A structural model for shape recognition using neural nets

Shape representation and recognition is an important topic in many applications of computer vision and artificial intelligence, including character recognition, pattern recognition, machine monitoring, robot manipulation and production part recognition. In this paper, a structural model based on boundary information is proposed to describe the silhouette of planar objects (especially machined parts). The structural model describes objects by a set of primitives, each of which is represented by three geometric features: its length, curvature, and relative orientation. This representation scheme not only compresses the data, but also provides a compact and meaningful form to facilitate further recognition operations. Based on this model, the object recognition is accomplished by using a multilayered feedforward neural network. The proposed model is transformation invariant, which offers the necessary flexibility for real-time implementation in automated manufacturing systems. In addition, the numerical results for a set of ten reference shapes indicate that the matching engine can achieve very high success rates using short recognition times.     

基于交线存储的快速实体边界生成算法

随着ＣＡＤ技术的日益普及，越来越复杂的设计对象进入系统，这就不可避免地导致了系统时空优化问题。目前实体造型中的主要表示方法，即边界模型及构造模型，由于各自特写的结构，无法较好地满足实际工程的需要。本文通过分析实体边界构造过程，提出了一个基个交线存储的实体存储方式及边界重构算法，该方法具有实体数据存储空间小，边界生成速度快等特点，较好地解决了系统存在的时空优化问题。     

An approach to geometric modeling of solids bounded by sculptured surfaces

In this paper we present the design and implementation of a geometric modeling system for modeling solids bounded by sculptured surfaces. The three most important solid representation schemes—constructive solid geometry, boundary representation, and octree representation—are combined together in our system in such a way that the resulting scheme enjoys many of the advantages of its individual schemes. We also developed algorithms for conversion of objects from their boundary representation to octree representation, and for the boundary evaluation of octree encoded objects. The system was implemented on a DEC 1090 computer in PASCAL, and we have presented a simple illustrative example to show the use of the system to create solid objects bounded by sculptured surfaces.     

A neural network approach to CSG-based 3-D object recognition

Describes the recognition subsystem of a computer vision system based on constructive solid geometry (CSG) representation scheme. Instead of using the conventional CSG trees to represent objects, the proposed system uses an equivalent representation scheme-precedence graphs-for object representation. Each node in the graph represents a primitive volume and each are between two nodes represents the relation between them. Object recognition is achieved by matching the scene precedence graph to the model precedence graph. A constraint satisfaction network is proposed to implement the matching process. The energy function associated with the network is used to enforce the matching constraints including match validity, primitive similarity, precedence graph preservation, and geometric structure preservation. The energy level is at its minimum only when the optimal match is reached. Experimental results on several range images are presented to demonstrate the proposed approach     

Mathematics mechanization and applications after thirty years

The aim of mathematics mechanization is to develop symbolic algorithms for manipulating mathematical objects, proving and discovering theorems in a mechanical way. This paper gives a brief review of the major advances in the field over the past thirty years. The characteristic set method for symbolic solution of algebraic, differential, and difference equation systems are first introduced. Methods for automated proving and discovering geometry theorems are then reviewed. Finally, applications in computer-aided geometric design, computer vision, intelligent computer-aided design, and robotics are surveyed.     

Computer aided analysis and derivation for artificial neuralsystems

The theoretical analysis and derivation of artificial neural systems, which consists essentially of manipulating symbolic mathematical objects according to certain mathematical and biological knowledge, can be done more efficiently with computer assistance by using and extending methods and systems of symbolic computation. After presenting the mathematical characteristics of neural systems and a brief review on Lyapunov stability theory, the authors present some features and capabilities of existing systems and the extension for manipulating objects occurring in the analysis of neural systems. Some strategies and a toolkit developed in MACSYMA for computer-aided analysis and derivation are described. A concrete example is given to demonstrate the derivation of a hybrid neural system, i.e. a system which in its learning rule combines elements of supervised and unsupervised learning. Future work and research directions are indicated     

A volumetric/iconic frequency domain representation for objectswith application for pose invariant face recognition

A novel method for representing 3D objects that unifies viewer and model centered object representations is presented. A unified 3D frequency-domain representation, called volumetric frequency representation (VFR), encapsulates both the spatial structure of the object and a continuum of its views in the same data structure. The frequency-domain image of an object viewed from any direction can be directly extracted employing an extension of the projection slice theorem, where each Fourier-transformed view is a planar slice of the volumetric frequency representation. The VFR is employed for pose-invariant recognition of complex objects, such as faces. The recognition and pose estimation is based on an efficient matching algorithm in a four-dimensional Fourier space. Experimental examples of pose estimation and recognition of faces in various poses are also presented     

Determining interference between pairs of solids defined constructively in computer animation

This paper presents theory and implementation of a method for detecting interference between a pair of solid objects. Often at times, when performing simulations, two solids may unwittingly interpenetrate each other. The two components of the system presented in this paper are: (1) a surface representation method to model solid objects; and (2) a method for detecting interference. Body representation of a solid in this system is based upon enveloping each solid with surfaces (called positive entities). Most computer aided design (CAD) systems use solid modeling techniques to represent solid objects. Since most solid models use Boolean operations to model complex objects, a method is presented to envelop complex objects with parametric surfaces. A method for tracing intersection curves between two surfaces is also presented. Discontinuities on surfaces are defined as negative entitics in order to extend the method to complex solids. Determining interference is based upon a numerical algorithm for computing points of intersection between boundary curves and parametrized entities. The existence of segments of these curves inside the boundary of positive and negative entities is established by computing the circulation of a function around the boundary curve. Interference between two solids is then detected. No limitations are imposed on the convexity or simplicity of the boundary curves treated.     

拓扑关系在图形对象表示与识别中的应用

工程图纸识别中,图形对象的表示方法是研究的重点和难点。在矢量化的基础上提出了一种基于拓扑结构的图形对象表示与识别方法,该方法通过矢量化获得各个图形对象的矢量基元,通过建立矢量基元的影响区域确定各矢量基元的拓扑邻接关系图,进而确定图形对象之间的位置关系,为进一步理解识别工程图纸提供了基本的信息。然后介绍了一个应用实例：圆弧角的识别。在实验中,分别用标准数据和真实图纸进行了测试,实验结果显示该算法具有较高的识别精度和识别效率。     

General sweep mathematical morphology

General sweep mathematical morphology provides a new class of morphological operations, which allow one to select varying shapes and orientations of structuring elements during the sweeping process. Such a class holds syntactic characteristics similar to algebraic morphology as well as sweep geometric modeling. The conventional morphology is a subclass of the general sweep morphology. The sweep morphological dilation/erosion provides a natural representation of sweep motion in the manufacturing processes, and the sweep opening/closing provides variant degrees of smoothing in image filtering. The theoretical framework for representation, computation and analysis of sweep morphology is presented in this paper. Its applications to the sweeping with deformations, image enhancement, edge linking, and shortest path planning for rotating objects are also discussed.     

Modeling and fabrication of artistic products based on IFS fractal representation

Contemporary computer-aided design (CAD) and computer-aided manufacturing (CAM) theories and systems are well developed for analytical and free-form objects, but neither can deal with all artistically appealing objects efficiently. Artistically appealing objects are common in, for example, jewellery and furniture decoration. Therefore, it is valuable to explore more suitable modeling and manufacturing method for artistically appealing objects.Fractal geometry has been employed for modeling natural objects that cannot be described easily by Euclidean geometry. In this paper, one type of fractal solid — the Iterated Function System (IFS) fractal — is proposed for modeling artistically appealing objects in a computational form. A Radial-Blossoming Tree (RBT) data structure is worked out for fractal solid modeling in a CAD platform. Traversal algorithms have been devised to extract necessary information from the RBT for generating the toolpath for a layered manufacturing (LM) process, so that a physical fractal object can be built and the fabrication of a fractally represented artistic product can be realized.     

Computer-aided design of porous artifacts

Heterogeneous structures represent an important new frontier for 21st century engineering. Human tissues, composites, ‘smart’ and multi-material objects are all physically manifest in the world as three-dimensional (3D) objects with varying surface, internal and volumetric properties and geometries. For instance, a tissue engineered structure, such as bone scaffold for guided tissue regeneration, can be described as a heterogeneous structure consisting of 3D extra-cellular matrices (made from biodegradable material) and seeded donor cells and/or growth factors.The design and fabrication of such heterogeneous structures requires new techniques for solid models to represent 3D heterogeneous objects with complex material properties. This paper presents a representation of model density and porosity based on stochastic geometry. While density has been previously studied in the solid modeling literature, porosity is a relatively new problem. Modeling porosity of bio-materials is critical for developing replacement bone tissues. The paper uses this representation to develop an approach to modeling of porous, heterogeneous materials and provides experimental data to validate the approach. The authors believe that their approach introduces ideas from the stochastic geometry literature to a new set of engineering problems. It is hoped that this paper stimulates researchers to find new opportunities that extend these ideas to be more broadly applicable for other computational geometry, graphics and computer-aided design problems.     

基于均衡化及柔性形态学的显微图像边缘检测

对显微图像进行噪声过滤和增强是对其进行的分类、识别、检测处理的基础,在分析、综合传统的图像增强和图像分割的算法的基础上,将直方图变换和柔性数学形态学组合,提出了基于均衡化及柔性数学形态学的显微图像边缘检测方法,并通过实现表明该方法能够有效的抑制微生物显微图像的噪声,提高检测精度,保护边缘细节,并且易于编程实现.     

Scalable representation for 3D object recognition using feature sharing and view clustering

Scalability is an important issue in object recognition as it reduces database storage and recognition time. In this paper, we propose a new scalable 3D object representation and a learning method to recognize many everyday objects. The key proposal for scalable object representation is to combine the concept of feature sharing with multi-view clustering in part-based object representation, in particular a common-frame constellation model (CFCM). In this representation scheme, we also propose a fully automatic learning method: appearance-based automatic feature clustering and sequential construction of clustered CFCMs from labeled multi-views and multiple objects. We evaluated the scalability of the proposed method to COIL-100 DB and applied the learning scheme to 112 objects with 620 training views. Experimental results show the scalable learning results in almost constant recognition performance relative to the number of objects.     

Connectivity in the regular polytope representation

In order to be able to draw inferences about real world phenomena from a representation expressed in a digital computer, it is essential that the representation should have a rigorously correct algebraic structure. It is also desirable that the underlying algebra be familiar, and provide a close modelling of those phenomena. The fundamental problem addressed in this paper is that, since computers do not support real-number arithmetic, the algebraic behaviour of the representation may not be correct, and cannot directly model a mathematical abstraction of space based on real numbers. This paper describes a basis for the robust geometrical construction of spatial objects in computer applications using a complex called the “Regular Polytope”. In contrast to most other spatial data types, this definition supports a rigorous logic within a finite digital arithmetic. The definition of connectivity proves to be non-trivial, and alternatives are investigated. It is shown that these alternatives satisfy the relations of a region connection calculus (RCC) as used for qualitative spatial reasoning, and thus introduce the rigor of that reasoning to geographical information systems. They also form what can reasonably be termed a “Finite Boolean Connection Algebra”. The rigorous and closed nature of the algebra ensures that these primitive functions and predicates can be combined to any desired level of complexity, and thus provide a useful toolkit for data retrieval and analysis. The paper argues for a model with two and three-dimensional objects that have been coded in Java and which implement a full set of topological and connectivity functions which is shown to be complete and rigorous.     

Using symbolic computation to find algebraic invariants

Implicit polynomials have proved themselves as having excellent representation power for complicated objects, and there is growing use of them in computer vision, graphics, and CAD. A must for every system that tries to recognize objects based on their representation by implicit polynomials are invariants, which are quantities assigned to polynomials that do not change under coordinate transformations. In the recognition system developed at the Laboratory for Engineering Man-Machine Studies in Brown University (LEMS), it became necessary to use invariants which are explicit and simple functions of the polynomial coefficients. A method to find such invariants is described and the new invariants presented. This work addresses only the problem of finding the invariants; their stability is studied in another paper     

Learning Sight from Sound: Ambient Sound Provides Supervision for Visual Learning

The sound of crashing waves, the roar of fast-moving cars—sound conveys important information about the objects in our surroundings. In this work, we show that ambient sounds can be used as a supervisory signal for learning visual models. To demonstrate this, we train a convolutional neural network to predict a statistical summary of the sound associated with a video frame. We show that, through this process, the network learns a representation that conveys information about objects and scenes. We evaluate this representation on several recognition tasks, finding that its performance is comparable to that of other state-of-the-art unsupervised learning methods. Finally, we show through visualizations that the network learns units that are selective to objects that are often associated with characteristic sounds. This paper extends an earlier conference paper, Owens et al. (in: European conference on computer vision, 2016b), with additional experiments and discussion.