JungHyun Han

Manufacturing feature recognition from solid models: a status report

The field of solid modeling has developed a variety of techniques for unambiguous representations of three-dimensional objects. Feature recognition is a sub-discipline of solid modeling that focuses on the design and implementation of algorithms for detecting manufacturing information from solid models produced by computer-aided design (CAD) systems. Examples of this manufacturing information include features such as holes, slots, pockets and other shapes that can be created on modern computer numerically controlled machining systems. Automated feature recognition has been an active research area in solid modeling for many years and is considered to be a critical component for integration of CAD and computer-aided manufacturing. The paper gives an overview of the state-of-the-art in feature recognition research. Rather than giving an exhaustive survey, we focus on the three of the major algorithmic approaches for feature recognition: graph-based algorithms, volumetric decomposition techniques, and hint-based geometric reasoning. For each approach, we present a detailed description of the algorithms being employed along with some assessments of the technology. We conclude by outlining important open research and development issues.

Integration of feature based design and feature recognition

Process planning for machined parts typically requires that a part be described through machining features such as holes, slots and pockets. This paper presents a novel feature finder, which automatically generates a part interpretation in terms of machining features, by utilizing information from a variety of sources such as nominal geometry, tolerances and attributes, and design features. The feature finder strives to produce a desirable interpretation of the part as quickly as possible. If this interpretation is judged unacceptable by a process planner, alternatives can be generated on demand. The feature finder uses a hint-based approach, and combines artificial intelligence techniques, such as blackboard architecture and uncertain reasoning, with the geometric completion procedures first introduced in the OOFF system previously developed at USC.

Role-Based Viewing Envelopes for Information Protection in Collaborative Modeling

Information security and assurance are new frontiers for collaborative design. In this context, information assurance (IA) refers to methodologies to protect engineering information by ensuring its availability, confidentiality, integrity, non-repudiation, authentication, access control, etc. In collaborative design, IA techniques are needed to protect intellectual property, establish security privileges and create “need to know” protections on critical features. Aside from 3D watermarking, research on how to provide IA to distributed collaborative engineering teams is largely non-existent. This paper provides a framework for information assurance within collaborative design, based on a technique we call role-based viewing, in which information security relationships are roles assigned to users based on their permissions and privileges. Role-based viewing is achieved through integration of multi-resolution geometry and with the security model. In this way, 3D models are geometrically partitioned, and the partitioning is used to create multi-resolution mesh hierarchies that obscure, obfuscate, or remove sensitive material from the view of users without appropriate permissions. This approach is the basis for our prototype system FACADE (the Framework for Access-control in Computer-Aided Design Environments), a synchronous, multi-user collaborative modeling environment. In FACADE, groups of users worked in a shared 3D modeling environment in which each user viewing and modeling privileges are managed by a central access control mechanism. In this manner, individual actors see only the data they are allowed to see, at the level of detailed they are permitted to see it.

Feature recognition from CAD models

This article presents a system dubbed IF/sup 2/ (Integrated Incremental Feature Finder), which functions as a feature recognizer or a feature model converter depending on whether the input CAD data contains design features. IF/sup 2/ recognizes three types of machining features: holes, slots and pockets. This article discusses recognition algorithms for slots and pockets. Recognition algorithms for holes are similar to those of slots (Han, 1996).

An approach for interlinking design and process planning

Abstract Interlinking design and process planning plays a key role in realizing Computer Integrated Manufacturing (CIM). Given a part geometry from a CAD system, CAPP generates a sequenced set of instructions to manufacture the specified part. In order to do that, CAPP has to recognize manufacturing features of the part and the relevant information about precision requirements such as surface roughness as well as dimensional and geometric tolerances. Since geometric models from most of the current CAD systems do not incorporate this manufacturing information, human intervention at the first stage of CAPP is inevitable. This has been a major hindrance to information flow between design and process planning. This paper proposes an approach for interlinking CAD and CAPP, and describes the relevant efforts towards it: recognition of machining features, handling of manufacturing information, and implementation of a neutral interface using ISO 10303-224.

Modeler-independent feature recognition in a distributed environment

Abstract Solid modelers and other CAD/CAM subsystems are moving to distributed heterogeneous computing environments, so as to support design and manufacturing processes that are temporally and spatially distributed. Communication and collaboration among the software components of such distributed systems require protocols for accessing remote objects. This paper discusses an approach that provides transparent access to diverse solid modelers in a distributed environment. A solid modeler is augmented with a software wrapper, called an adaptor, so as to provide a uniform application programming interface (API). Applications interact with the uniform API and need not concern themselves with the specifics of the modeling systems used. API calls are implemented in a client-server architecture, in which a modeler and its adaptor function as a geometry server, and various applications communicate with the server through remote procedure calls (RPCs). A few adaptors have been implemented at the University of Southern Californias Programmable Automation Laboratory, and have been used routinely for several years. This paper discusses adaptor design problems and our approach to their solutions. It illustrates the application of our methods through an example that involves the incremental recognition of machinable features in a distributed environment. This environment includes a geometry server, a simple feature-based design system, a state-of-the-art feature recognizer, and a graphics renderer, all running as separate processes in different machines. To our knowledge, this is the first documented effort in which a complex application such as feature recognition is capable of running, unmodified, on top of modelers based on constructive solid geometry or on boundary representations, which are fundamentally different.

Hybrid approach to efficient text extraction in complex color images

Texture-based methods and connected component (CC) methods have been widely used for text localization. However, these two primary methods have their own strength and weakness. This paper proposes a hybrid approach of the two methods for text localization in complex images. An automatically constructed MLP-based texture classifier can increase the recall rates for complex images with much less user intervention and no explicit feature extraction. The CC-based filtering based on the geometry and shape information enhances the precision rates without affecting overall performance. Then, the time-consuming texture analysis for less relevant pixels is avoided by using CAM Shift. Our experimentation shows that the proposed hybrid approach leads to not only robust but also efficient text localization.

A real-time 3D workspace modeling with stereo camera

This paper presents a novel approach to real-time 3D modeling of workspace for manipulative robotic tasks. First, we establish the three fundamental principles that human uses for modeling and interacting with environment. These principles have led to the development of an integrated approach to real-time 3D modeling, as follows: 1) It starts with a rapid but approximate characterization of the geometric configuration of workspace by identifying global plane features. 2) It quickly recognizes known objects in workspace and replaces them by their models in database based on in-situ registration. 3) It models the geometric details on the fly adaptively to the need of the given task based on a multi-resolution octree representation. SIFT features with their 3D position data, referred to here as stereo-sis SIFT, are used extensively, together with point clouds, for fast extraction of global plane features, for fast recognition of objects, for fast registration of scenes, as well as for overcoming incomplete and noisy nature of point clouds. The experimental results show the feasibility of real-time and behavior-oriented 3D modeling of workspace for robotic manipulative tasks.

Hint-based reasoning for feature recognition: Status report

Abstract This technical note reports the feature recognition test results using three systems: IF2, F-Rex and FBMach. Hint-based reasoning extends the portions of an objects boundary that are associated with a machining feature so as to create a material removal volume. Portions of the feature that are missing because of spatial intersections with others are reconstructed. Recognized features also provide useful information about accessibility and other machinability constraints.

Text scanner with text detection technology on image sequences

We propose a text scanner which detects wide text strings in a sequence of scene images. For scene text detection, we use a multiple-CAMShift algorithm on a text probability image produced by a multi-layer perceptron. To provide enhanced resolution of the extracted text images, we perform the text detection process after generating a mosaic image in a fast and robust image registration method.