Saifeng; Zhong, Z.; Huang, J.; Wang, W.; Guo, Xiaohu; Ni, Saifeng; Guo, Xiaohu
    We present a particle-based approach to generate field-aligned tetrahedral meshes, guided by cubic lattices, including BCC and FCC lattices. Given a volumetric domain with an input frame field and a user-specified edge length for the cubic lattice, we optimize a set of particles to form the desired lattice pattern. A Gaussian Hole Kernel associated with each particle is constructed. Minimizing the sum of kernels of all particles encourages the particles to form a desired layout, e.g., field-aligned BCC and FCC. The resulting set of particles can be connected to yield a high quality field-aligned tetrahedral mesh. As demonstrated by experiments and comparisons, the field-aligned and lattice-guided approach can produce higher quality isotropic and anisotropic tetrahedral meshes than state-of-the-art meshing methods.
  • Item
    Three‐Dimensional Lesion Phenotyping and Physiologic Characterization Inform Remyelination Ability in Multiple Sclerosis
    (American Society of Neuroimaging) Sivakolundu, Dinesh K.; Hansen, Madison R.; West, Kathryn L.; Wang, Yeqi; Stanley, Thomas; Wilson, Andrew; McCreary, Morgan; Turner, Monroe P.; Pinho, Marco C.; Newton, Braeden D.; Guo, Xiaohu; Rypma, Bart; Okuda, Darin T.; Sivakolundu, Dinesh K.; West, Kathryn L.; Wang, Yeqi; Stanley, Thomas; Wilson, Andrew; Turner, Monroe P.; Guo, Xiaohu; Rypma, Bart
    BACKGROUND AND PURPOSE Multiple sclerosis (MS) clinical management is based upon lesion characterization from 2‐dimensional (2D) magnetic resonance imaging (MRI) views. Such views fail to convey the lesion‐phenotype (ie, shape and surface texture) complexity, underlying metabolic alterations, and remyelination potential. We utilized a 3‐dimensional (3D) lesion phenotyping approach coupled with imaging to study physiologic profiles within and around MS lesions and their impacts on lesion phenotypes. METHODS Lesions were identified in 3T T₂‐FLAIR images and segmented using geodesic active contouring. A calibrated fMRI sequence permitted measurement of cerebral blood flow (CBF), blood‐oxygen‐level‐dependent signal (BOLD), and cerebral metabolic rate of oxygen (CMRO₂). These metrics were measured within lesions and surrounding tissue in concentric layers exact to the 3D‐lesion shape. BOLD slope was calculated as BOLD changes from a lesion to its surrounding perimeters. White matter integrity was measured using diffusion kurtosis imaging. Associations between these metrics and 3D‐lesion phenotypes were studied. RESULTS One hundred nine lesions from 23 MS patients were analyzed. We identified a noninvasive biomarker, BOLD slope, to metabolically characterize lesions. Positive BOLD slope lesions were metabolically active with higher CMRO₂ and CBF compared to negative BOLD slope or inactive lesions. Metabolically active lesions with more intact white matter integrity had more symmetrical shapes and more complex surface textures compared to inactive lesions with less intact white matter integrity. CONCLUSION The association of lesion phenotypes with their metabolic signatures suggests the prospect for translation of such data to clinical management by providing information related to metabolic activity, lesion age, and risk for disease reactivation and self‐repair. Our findings also provide a platform for disease surveillance and outcome quantification involving myelin repair therapeutics.
  • Item
    GPU-Based Computation of Discrete Periodic Centroidal Voronoi Tessellation in Hyperbolic Space
    (2012-02) Shuai, Liang; Guo, Xiaohu
    Periodic centroidal Voronoi tessellation (CVT) in hyperbolic space provides a nice theoretical framework for computing the constrained CVT on high-genus (genus > 1) surfaces. This paper addresses two computational issues related to such hyperbolic CVT framework: (1) efficient reduction of unnecessary site copies in neighbor domains on the universal covering space, based on two special rules; (2) GPU-based parallel algorithms to compute a discrete version of the hyperbolic CVT. Our experiments show that with the dramatically reduced number of unnecessary site copies in neighbor domains and the GPU-based parallel algorithms, we significantly speed up the computation of CVT for high-genus surfaces. The proposed discrete hyperbolic CVT guarantees to converge and produces high-quality results.