Section: New Results

Geometry Modeling and Processing

Relaxed lightweight assembly retrieval using vector space model

Participants : Kai-Mo Hu, Bin Wang, Jun-Hai Yong, Jean-Claude Paul.

Assembly searching technologies are important for the improvement of design reusability. However, existing methods require that assemblies possess high-level information, and thus cannot be applied in lightweight assemblies. In this paper, we propose a novel relaxed lightweight assembly retrieval approach based on a vector space model (VSM). By decomposing the assemblies represented in a watertight polygon mesh into bags of parts, and considering the queries as a vague specification of a set of parts, the resilient ranking strategy in VSM is successfully applied in the assembly retrieval. Furthermore, we take the scale-sensitive similarities between parts into the evaluation of matching values, and extend the original VSM to a relaxed matching framework. This framework allows users to input any fuzzy queries, is capable of measuring the results quantitatively, and performs well in retrieving assemblies with specified characteristics. To accelerate the online matching procedure, a typical parts based matching process, as well as a greedy strategy based matching algorithm is presented and integrated in the framework, which makes our system achieve interactive performance. We demonstrate the efficiency and effectiveness of our approach through various experiments on the prototype system. [19]

Calculating Jacobian coefficients of primitive constraints with respect to Euler parameters

Participants : Hai-Chuan Song, Jun-Hai Yong.

It is a fundamental problem to calculate Jacobian coefficients of constraint equations in assembly constraint solving because most approaches to solving an assembly constraint system will finally resort to a numerical iterative method that requires the first-order derivatives of the constraint equations. The most-used method of deriving the Jacobian coefficients is to use virtual rotation which is originally presented to derive the equations of motion of constrained mechanical systems. However, when Euler parameters are adopted as the state variables to represent the transformation matrix, using the virtual rotation will yield erroneous formulae of Jacobian coefficients. The reason is that Euler parameters are incompatible with virtual rotation. In this paper, correct formulae of Jacobian coefficients of geometric constraints with respect to Euler parameters are presented in both Cartesian coordinates and relative generalized coordinates. Experimental results show that our proposed formulae make Newton-Raphson iterative method converge faster and more stable. [22]

An extended schema and its production rule-based algorithms for assembly data exchange using IGES

Participants : Kai-Mo Hu, Bin Wang, Jun-Hai Yong.

Assembly data exchange and reuse play an important role in CAD and CAM in shortening the product development cycle. However, current CAD systems cannot transfer mating conditions via neutral file format, and their exported IGES files are heterogeneous. In this paper, a schema for the full data exchange of assemblies is presented based on IGES. We first design algorithms for the pre-and-post processors of parts based on solid model, in which the topologies are explicitly specified and will be referred by mating conditions, and then extend the IGES schema by introducing the Associativity Definition Entity and Associativity Instance Entity defined in IGES standard, so as to represent mating conditions. Finally, a production rule-based method is proposed to analyze and design the data exchange algorithms for assemblies. Within this schema, the heterogeneous representations of assemblies exported from different CAD systems can be processed appropriately, and the mating conditions can be properly exchanged. Experiments on the prototype system verify the robustness, correctness, and flexibility of our schema. [18]

Robust shape normalization of 3D articulated volumetric models

Participants : Yu-Shen Liu, Jun-Hai Yong, Jean-Claude Paul.

3D shape normalization is a common task in various computer graphics and pattern recognition applications. It aims to normalize different objects into a canonical coordinate frame with respect to rigid transformations containing translation, rotation and scaling in order to guarantee a unique representation. However, the conventional normalization approaches do not perform well when dealing with 3D articulated objects.

To address this issue, we introduce a new method for normalizing a 3D articulated object in the volumetric form. We use techniques from robust statistics to guide the classical normalization computation. The key idea is to estimate the initial normalization by using implicit shape representation, which produces a novel articulation insensitive weight function to reduce the influence of articulated deformation. We also propose and prove the articulation insensitivity of implicit shape representation. The final solution is found by means of iteratively reweighted least squares. Our method is robust to articulated deformation without any explicit shape decomposition. The experimental results and some applications are presented for demonstrating the effectiveness of our method. [24]

$G^1$ continuous approximate curves on NURBS surfaces

Participant : Jun-Hai Yong.

Curves on surfaces play an important role in computer aided geometric design. In this paper, we present a parabola approximation method based on the cubic reparameterization of rational Bézier surfaces, which generates $G^1$ continuous approximate curves lying completely on the surfaces by using iso-parameter curves of the reparameterized surfaces. The Hausdorff distance between the approximate curve and the exact curve is controlled under the user-specified tolerance. Examples are given to show the performance of our algorithm. [28]

The IFC-based path planning for 3D indoor spaces

Participant : Yu-Shen Liu.

Path planning is a fundamental problem, especially for various AEC applications, such as architectural design, indoor and outdoor navigation, and emergency evacuation. However, the conventional approaches mainly operate path planning on 2D drawings or building layouts by simply considering geometric information, while losing abundant semantic information of building components. To address this issue, this paper introduces a new method to cope with path planning for 3D indoor space through an IFC (Industry Foundation Classes) file as input. As a major data exchange standard for Building Information Modeling (BIM), the IFC standard is capable of restoring both geometric information and rich semantic information of building components to support lifecycle data sharing. The method consists of three main steps: (1) extracting both geometric and semantic information of building components defined within the IFC file, (2) discretizing and mapping the extracted information into a planar grid, (3) and finally finding the shortest path based on the mapping for path planning using Fast Marching Method. The paper aims to process different kinds of building components and their corresponding properties to obtain rich semantic information that can enhance applications of path planning. In addition, the IFC-based distributed data sharing and management is implemented for path planning. The paper also presents some experiments to demonstrate the accuracy, efficiency and adaptability of the method. Video demonstration is available from http://cgcad.thss.tsinghua.edu.cn/liuyushen/ifcpath/ . [20]

Recovering Geometric Detail by Octree Normal Maps

Participants : Bin Wang, Jean-Claude Paul.

This paper presents a new approach for constructing normal maps that capture high-frequency geometric detail from dense models of arbitrary topology and are applied to the simplified version of the same models generated by any simplification method to mimic the same level of detail. A variant of loose octree scheme is used to optimally calculate the mesh normals. A B-spline surface fitting based method is employed to solve the issue of thin plate. A memory saving Breadth-First Search (BFS) order construction is designed. Furthermore, a speedup scheme that exploits access coherence is used to accelerate filtering operation. The proposed method can synthesize good quality images of models with extremely high number of polygons while using much less memory and render at much higher frame rate. [31]

An improved example-driven symbol recognition approach in engineering drawings

Participants : Hui Zhang, Ya-Mei Wen.

In this paper, an improved example-driven symbol recognition algorithm is proposed for CAD engineering drawings. First, in order to represent the structure of symbols more clearly and simply, we involve the text entity as one of the basic elements and redefine the relation representation mechanism, which is the foundation for the following algorithms. Then, the structure graph and a constrained tree can be established automatically for the target symbol, using the knowledge acquisition algorithm. In this process, the highest priority element is considered as the key feature, which will be regarded as the root node of the tree. The sequence of breadth first traveling will be recorded to be the recognition rule and saved in the symbol library. In the recognition process, the nodes with the same type as the key features can be located first in the drawing. Unnecessary matching calculations would be greatly reduced because of the accurate location. The other elements around, which satisfy the topology structure of the constrained tree, will be found next. The target symbol is recognized if all of the elements and constraints in the tree are found. Moreover, an extra preprocessing analysis approach is proposed to address repeat modes in a symbol. Thus, similar symbols can be recognized by one rule. We evaluate the proposed approach on the GREC databases and the real engineering drawings. The experimental results validate its effectiveness and efficiency. [17]

3DMolNavi: A web-based retrieval and navigation tool for flexible molecular shape comparison

Participants : Yu-Shen Liu, Jean-Claude Paul.

Many molecules of interest are flexible and undergo significant shape deformation as part of their function, but most existing methods of molecular shape comparison treat them as rigid shapes, which may lead to incorrect measure of the shape similarity of flexible molecules. Currently, there still is a limited effort in retrieval and navigation for flexible molecular shape comparison, which would improve data retrieval by helping users locate the desirable molecule in a convenient way. To address this issue, we develop a web-based retrieval and navigation tool, named 3DMolNavi, for flexible molecular shape comparison. This tool is based on the histogram of Inner Distance Shape Signature (IDSS) for fast retrieving molecules that are similar to a query molecule, and uses dimensionality reduction to navigate the retrieved results in 2D and 3D spaces. We tested 3DMolNavi in the Database of Macromolecular Movements (MolMovDB) and CATH. Compared to other shape descriptors, it achieves good performance and retrieval results for different classes of flexible molecules. The advantages of 3DMolNavi, over other existing softwares, are to integrate retrieval for flexible molecular shape comparison and enhance navigation for user's interaction. [23]

Manifold-ranking based retrieval using k-regular nearest neighbor graph

Participants : Bin Wang, Kai-Mo Hu, Jean-Claude Paul.

Manifold-ranking is a powerful method in semi-supervised learning, and its performance heavily depends on the quality of the constructed graph. In this paper, we propose a novel graph structure named k-regular nearest neighbor (k-RNN) graph as well as its constructing algorithm, and apply the new graph structure in the framework of manifold-ranking based retrieval. We show that the manifold-ranking algorithm based on our proposed graph structure performs better than that of the existing graph structures such as k-nearest neighbor (k-NN) graph and connected graph in image retrieval, 2D data clustering as well as 3D model retrieval. In addition, the automatic sample reweighting and graph updating algorithms are presented for the relevance feedback of our algorithm. Experiments demonstrate that the proposed algorithm outperforms the state-of-the-art algorithms. [25]