Automatic form-feature recognition using neural-network-based techniques on boundary representations of solid models

1992 ◽  
Vol 24 (7) ◽  
pp. 381-393 ◽  
Author(s):  
S. Prabhakar ◽  
M.R. Henderson
Keyword(s):  
Neural Network ◽  
Feature Recognition ◽  
Solid Models ◽  
Boundary Representations ◽  
Form Feature

Form Feature Recognition Using Base Volume Decomposition

1994 ◽  
Author(s):  
James K. Coles ◽  
Richard H. Crawford ◽  
Kristin L. Wood
Keyword(s):  
Real World ◽  
Feature Recognition ◽  
Recognition Method ◽  
Mechanical Parts ◽  
Feature Representations ◽  
Volume Decomposition ◽  
Boundary Representations ◽  
New Feature ◽  
Form Feature

Abstract A new feature recognition method is presented that generates volumetric feature representations from conventional boundary representations of mechanical parts. Recognition is accomplished by decomposing the known total feature volume of a part into a set of smaller volumes through analytic face extension. The decomposed volumes are combined to generate an initial set of features. Alternative sets of features are generated by maintaining and evaluating information on intersections of the initial feature set. The capabilities of the method are demonstrated through both a hypothetical and a real world design example. The method’s ability to locate features despite interactions with other features, and its ability to generate alternative sets of features, distinguishes it from existing recognition techniques.

MMCs and PPCs As Constructs of Curvature Regions for Form Feature Determination

1997 ◽  
Author(s):  
Ratnakar Sonthi ◽  
Rajit Gadh
Keyword(s):  
Feature Recognition ◽  
Geometric Model ◽  
Boundary Representation ◽  
Solid Model ◽  
Cad Model ◽  
Solid Models ◽  
Feature Information ◽  
Primitive Shape ◽  
Region Representation ◽  
Form Feature

Abstract Shape feature information about a part is required to analyze the part for downstream issues such as manufacturability and assemblability. One method of obtaining the feature information is feature recognition from the geometric model. This paper presents an approach called Curvature Region (CR) approach for feature determination in solid models. The CR-approach categorizes features into two primitive shape classes: protrusions and depressions. In the first step, these primitive shape classes are recognized from the solid model. In the next step, the primitive shape classes are analyzed using rules to obtain features. Primitive features are found by first converting the boundary representation (B-Rep) of the CAD model to a higher level of representation called Curvature Region Representation (CR-Rep). Curvature Regions are then grouped together to form Minus-Minus Centers (MMCs) and Plus-Plus Centers (PPCs). Primitive shapes are then defined in terms of these centers.

Midsurface Abstraction From 3D Solid Models for Engineering Analysis

1996 ◽  
Author(s):  
Mohsen Rezayat
Keyword(s):  
Feature Recognition ◽  
Analysis Model ◽  
Thin Sections ◽  
New Approach ◽  
Solid Models ◽  
Product And Process ◽  
Plastic Parts ◽  
3D Solid ◽  
User Friendly ◽  
Analysis Models

Abstract An integral part of implementing parallel product and process designs is simulation through numerical analysis. This simulation-driven design requires discretization of the 3D part in an appropriate manner. If the part is thin or has thin sections (e.g., plastic parts), then an analysis model with reduced dimensionality may be more accurate and economical than a standard 3D model. In addition, substantial simplification of some details in the design geometry may be beneficial and desirable in the analysis model. Unfortunately, the majority of CAD systems do not provide the means for abstraction of appropriate analysis models. In this paper we present a new approach, based on midsurface abstraction, which holds significant promise in simplifying simulation-driven design. The method is user-friendly because very little interaction is required to guide the software in its automatic creation of the desired analysis model. It is also robust because it handles typical parts with complex and interacting features. Application of the method for feature recognition and abstraction is also briefly discussed.

scholarly journals Unsupervised feature recognition in single-molecule break junction data

Nanoscale ◽  
2020 ◽  
Vol 12 (15) ◽  
pp. 8355-8363 ◽  
Author(s):  
András Magyarkuti ◽  
Nóra Balogh ◽  
Zoltán Balogh ◽  
Latha Venkataraman ◽  
András Halbritter
Keyword(s):  
Neural Network ◽  
Network Analysis ◽  
Single Molecule ◽  
Feature Recognition ◽  
Principal Component ◽  
Break Junction ◽  
Neural Network Analysis ◽  
Efficient Tool

A combined principal component and neural network analysis serves as an efficient tool for the unsupervised recognition of unobvious but highly relevant trace classes in single-molecule break junction data.

scholarly journals Chinese Language Feature Analysis Based on Multilayer Self-Organizing Neural Network and Data Mining Techniques

2021 ◽  
Vol 2021 ◽  
pp. 1-9
Author(s):  
Xiujin Yu ◽  
Shengfu Liu ◽  
Hui Zhang
Keyword(s):  
Neural Network ◽  
Data Mining ◽  
Feature Recognition ◽  
High Efficiency ◽  
Chinese Language ◽  
Strong Support ◽  
Superior Performance ◽  
Feature Analysis ◽  
Data Mining Techniques ◽  
Self Organizing

As one of the oldest languages in the world, Chinese has a long cultural history and unique language charm. The multilayer self-organizing neural network and data mining techniques have been widely used and can achieve high-precision prediction in different fields. However, they are hardly applied to Chinese language feature analysis. In order to accurately analyze the characteristics of Chinese language, this paper uses the multilayer self-organizing neural network and the corresponding data mining technology for feature recognition and then compared it with other different types of neural network algorithms. The results show that the multilayer self-organizing neural network can make the accuracy, recall, and F1 score of feature recognition reach 68.69%, 80.21%, and 70.19%, respectively, when there are many samples. Under the influence of strong noise, it keeps high efficiency of feature analysis. This shows that the multilayer self-organizing neural network has superior performance and can provide strong support for Chinese language feature analysis.

Handling Blending Features in Form Feature Recognition Using Convex Decomposition

1994 ◽  
Author(s):  
Sreekumar Menon ◽  
Yong Se Kim
Keyword(s):  
Feature Recognition ◽  
Boundary Representation ◽  
Recognition Method ◽  
Geometric Information ◽  
Polyhedral Model ◽  
Curved Boundary ◽  
Convex Decomposition ◽  
Geometric Objects ◽  
Form Features ◽  
Form Feature

Abstract Form features intrinsic to the product shape can be recognized using a convex decomposition called Alternating Sum of Volumes with Partitioning (ASVP). However, the domain of geometric objects to which ASVP decomposition can be applied had been limited to polyhedral solids due to the difficulty of convex hull construction for solids with curved boundary faces. We develop an approach to extend the geometric domain to solids having cylindrical and blending features. Blending surfaces are identified and removed from the boundary representation of the solid, and a polyhedral model of the unblended solid is generated while storing the cylindrical geometric information. From the ASVP decomposition of the polyhedral model, polyhedral form features are recognized. Form feature decomposition of the original solid is then obtained by reattaching the stored blending and cylindrical information to the form feature components of its polyhedral model. In this way, a larger domain of solids can be covered by the feature recognition method using ASVP decomposition. In this paper, handling of blending features in this approach is described.

scholarly journals Key Feature Recognition Algorithm of Network Intrusion Signal Based on Neural Network and Support Vector Machine

Symmetry ◽  
2019 ◽  
Vol 11 (3) ◽  
pp. 380 ◽  
Author(s):  
Kai Ye
Keyword(s):  
Neural Network ◽  
Support Vector Machine ◽  
Feature Recognition ◽  
False Positive Rate ◽  
Principal Component ◽  
Recognition Algorithm ◽  
Support Vector ◽  
Data Set ◽  
Network Intrusion ◽  
Key Features

When identifying the key features of the network intrusion signal based on the GA-RBF algorithm (using the genetic algorithm to optimize the radial basis) to identify the key features of the network intrusion signal, the pre-processing process of the network intrusion signal data is neglected, resulting in an increase in network signal data noise, reducing the accuracy of key feature recognition. Therefore, a key feature recognition algorithm for network intrusion signals based on neural network and support vector machine is proposed. The principal component neural network (PCNN) is used to extract the characteristics of the network intrusion signal and the support vector machine multi-classifier is constructed. The feature extraction result is input into the support vector machine classifier. Combined with PCNN and SVM (Support Vector Machine) algorithms, the key features of network intrusion signals are identified. The experimental results show that the algorithm has the advantages of high precision, low false positive rate and the recognition time of key features of R2L (it is a common way of network intrusion attack) data set is only 3.18 ms.

Sign in / Sign up

Export Citation Format

Share Document