A Skeleton-based Approach for Detection of Perceptually Salient Features on Polygonal Surfaces

The paper presents a skeleton‐based approach for robust detection of perceptually salient shape features. Given ashape approximated by a polygonal surface, its skeleton is extracted using a three‐dimensional Voronoi diagramtechnique proposed recently by Amenta et al. [ 3 ]. Shape creases, ridges and ravines, are detected as curvescorresponding to skeletal edges. Salient shape regions are extracted via skeleton decomposition into patches.The approach explores the singularity theory for ridge and ravine detection, combines several filtering methodsfor skeleton denoising and for selecting perceptually important ridges and ravines, and uses a topological analysisof the skeleton for detection of salient shape regions. ACM CSS: I.3.5 Computational Geometry and Object Modeling     

Separatrix Persistence: Extraction of Salient Edges on Surfaces Using Topological Methods

Salient edges are perceptually prominent features of a surface. Most previous extraction schemes utilize the notion of ridges and valleys for their detection, thereby requiring curvature derivatives which are rather sensitive to noise. We introduce a novel method for salient edge extraction which does not depend on curvature derivatives. It is based on a topological analysis of the principal curvatures and salient edges of the surface are identified as parts of separatrices of the topological skeleton. Previous topological approaches obtain results including non-salient edges due to inherent properties of the underlying algorithms. We extend the profound theory by introducing the novel concept of separatrix persistence , which is a smooth measure along a separatrix and allows to keep its most salient parts only. We compare our results with other methods for salient edge extraction.     

Visual salience guided feature-aware shape simplification

In the area of 3D digital engineering and 3D digital geometry processing, shape simplification is an important task to reduce their requirement of large memory and high time complexity. By incorporating the content-aware visual salience measure of a polygonal mesh into simplification operation, a novel feature-aware shape simplification approach is presented in this paper. Owing to the robust extraction of relief heights on 3D highly detailed meshes, our visual salience measure is defined by a center-surround operator on Gaussian-weighted relief heights in a scale-dependent manner. Guided by our visual salience map, the feature-aware shape simplification algorithm can be performed by weighting the high-dimensional feature space quadric error metric of vertex pair contractions with the weight map derived from our visual salience map. The weighted quadric error metric is calculated in a six-dimensional feature space by combining the position and normal information of mesh vertices. Experimental results demonstrate that our visual salience guided shape simplification scheme can adaptively and effectively re-sample the underlying models in a feature-aware manner, which can account for the visually salient features of the complex shapes and thus yield better visual fidelity.     

Visual saliency guided normal enhancement technique for 3D shape depiction

Visual saliency can effectively guide the viewer's visual attention to salient regions of a 3D shape. Incorporating the visual saliency measure of a polygonal mesh into the normal enhancement operation, a novel saliency guided shading scheme for shape depiction is developed in this paper. Due to the visual saliency measure of the 3D shape, our approach will adjust the illumination and shading to enhance the geometric salient features of the underlying model by dynamically perturbing the surface normals. The experimental results demonstrate that our non-photorealistic shading scheme can enhance the depiction of the underlying shape and the visual perception of its salient features for expressive rendering. Compared with previous normal enhancement techniques, our approach can effectively convey surface details to improve shape depiction without impairing the desired appearance.     

基于多模态多级特征聚合网络的光场显著性目标检测

现有基于深度学习的显著性检测算法主要针对二维RGB图像设计,未能利用场景图像的三维视觉信息,而当前光场显著性检测方法则多数基于手工设计,特征表示能力不足,导致上述方法在各种挑战性自然场景图像上的检测效果不理想。提出一种基于卷积神经网络的多模态多级特征精炼与融合网络算法,利用光场图像丰富的视觉信息,实现面向四维光场图像的精准显著性检测。为充分挖掘三维视觉信息,设计2个并行的子网络分别处理全聚焦图像和深度图像。在此基础上,构建跨模态特征聚合模块实现对全聚焦图像、焦堆栈序列和深度图3个模态的跨模态多级视觉特征聚合,以更有效地突出场景中的显著性目标对象。在DUTLF-FS和HFUT-Lytro光场基准数据集上进行实验对比,结果表明,该算法在5个权威评估度量指标上均优于MOLF、AFNet、DMRA等主流显著性目标检测算法。     

融合双目多维感知特征的立体视频显著性检测

目的 立体视频能提供身临其境的逼真感而越来越受到人们的喜爱,而视觉显著性检测可以自动预测、定位和挖掘重要视觉信息,可以帮助机器对海量多媒体信息进行有效筛选。为了提高立体视频中的显著区域检测性能,提出了一种融合双目多维感知特性的立体视频显著性检测模型。方法 从立体视频的空域、深度以及时域3个不同维度出发进行显著性计算。首先,基于图像的空间特征利用贝叶斯模型计算2D图像显著图;接着,根据双目感知特征获取立体视频图像的深度显著图;然后,利用Lucas-Kanade光流法计算帧间局部区域的运动特征,获取时域显著图;最后,将3种不同维度的显著图采用一种基于全局-区域差异度大小的融合方法进行相互融合,获得最终的立体视频显著区域分布模型。结果 在不同类型的立体视频序列中的实验结果表明,本文模型获得了80%的准确率和72%的召回率,且保持了相对较低的计算复杂度,优于现有的显著性检测模型。结论 本文的显著性检测模型能有效地获取立体视频中的显著区域,可应用于立体视频/图像编码、立体视频/图像质量评价等领域。     

基于卷积神经网络与视觉显著性的图像融合

传统的红外与可见光图像融合方法,多数需要手动提取特征且特征提取单一。而深度学习可以自动选择图像特征,改善特征提取的单一性,因此提出一种基于卷积神经网络与视觉显著性的红外和可见光图像融合方法。利用卷积神经网络获得红外目标与背景的二分类图;利用条件随机场对分类图进行精分割得到显著性目标提取图;采用非下采样轮廓波变换并结合目标提取图,得到融合图像。实验结果表明,该方法在主观视觉和客观评价方面均优于传统非智能方法,并且5个客观评价指标(边缘信息保留量,结构相似度,互信息,信息熵和标准差)均有显著提高。     

Perceptual-saliency extremum lines for 3D shape illustration

Owing to their efficiency for conveying perceptual information of the underlying shape and their pleasing perceiving in visual aesthetics experience, line drawings are now becoming a widely used technique for illustrating 3D shapes. Using a center-surrounding bilateral filter operator on Gaussian-weighted average of local projection height between mesh vertices and their neighbors, a new perceptual-saliency measure which can depict surface salient features, is proposed in this paper. Due to the definition of perceptual-saliency measure, our perceptual-saliency extremum lines can be considered as the ridge-valley lines of perceptual-saliency measure along the principal curvature directions on triangular meshes. The experimental results demonstrate that these extremum lines effectively capture and depict 3D shape information visually, especially for archaeological artifacts.     

An effective illustrative visualization framework based on photic extremum lines (PELs)

Conveying shape using feature lines is an important visualization tool in visual computing. The existing feature lines (e.g., ridges, valleys, silhouettes, suggestive contours, etc.) are solely determined by local geometry properties (e.g., normals and curvatures) as well as the view position. This paper is strongly inspired by the observation in human vision and perception that a sudden change in the luminance plays a critical role to faithfully represent and recover the 3D information. In particular, we adopt the edge detection techniques in image processing for 3D shape visualization and present Photic Extremum Lines (PELs) which emphasize significant variations of illumination over 3D surfaces. Comparing with the existing feature lines, PELs are more flexible and offer users more freedom to achieve desirable visualization effects. In addition, the user can easily control the shape visualization by changing the light position, the number of light sources, and choosing various light models. We compare PELs with the existing approaches and demonstrate that PEL is a flexible and effective tool to illustrate 3D surface and volume for visual computing.     

High Fidelity Digital Terrain Modelling for the Reconstruction of the Antarctic Sea Floor

Shape-based modelling is a general approach to surface representation, which has a great importance in the specific context of the Antarctic sea floor reconstruction, where measurements can involve critical operations. Here, a method is proposed where shape-based surface reconstruction is achieved performing a geometric reasoning on the raw data to delineate a shape structure on which the final surface model can be built. Data of the Antarctic sea floor are collected by surveys carried out along parallel courses during which the depth of the sea is measured at almost regular intervals. The seabed is then represented by a set of profiles, corresponding to almost vertical cross sections. The surface reconstruction is performed in three steps. First, a shape-based simplification is carried out on the profiles, using a combination of the wavelet theory and the classical Douglas and Peucker algorithm. The second step consists of finding similarities in the morphology of adjacent profiles, which may suggest the presence of surface features, such as ridges and ravines. Finally, the deduced surface features are used to build a kind of skeleton on which the most appropriate triangulation can be constructed.     

注意力引导网络的显著性目标检测

目的 全卷积模型的显著性目标检测大多通过不同层次特征的聚合实现检测,如何更好地提取和聚合特征是一个研究难点。常用的多层次特征融合策略有加法和级联法,但是这些方法忽略了不同卷积层的感受野大小以及产生的特征图对最后显著图的贡献差异等问题。为此,本文结合通道注意力机制和空间注意力机制有选择地逐步聚合深层和浅层的特征信息,更好地处理不同层次特征的传递和聚合,提出了新的显著性检测模型AGNet(attention-guided network),综合利用几种注意力机制对不同特征信息加权解决上述问题。方法 该网络主要由特征提取模块(feature extraction module, FEM)、通道—空间注意力融合模块(channel-spatial attention aggregation module, C-SAAM)和注意力残差细化模块(attention residual refinement module,ARRM)组成,并且通过最小化像素位置感知(pixel position aware, PPA)损失训练网络。其中,C-SAAM旨在有选择地聚合浅层的边缘信息以及深层抽象的语义特征,利用通道注意力和空间注意力避免融合冗余的背景信息对显著性映射造成影响;ARRM进一步细化融合后的输出,并增强下一个阶段的输入。结果 在5个公开数据集上的实验表明,AGNet在多个评价指标上达到最优性能。尤其在DUT-OMRON(Dalian University of Technology-OMRON)数据集上,F-measure指标相比于排名第2的显著性检测模型提高了1.9%,MAE(mean absolute error)指标降低了1.9%。同时,网络具有不错的速度表现,达到实时效果。结论 本文提出的显著性检测模型能够准确地分割出显著目标区域,并提供清晰的局部细节。     

基于显著特征谱嵌入的三维模型相似性分析*

面向快速、高效的三维模型检索技术的迫切需求,提出基于显著特征谱嵌入的三维模型相似性分析方法.首先通过局部曲率及凸凹性检测,有效提取模型的显著特征点,构建模型的显著特征描述算子.然后基于拉普拉斯映射及谱分析原理进一步提取模型的内蕴形状特征.最后,结合薄板样条函数实现模型间的配准与相似性分析.通过实验验证文中方法不仅有效提高模型匹配的效率,而且能有效识别同一类模型的结构特征,同时对于残缺模型间的匹配具有较强的鲁棒性.     

Curve skeleton extraction by coupled graph contraction and surface clustering

In this paper, we present a practical algorithm to extract a curve skeleton of a 3D shape. The core of our algorithm comprises coupled processes of graph contraction and surface clustering. Given a 3D shape represented by a triangular mesh, we first construct an initial skeleton graph by directly copying the connectivity and geometry information from the input mesh. Graph contraction and surface clustering are then performed iteratively. The former merges certain graph nodes based on computation of an approximate centroidal Voronoi diagram, seeded by subsampling the graph nodes from the previous iteration. Meanwhile, a coupled surface clustering process serves to regularize the graph contraction. Constraints are used to ensure that extremities of the graph are not shortened undesirably, to ensure that skeleton has the correct topological structure, and that surface clustering leads to an approximately-centered skeleton of the input shape. These properties lead to a stable and reliable skeleton graph construction algorithm.Experiments demonstrate that our skeleton extraction algorithm satisfies various desirable criteria. Firstly, it produces a skeleton homotopic with the input (the genus of both shapes agree) which is both robust (results are stable with respect to noise and remeshing of the input shape) and reliable (every boundary point is visible from at least one curve-skeleton location). It can also handle point cloud data if we first build an initial skeleton graph based on k-nearest neighbors. In addition, a secondary output of our algorithm is a skeleton-to-surface mapping, which can e.g. be used directly for skinning animation.Highlights(1) An algorithm for curve skeleton extraction from 3D shapes based on coupled graph contraction and surface clustering. (2) The algorithm meets various desirable criteria and can be extended to work for incomplete point clouds.     

3D Skeletons: A State‐of‐the‐Art Report

Given a shape, a skeleton is a thin centered structure which jointly describes the topology and the geometry of the shape. Skeletons provide an alternative to classical boundary or volumetric representations, which is especially effective for applications where one needs to reason about, and manipulate, the structure of a shape. These skeleton properties make them powerful tools for many types of shape analysis and processing tasks. For a given shape, several skeleton types can be defined, each having its own properties, advantages, and drawbacks. Similarly, a large number of methods exist to compute a given skeleton type, each having its own requirements, advantages, and limitations. While using skeletons for two‐dimensional (2D) shapes is a relatively well covered area, developments in the skeletonization of three‐dimensional (3D) shapes make these tasks challenging for both researchers and practitioners. This survey presents an overview of 3D shape skeletonization. We start by presenting the definition and properties of various types of 3D skeletons. We propose a taxonomy of 3D skeletons which allows us to further analyze and compare them with respect to their properties. We next overview methods and techniques used to compute all described 3D skeleton types, and discuss their assumptions, advantages, and limitations. Finally, we describe several applications of 3D skeletons, which illustrate their added value for different shape analysis and processing tasks.     

Domain connected graph: the skeleton of a closed 3D shape for animation

In previous research, three main approaches have been employed to solve the skeleton extraction problem: medial axis transform (MAT), generalized potential field and decomposition-based methods. These three approaches have been formulated using three different concepts, namely surface variation, inside energy distribution, and the connectivity of parts. By combining the above mentioned concepts, this paper creates a concise structure to represent the control skeleton of an arbitrary object. First, an algorithm is proposed to detect the end, connection and joint points of an arbitrary 3D object. These three points comprise the skeleton, and are the most important to consider when describing it. In order to maintain the stability of the point extraction algorithm, a prong-feature detection technique and a level iso-surfaces function-based on the repulsive force field was employed. A neighborhood relationship inherited from the surface able to describe the connection relationship of these positions was then defined. Based on this relationship, the skeleton was finally constructed and named domain connected graph (DCG). The DCG not only preserves the topology information of a 3D object, but is also less sensitive than MAT to the perturbation of shapes. Moreover, from the results of complicated 3D models, consisting of thousands of polygons, it is evident that the DCG conforms to human perception.     

基于深度学习框架的多模态动作识别

提出一种基于深度神经网络的多模态动作识别方法,根据不同模态信息的特性分别采用不同的深度神经网络,适应不同模态的视频信息,并将多种深度网络相结合,挖掘行为识别的多模态特征。主要考虑人体行为静态和动态2种模态信息,结合微软Kinect的多传感器摄像机获得传统视频信息的同时也能获取对应的深度骨骼点信息。对于静态信息采用卷积神经网络模型,对于动态信息采用递归循环神经网络模型。最后将2种模型提取的特征相融合进行动作识别和分类。在MSR 3D的行为数据库上实验结果表明,本文的方法对动作识别具有良好的分类效果。