检索结果-南通市图书馆

Surveys on line drawing simplification for Chinese cultural computing

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

The Journal of China Universities of Posts and Telecommunications 2022年第2期29卷 33-42页

作者： Chen Jiazhou Amal Ahmed Hasan Mohammed Huang Keyu miao yongwei School of Computer Science and Technology Zhejiang University of TechnologyHangzhou 310023China School of Information Science and Technology Hangzhou Normal UniversityHangzhou 311121China

In the recent decade,many approaches of rough line drawing simplification were proposed,but they are not well summarized yet,especially from the perspective of Chinese cultural *** this paper,a comprehensive review of existing line drawing simplification methods was presented,including their algorithms,advantages/disadvantages,inputs/outputs,datasets and source codes,*** raster line drawings,related simplification work was discussed according to four main categories:fitting-based methods,tracing-based methods,field-based methods,and learning-based *** vector line drawings,a deep investigation was introduced for two major steps of simplification:stroke grouping and stroke ***,conclusions were given,key challenges and future directions of line drawing simplification for Chinese traditional art were thoroughly discussed.

关键词： line drawing sketch simplification cultural computing

Relief generation from 3D scenes guided by geometric texture richness

在线全文

学校读者我要写书评

暂无评论

Computational Visual Media 2018年第3期4卷 209-221页

作者： yongwei miao Yuliang Sun Xudong Fang Jiazhou Chen Xudong Zhang Renato Pajarola College of Information Science and Technology Zhejiang Sci-Tech University College of Computer Science and Technology Zhejiang University of Technology Department of Informatics University of Zrich

Typically, relief generation from an input 3D scene is limited to either bas-relief or high-relief modeling. This paper presents a novel unified scheme for synthesizing reliefs guided by the geometric texture richness of 3D scenes; it can generate both bas- and high-reliefs. The type of relief and compression coefficient can be specified according to the user＇s artistic needs. We use an energy minimization function to obtain the surface reliefs, which contains a geometry preservation term and an edge constraint term. An edge relief measure determined by geometric texture richness and edge z-depth is utilized to achieve a balance between these two terms. During relief generation, the geometry preserwtion term keeps local surface detail in the original scenes, while the edge constraint term maintains regions of the original models with rich geometric texture. Elsewhere, in high- reliefs, the edge constraint term also preserves depth discontinuities in the higher parts of the original scenes. The energy function can be discretized to obtain a sparse linear system. The reliefs are obtained by solving it by an iterative process. Finally, we apply non-linear compression to the relief to meet the user＇s artistic needs. Experimental results show the method＇s effectiveness for generating both bas- and high-reliefs for complex 3D scenes in a unified manner.

关键词： geometric texture richness reliefgeneration energy minimization bas-relief high-relief

Local imperceptible adversarial attacks against human pose estimation networks

在线全文

学校读者我要写书评

暂无评论

Visual Computing for Industry,Biomedicine,and Art 2023年第1期6卷 318-328页

作者： Fuchang Liu Shen Zhang Hao Wang Caiping Yan yongwei miao School of Information Science and Technology Hangzhou Normal UniversityHangzhou 311121ZhejiangChina

Deep neural networks are vulnerable to attacks from adversarial *** attack research on human pose estimation(HPE),particularly for body joint detection,has been largely *** classification-based attack methods to body joint regression tasks is not *** issue is that the attack effectiveness and imperceptibility contradict each *** solve these issues,we propose local imperceptible attacks on HPE *** particular,we reformulate imperceptible attacks on body joint regression into a constrained maximum allowable ***,we approximate the solution using iterative gradient-based strength refinement and greedy-based pixel *** method crafts effective perceptual adversarial attacks that consider both human perception and attack *** conducted a series of imperceptible attacks against state-of-the-art HPE methods,including HigherHRNet,DEKR,and *** experimental results demonstrate that the proposed method achieves excellent imperceptibility while maintaining attack effectiveness by significantly reducing the number of perturbed *** 4%of the pixels can achieve sufficient attacks on HPE.

关键词： Adversarial attack Human pose estimation White-box attack Imperceptibility Local perturbation

Symm Sketch: Creating symmetric 3D free-form shapes from 2D sketches

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2015年第1期1卷 3-16页

作者： yongwei miao Feixia Hu Xudong Zhang Jiazhou Chen Renato Pajarola College of Computer Science and Technology Zhejiang University of Technology Department of Informatics University of Zürich

This paper presents Symm Sketch—a system for creating symmetric 3D free-form shapes from 2D sketches. The reconstruction task usually separates a 3D symmetric shape into two types of shape components, that is, the self-symmetric shape component and the mutual-symmetric shape components. Each type can be created in an intuitive manner. Using a uniform symmetry plane, the user first draws 2D sketch lines for each shape component on a sketching plane. The z-depth information of the hand-drawn input sketches can be calculated using their property of mirror symmetry to generate 3D construction curves. In order to provide more freedom for controlling the local geometric features of the reconstructed free-form shapes(e.g., non-circular crosssections), our modeling system creates each shape component from four construction curves. Using one pair of symmetric curves and one pair of general curves, an improved cross-sectional surface blending scheme is applied to generate a parametric surface for each component. The final symmetric free-form shape is progressively created, and is represented by 3D triangular mesh. Experimental results illustrate that our system can generate complex symmetric free-form shapes effectively and conveniently.

关键词： 3D reconstruction construction curves free-form shapes mirror symmetry sketch-based modeling

A Fast Algorithm for Content-aware Saliency Detection and Stylized Rendering

在线全文

学校读者我要写书评

暂无评论

A Fast Algorithm for Content-aware Saliency Detection and St...

基于深度信息局部二值模式特征的室内场景边缘检测

2012年计算机应用与系统建模国际会议

作者： yongwei miao Ruifeng Han Huahao Shou College of Computer Science and Technology Zhejiang University of Technology

Image saliency plays an important role in image processing and recognizing tasks. Owing to color difference computation between the patch centered at the current pixel and the patches centered at its neighboring pixels, a fast algorithm for multi-scale content-aware saliency detection is presented in this paper. Furthermore, the histogram information of the image is also incorporated into our saliency detection to eliminate the loss of accuracy introduced in our local dissimilarity comparison. Due to our content-aware saliency estimation, a saliency guided stylized rendering scheme is given by combing the saliency guided bilateral filtering and saliency guided contour detection techniques.

关键词： Content-aware Image saliency Bilateral filtering Edge detection Stylized rendering

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

计算机辅助设计与图形学学报 2017年第12期29卷 2162-2170页

作者：于莉洁孙瑜亮缪永伟浙江工业大学计算机科学与技术学院杭州310023 浙江农林大学信息与教育技术中心杭州311300 浙江理工大学信息学院杭州310018

针对室内场景深度图像,检测场景中的面片边缘信息并确定场景中的完整面信息是实现室内场景分析和理解的基础.基于深度信息局部二值模式特征,提出一种室内场景深度图像边缘检测的方法.首先对场景深度信息图分别求X方向和Y方向的梯度信息... 详细信息

针对室内场景深度图像,检测场景中的面片边缘信息并确定场景中的完整面信息是实现室内场景分析和理解的基础.基于深度信息局部二值模式特征,提出一种室内场景深度图像边缘检测的方法.首先对场景深度信息图分别求X方向和Y方向的梯度信息,结合2个梯度图找到深度信息图的基本边缘信息;然后计算基本边缘附近的局部二值模式特征信息,并计算深度图每个点的法线信息;最后利用深度信息局部二值模式特征和法线信息对边缘信息进行判断和矫正,以提取深度图像的面片边缘信息.实验结果表明,该方法能够高效、准确地检测室内场景深度图像的边缘信息,避免边缘信息的过检测和欠检测.

关键词：室内场景深度图像边缘检测局部二值模式特征

浮雕显著性信息熵驱动的三维模型最优视点选择

在线全文

学校读者我要写书评

暂无评论

计算机辅助设计与图形学学报 2011年第12期23卷 2033-2039页

作者：缪永伟王洪军寿华好浙江工业大学计算机科学与技术学院杭州310023 浙江工业大学理学院杭州310023

在三维模型最优视点选择中,将模型的浮雕显著性信息度量引入模型视点互信息的信息熵分析,提出一种结合视觉显著性的模型视见描述子刻画方法.针对模型视点球的各个视点计算模型可见面片浮雕显著性信息熵驱动的视见描述子,并从中选择视见... 详细信息

在三维模型最优视点选择中,将模型的浮雕显著性信息度量引入模型视点互信息的信息熵分析,提出一种结合视觉显著性的模型视见描述子刻画方法.针对模型视点球的各个视点计算模型可见面片浮雕显著性信息熵驱动的视见描述子,并从中选择视见描述子值最大的视点作为模型的最优观察视点.实验结果表明,由该方法获取的最优视点得到的视图能够反映和展示三维模型的视觉显著特征.

关键词：最优视点浮雕显著性视点互信息信息熵视见描述子

在线全文

学校读者我要写书评

暂无评论

基于显著性图的点云替换对抗攻击

中国图象图形学报 2022年第2期27卷 500-510页

作者：刘复昌南博缪永伟杭州师范大学信息科学与技术学院杭州311121 浙江理工大学信息学院杭州310018

目的传统针对对抗攻击的研究通常集中于2维图像领域,而对3维物体进行修改会直接影响该物体的3维特性,生成令人无法察觉的扰动是十分困难的,因此针对3维点云数据的对抗攻击研究并不多。点云对抗样本,如点云物体分类、点云物体分割等的深... 详细信息

目的传统针对对抗攻击的研究通常集中于2维图像领域,而对3维物体进行修改会直接影响该物体的3维特性,生成令人无法察觉的扰动是十分困难的,因此针对3维点云数据的对抗攻击研究并不多。点云对抗样本,如点云物体分类、点云物体分割等的深度神经网络通常容易受到攻击,致使网络做出错误判断。因此,提出一种基于显著性图的点云替换对抗攻击方法。方法由于现有点云分类网络通常需要获取点云模型中的关键点,该方法通过将点移动到点云中心计算点的显著性值,从而构建点云显著性图,选择具有最高显著性值的采样点集作为关键点集,以确保对网络分类结果造成更大的影响;利用Chamfer距离衡量点云模型之间的差异性,并选择与点云模型库中具有最近Chamfer距离的模型关键点集进行替换,从而实现最小化点云扰动并使得人眼难以察觉。结果使用ModelNet40数据集,分别在点云分类网络Point Net和Point Net++上进行对比实验。在Point Net网络上,对比FGSM(fast gradient sign method)、I-FGSM(iterative fast gradient sign method)和JSMA(Jacobian-based saliency map attack)方法,本文方法攻击成功率分别提高38.6%、7.3%和41%;若扰动100个采样点,本文方法将使网络准确率下降到6.2%。在Point Net++网络上,对比FGSM和JSMA,本文方法的攻击成功率分别提高58.6%和85.3%;若扰动100个采样点,本文方法将使网络准确率下降到12.8%。结论本文提出的点云对抗攻击方法,不仅考虑到对抗攻击的效率,而且考虑了对抗样本的不可察觉性,能够高效攻击主流的点云深度神经网络。

关键词：点云对抗攻击显著性图 Chamfer距离 PointNet PointNet++

在线全文

学校读者我要写书评

暂无评论

基于几何基元定标的单幅图像组合式建筑建模

浙江工业大学学报 2019年第6期47卷 611-617,647页

作者：缪永伟汪逊浙江工业大学计算机科学与技术学院

针对单幅图像几何结构复杂的组合式建筑建模,为克服单幅图像提供的信息量少、建筑物部件之间相互遮挡关系等局限性,提出了一种基于长方体和圆柱体几何基元定标的组合式建筑交互建模方法。首先对建筑物每个几何部件单独进行场景定标与建... 详细信息

针对单幅图像几何结构复杂的组合式建筑建模,为克服单幅图像提供的信息量少、建筑物部件之间相互遮挡关系等局限性,提出了一种基于长方体和圆柱体几何基元定标的组合式建筑交互建模方法。首先对建筑物每个几何部件单独进行场景定标与建模。对长方体部件,用户标注图像投影的6个顶点,系统根据顶点坐标计算相机参数,利用顶点与三维几何体的对应关系构建长方体部件模型;对圆柱体部件,用户标注投影轮廓的4个顶点,计算灭点坐标并利用灭点性质求解相机参数,再构建圆柱体部件模型。然后,利用投影误差和部件几何约束关系引进能量函数对部件进行联合优化生成三维模型,实现组合式建筑的精确重建。最后,利用建筑物纹理的对称信息合成真实的三维组合式建筑模型。实验结果表明:该方法交互简单、建模效率高且扩展性好,可以满足不同复杂结构的组合式建筑物的建模需求。

关键词：组合式建筑部件建模联合优化场景估计