曲面体线图的标记与匹配

三维曲面体的标记与匹配,是计算机视觉研究领域中一项相当有用但又是非常困难的技术。本文讨论了一种曲面体标记的新方法。通过引入线段的一些新的标记类型,圆满地解决了利用 Huffman 标记技术描述曲面体时,可能出现的不一致性。根据以上的标记结果,本文还给出了一匹配曲面体线图的新方法。在匹配过程中,我们注重的不是图象的二维特征,而是其对应的物体的三维特征,因而对物体的旋转、遮挡、尺度变化及远视失真等极不敏感。实验结果表明本算法是相当令人满意的。     

基于单幅线条图的三维立体重建方法综述

从单幅线条图重建物体的三维立体结构是机器视觉领域的一个重要问题。它的应用范围包括基于手工绘制的立体设计与创作、将现有的工程线框模型图转换为三维模型、基于自然图像的三维建模、图像理解和检索等。从不同的角度对现有的相关方法进行了分类,并重点从所用算法分类的角度回顾了现有的相关方法。最后总结了基于幅线条图的三维立体重建方法的现状和不足,指出了未来的研究发展方向。     

Constraints Between Distant Lines in the Labelling of Line Drawings of Polyhedral Scenes

The machine interpretation of line drawings has applications both in vision and geometric modelling. This paper extends the classic technique of assigning semantic labels to lines subject to junction constraints, by introducing new constraints (often between distant lines). These include generic constraints between lines lying on a path in the drawing as well as preference constraints between the labellings of pairs of junctions lying on parallel lines. Such constraints are essential to avoid an exponential number of legal labellings of drawings of objects with non-trihedral vertices. The strength of these constraints is demonstrated by their ability to identify the unique correct labelling of many drawings of polyhedral objects with tetrahedral vertices. These new constraints also allowed us to deduce a general polyhedral junction constraint for the case when there is no limit on the number of faces which can meet at a junction.     

A Survey of 3D Solid Reconstruction from 2D Projection Line Drawings

The reconstruction of a 3D object from its 2D projection(s) and its corresponding problem of 3D object recognition are two of the important research areas in the field of computer vision and artificial intelligence. Reconstruction involves determining the geometric and topological relationship of an object's atomic parts whereas recognition involves identifying an object by some form of template matching. Nagendra and Gujar¹ gave a survey of several papers on reconstruction of 3D object from its 2D views. In this paper we present a taxonomy of 3D object reconstruction from 2D projection line drawings. We base the classification on the number of 2D views of the 3D solid object, the degree of user interaction necessary for correct reconstruction, and the internal representation used in the reconstruction process. We discuss the basic issues associated with this problem, review the relevant literature and present topics for future research.     

基于形状演化的线条画风格转换与变形

根据线条画和线条画中笔画的形状特征,运用平面形状演化理论对线条画和笔画的形状进行平滑或增强,并对宽度进行相应的调整,最后得到不同风格的线条画．该风格转换水需要任何样本,而且能够控制转换的程度,提高了线条画风格转换的自由度．     

标记平面立体线图的自适应遗传算法

针对由理查德·迈尔斯提出的标记线图的遗传算法进行改进：采取自适应参数调 整法,同一代中适应度高于平均的个体杂交和变异率动态变化,适应度低于平均的个体杂交和 变异率设为定值;在创建初始种群时加入了约束条件,旨在改善初始种群覆盖空间的不确定性 和个体分布的相对不合理性;修正了遗传算法的适应度函数,使得以个体适应度为指标的选择 算子能正确引导算法搜索解空间。用遗传算法标记 6 幅不同的线图,变量为杂交率、变异率公 式中的参数 a 和 c,分析算法标记成功率曲线的变化趋势,探讨算子参数设置对遗传算法性能 的影响,结果表明 c 属于区间[0,0.05],a 属于区间[0.8,1.0]且为标记线图的遗传算法的最优 参数设置。     

3D Reconstruction from 2D Line Drawings only with Visible Vertices and Edges

The human vision system can reconstruct a 3D object easily from single 2D line drawings even if the hidden lines of the object are invisible. Now, there are many methods have emulated this ability, but when the hidden lines of the object are invisible, these methods cannot reconstruct a complete 3D object. Therefore, we put forward a new algorithm to settle this hard problem. Our approach consists of two steps： （1） infer the invisible vertices and edges to complete the line drawing, （2） propose a vertex-based optimization method to reconstruct a 3D object.     

从单幅正轴测投影线图建立平面立体的模型

提出了一种从单幅二维正轴测投影线图建立平面立体线框模型的新方法。提出一种新的空间直线坐标表示法，根据正轴测投影线图中隐含的直线与平面之间的从属关系建立约束方程。考虑画出隐藏线的正轴测投影线图，根据线画图中隐含的约束条件，建立一个线性系统，通过求解该线性系统，得到平面立体的三维信息。证明解释平面立体正轴测投影线图至少有4个自由度。     

基于规则的建筑结构图钢筋用量自动识别系统

传统的手工建筑工程量统计方法不仅费时而且容易出错,利用计算机自动完成工程量统计工作则可以很好地解决这一问题.介绍了一个基于规则的建筑结构图自动识别系统(automatic interpretation of structure drawings,简称AISD).该系统以矢量化后的电子图档为基础,通过总结建筑工程图结构特征及绘图规则,自动分析图中的各种图形元素、符号以及其关系,理解各种部件信息,并加以综合,以获取正确的建筑工程钢筋用量.通过对工程图的特征进行详细的研究,总结出一套基于规则的适应不同类型工程图     

一种高效的基于约束网络的工程图符号识别方法

将工程图中所有符号的几何特征转换为一个约束网格，用结点记录判断条件，边指示判断次序，实现以层次化约束判断为核心的过滤式识别方法，基于逐步简化的思想，该方法引入多层优先级和边属性的定义，通过优先级调整和边属性查询的配合，优化网络搜索；通过将巳经过约束检查的图元编号记录在结点中而大幅度减少图元遍历和判断。该方法应用在建筑工程图识别中效率较高，对符号种类多、图低内容复杂等情况适应能力较强。     

The Interpretation of Line Drawings with Contrast Failure and Shadows

In line drawings derived from real images, lines may be missing due to contrast failure and objects with curved surfaces may cast shadows from multiple light sources.This paper shows that it is the presence of shadows, rather than contrast failure, that renders the line drawing labelling problem NP-complete. However, shadows are a valuable visual cue, since their presence is formally shown to reduce the average ambiguity of drawings. This is especially true when constraints concerning shadow formation are employed to differentiate shadow and non-shadow lines.The extended junction constraint, concerning straight lines colinear with junctions, compensates the loss of information caused by contrast failure. In fact, we observe the contrast failure paradox: a drawing is sometimes less ambiguous when lines are partly missing due to contrast failure.It is known that the coplanarity of sets of object vertices can be deduced from the presence of straight lines in the drawing. This paper shows that these coplanarity constraints are robust to the presence of contrast failure.     

建筑工程图识别与理解--模型与算法

建筑工程图采用二维示意式表达方式,识别和三维精确还原相对困难、提出了一种基于轴网、结构语义驱动的层次式自生长识别模型(SINEHIR模型),在融合结构语义的基础上从先导建筑构件开始,在削减后的识别空间中结合当前识别层次的特点,分别采用基于特征的建筑符号识别方法、基于符号的轴网识别方法、基于轮廓的结点构件识别方法、基于关联语义的段构件识别方法、基于几何的组合构件识别方法和基于继承的信息传递方法,在模型约束下完成后继识别,以较小的计算量有效排除绘图噪音、提高识别率．     

灰色关联分析在工程图纸除噪预处理中的应用研究

提出了一种基于灰色关联分析的工程图像除噪预处理算法.该算法首先对含噪工程图像的特点进行分析,利用图像中噪声点与边缘点的空间分布连续性不同和图像为近似二值图像的特征,确定了两个非边缘点的参考序列和一组待比较序列;然后通过这两个参考序列与各比较序列之间的灰色关联度将含噪图像的像素重新分类,从而在保留尽可能多边缘的情况下除去噪声.实验表明该除噪方法具有简单易行、效果显著,而且能同时除去多种不同类型的噪声及可动态控制去除噪声与保持边缘的折中度等优点.这对于改善细化、骨架化、识别和矢量化等后继处理结果有着重要意义和实用价值.     

Reconstruction of 3D Object Models from Vectorised Engineering Drawings

In this paper, the problem of 3D-object model reconstruction from engineering drawing projections is analysed, and its main stages are shown. Image vectorisation and entity recognition is mentioned briefly, the main focus being editing or the parameterisation of vectorised drawings and 3D object model reconstruction from vectorised ED projections. Vectorised drawing, as a rule, do not exactly correspond to sizes and other features (touching, parallelity, perpendicularly, symmetry, collinearity, etc.) being available on the initial drawing, and this ED vector model is not suitable for direct use in CAD systems. That is why the parameterisation stage is introduced and considered in detail. An algorithm for 3D-object reconstruction from the vectorised and parameterised drawing is proposed. The algorithm is based on the detection of volumetric solid-state object components (primitives), and performing theoretic-set operations with the components. Practical experience in realising these stages is shown. Received: 23 April 1999, Received in revised form: 29 November 2000, Accepted: 07 December 2000     

基于样本的线条画风格转换与定制方法的研究

提出了一种利用线条画作为样本对任意线条画进行风格转换和定制的方法,该方法不仅可以将样本图风格转移到目标图,还可以方便地控制目标图与样本图之间风格的相似程度,该方法主要采用基于细分笔划的风格相似性比较算法,可在多层次上把目标图转换为样本图风格,不仅如此,还通过定义风格的插值操作实现对风格的定制,与已有方法相比,文中方法只需少量的样本笔划即可进行风格转换,并且通过风格定制,使风格转换具较强的交互可控性,文章最后通过实例证明了方法的有效性。     

一个快速有效的直线反走样算法

直线f(x)=mx b，0≤m≤1；传统的生成算法是取样点(x，f(x))，其中x为整型，并将f(x)约束成整型，其实现过程有著名的DDA、Bresenham及Wu算法，本文针对直线的生成，分析了反走样技术，提出了直线的反走样生成算法，其取样点的光强由与之相邻的四个象素决定，且x和f(x)不必约束为整型，同时推导出点的光强与取样点数量无关而与直线的斜率有着直接的关系．并给出了具体的公式，实践证明，该算法具有更强的反走样效果．     

煤矿地质剖面图的计算机矢量化方法探讨

本文在分析地质剖面图计算机矢量化现状和存在问题的基础上,提出了快速矢量化煤矿地质剖面图的新思路：首先删除各种岩性符号,保留各层的边界线,最后实现地层边界 线的自动矢量化。     

基于样本区域的线条画风格转换

针对某些线条画存在多于一种风格的线条的情况，提出了一种利用线条画作为样本对任意线条画进行风格转换的方法。该方法基于指定的样本区域和目标区域，通过指定单位长度对笔划分段进行风格相似性比较，并引入旋转和缩放参数进行笔划的规范化表示。与已有的方法比较，该方法算法简单并能产生新的风格转换效果。     

基于样本区域的线条画风格转换

针对某些线条画存在多于一种风格的线条的情况,提出了一种利用线条画作为样本对任意线条画进行风格转换的方法。该方法基于指定的样本区域和目标区域,通过指定单位长度对笔划分段进行风格相似性比较,并引入旋转和缩放参数进行笔划的规范化表示。与已有的方法比较,该方法算法简单并能产生新的风格转换效果。