Geometric Feature Extraction Research Articles

Self-learning fuzzy control of scan tracking measurement in copying manufacture

Aiming at the effect of the geometric shape of the model surface on the precision of scan tracking control, technologies of model surface geometric feature extraction and predictive control technology are researched. A self-learning fuzzy controller is designed and realized, which automatically adjusts the scan tracking velocity according to the smoothness of the model surface. Finally, through the Matlab simulation experiment, the realizability and application effect of the studied self-learning control method of surface scan tracking velocity are verified.

An Integrated Approach to Computer-Based Automated Tracing and Its Validation for 200 Common Carotid Arterial Wall Ultrasound Images

Most of the algorithms for the segmentation of the common carotid artery (CCA) wall require human interaction to locate the vessel in the ultrasound image. The aim of this article is to show an accurate algorithm for the computer-based automated tracing of the CCA in longitudinal B-mode ultrasound images. Two hundred images (100 normal CCAs, 50 CCAs with an increased intima-media thickness, 30 with fibrous plaques, and 20 with anechoic plaques) were processed to delineate the region of interest containing the CCA. The strategy is an integrated approach (carotid artery layer extraction using an integrated approach [CALEXia]) consisting of geometric feature extraction, line fitting, and classification. The output of the algorithm is the tracings of the proximal and distal adventitia layers. Performance of the algorithm was validated against human tracings considered the ground truth. The mean distance errors +/- SD using this integrated approach were 1.05 +/- 1.04 pixels (0.07 +/- 0.07 mm) for proximal or near adventitia and 2.68 +/- 3.94 pixels (0.17 +/- 0.24 mm) for distal or far adventitia. Sixteen of 200 images were not perfectly traced because of the presence of both plaques and blood backscattering. The computational cost ensures the possibility for near real-time detection. Conclusions. Although the CALEXia algorithm automatically detects the CCA, it is also robust and validated over a large database. This can constitute a general basis for a completely automated segmentation procedure widely applicable to other anatomies.

Feature Extraction from High-density Point Clouds: Toward Automation of an Intelligent 3D Contactless Digitizing Strategy

AbstractThis paper deals with a global intelligent 3D digitizing algorithm, which allows increasing the quality of the resulting cloud of points. Built from quality analysis and characteristic line extraction, the algorithm computes a new path belonging to an admissible space with the objective of increasing the quality of the new resulting points cloud. The extraction work is performed thanks to a scale space algorithm, based on an iterative projection algorithm and the concept of mean curvature motion (MCM). The scale space framework allows us to perform the detection at a coarse scale without any noise or digitizing error interference and to project the result back onto the original point cloud. Thus all the details of the real object’s shape can be identified. Several applications illustrate geometrical feature extraction and global intelligent 3D digitizing within the context of RE. An application is also proposed to compute the distance between the real object shape and an existent CAD-model for con...

SEARCH SIMILAR PROTEIN STRUCTURES WITH CLASSIFICATION, SEQUENCE AND 3D ALIGNMENTS

We have developed an algorithm and web tool to search similar protein structures in the PDB (Protein Data Bank). The algorithm is a combination of a series of methods including protein classification, geometric feature extraction, sequence alignment, and 3D structure alignment. Given a protein structure, the tool can efficiently discover similar structures from hundreds of thousands of structures stored in the PDB. Our experimental results show that it is more accurate than other well-known protein search systems including PSI-BLAST, 3D-BLAST, and SSM in finding proteins that are structurally similar to the query protein, and its speed is also competitive with those systems. The algorithm has been fully implemented and is accessible online at the address http://fpsa.cs.panam.edu/, which is supported by a cluster of computers.

Extraction and Selection of Geometric Moment Features for Forensic Ballistic Specimen Identification

This paper introduces a recognition system for firearms identification based on cartridge case image. When a gun is fired, a unique characteristic on the cartridge and projectile of a bullet are produced. There are more than 30 marks left on the surface of a fired bullet which are invaluable towards identifying the firearm used. These characteristics are known as the mechanical fingerprint of firearm. Therefore, this research has extracted forty eight significantly different features of geometric moment from cartridge case images. Then correlation analysis was used to select the vital and relevant features for classification using discriminant analysis. Keywords: Firearms; cartridge case; geometric moment features; correlation analysis; discriminant analysis

Automatic Target Recognition in Synthetic Aperture Sonar Images Based on Geometrical Feature Extraction

This paper presents a new supervised classification approach for automated target recognition (ATR) in SAS images. The recognition procedure starts with a novel segmentation stage based on the Hilbert transform. A number of geometrical features are then extracted and used to classify observed objects against a previously compiled database of target and non-target features. The proposed approach has been tested on a set of 1528 simulated images created by the NURC SIGMAS sonar model, achieving up to 95% classification accuracy.

A flexible visual inspection system based on neural networks

In this work, we propose a neural networks-based machine vision system, which is intended to act as a reconfigurable inspection tool, for use in manufacturing environments. The processing engine of the system is a second-order neural network, which extracts geometric features invariant to translation and rotation. A major issue with the use of higher-order neural networks is the combinatorial explosion of the higher-order terms, which is addressed here with the use of the alternative image representation strategy of coarse coding. We developed a genetic algorithms tool, which allows the automated determination of the optimal number of hidden units in the neural networks architecture. The inspection system is tested in two application areas, namely inspection of axisymmetric components and classification of rivets. Numerous tests are carried out to evaluate the robustness of the proposed system to complex noise sources.

Real-Time Automatic ECG Diagnosis Method Dedicated to Pervasive Cardiac Care

Recent developments of the wireless sensor network will revolutionize the way of remote monitoring in dif-ferent domains such as smart home and smart care, particularly remote cardiac care. Thus, it is challenging to propose an energy efficient technique for automatic ECG diagnosis (AED) to be embedded into the wireless sensor. Due to the high resource requirements, classical AED methods are unsuitable for pervasive cardiac care (PCC) applications. This paper proposes an embedded real-time AED algorithm dedicated to PCC sys-tems. This AED algorithm consists of a QRS detector and a rhythm classifier. The QRS detector adopts the linear time-domain statistical and syntactic analysis method and the geometric feature extraction modeling technique. The rhythm classifier employs the self-learning expert system and the confidence interval method. Currently, this AED algorithm has been implemented and evaluated on the PCC system for 30 patients in the Gabriel Monpied hospital (CHRU of Clermont-Ferrand, France) and the MIT-BIH cardiac arrhythmias da-tabase. The overall results show that this energy efficient algorithm provides the same performance as the classical ones.

APPLICATIONS OF 3D PCA METHOD FOR EXTRACTION OF MEAN SHAPE AND GEOMETRICAL FEATURES OF BIOLOGICAL OBJECTS SET

This article presents application of the Principal Component Analysis method for analysis of geometry of biological objects and computation of three dimensional anthropometric database. In this work as the biological objects the fifteen human femur bones were used. The geometry of each bone was obtained by using 3D structural light scanner. For PCA analysis all objects have to be described with the same FEM mesh. To achieve this, the modified fluid registration was used. PCA decomposes the set of 3D objects into mean geometry and individual features (empirical modes) describing deviations from mean value. In this paper the mean shape and features of real bones were presented and discussed.

Rotations, translations and symmetry detection for complexified curves

A plane algebraic curve can be represented as the zero-set of a polynomial in two—or if one takes homogeneous coordinates: three—variables. The coefficients of the polynomial determine the curve uniquely. Thus features of the curve, like for instance rotation symmetry, must find their correspondence in the algebraic structure of the coefficients of the polynomial. In this article we will investigate how one can extract geometric curve features from the algebraic description of the curve. In particular, we will use a certain complex representation of polynomials introduced by (Tarel, J.-P., Cooper, D.B., 1998. A new complex basis for implicit polynomial curves and its simple exploitation for pose estimation and invariant recognition. In: Conference on Computer Vision and Pattern Recognition (CVPR'98), pp. 111–117; Unel, M., Wolovich, W.A., 1998. Complex representations of algebraic curves. In: International Conference on Image Processing ICIP 1998, pp. 272–276), which is very appropriate for the task of feature detection. In this complex representation actions on the curve parameters induced by geometric rotations or translations of the plane become very simple. Invariant expressions in the complexified parameters and also normal forms are easily accessible. Furthermore this representation allows the detection of rotation symmetry simply by looking at the indices of all non-vanishing complex parameters.

Live facial feature extraction

Precise facial feature extraction is essential to the high-level face recognition and expression analysis. This paper presents a novel method for the real-time geometric facial feature extraction from live video. In this paper, the input image is viewed as a weighted graph. The segmentation of the pixels corresponding to the edges of facial components of the mouth, eyes, brows, and nose is implemented by means of random walks on the weighted graph. The graph has an 8-connected lattice structure and the weight value associated with each edge reflects the likelihood that a random walker will cross that edge. The random walks simulate an anisotropic diffusion process that filters out the noise while preserving the facial expression pixels. The seeds for the segmentation are obtained from a color and motion detector. The segmented facial pixels are represented with linked lists in the original geometric form and grouped into different parts corresponding to facial components. For the convenience of implementing high-level vision, the geometric description of facial component pixels is further decomposed into shape and registration information. Shape is defined as the geometric information that is invariant under the registration transformation, such as translation, rotation, and isotropic scale. Statistical shape analysis is carried out to capture global facial features where the Procrustes shape distance measure is adopted. A Bayesian approach is used to incorporate high-level prior knowledge of face structure. Experimental results show that the proposed method is capable of real-time extraction of precise geometric facial features from live video. The feature extraction is robust against the illumination changes, scale variation, head rotations, and hand interference.

Perspective methods of human identification: Ear biometrics

AbstractGeometrical methods of feature extraction from ear images in order to perform human identification are presented. Geometrical approach is motivated by the actual procedures used by police and forensic experts (so-called ear otoscopy). In their work, geometrical features of ears such as size, height, width, and shapes of earlobe are useful and valid proofs of identity. The contribution of the article is development of the new and original methods of geometrical feature extraction from 2D ear images. Four novel algorithms of ear feature extraction from contour images are described in detail. Moreover, identification results obtained for each of the methods, based on the distance of feature vectors in the feature space, are presented.

Automated video-based facial expression analysis of neuropsychiatric disorders

Deficits in emotional expression are prominent in several neuropsychiatric disorders, including schizophrenia. Available clinical facial expression evaluations provide subjective and qualitative measurements, which are based on static 2D images that do not capture the temporal dynamics and subtleties of expression changes. Therefore, there is a need for automated, objective and quantitative measurements of facial expressions captured using videos. This paper presents a computational framework that creates probabilistic expression profiles for video data and can potentially help to automatically quantify emotional expression differences between patients with neuropsychiatric disorders and healthy controls. Our method automatically detects and tracks facial landmarks in videos, and then extracts geometric features to characterize facial expression changes. To analyze temporal facial expression changes, we employ probabilistic classifiers that analyze facial expressions in individual frames, and then propagate the probabilities throughout the video to capture the temporal characteristics of facial expressions. The applications of our method to healthy controls and case studies of patients with schizophrenia and Asperger's syndrome demonstrate the capability of the video-based expression analysis method in capturing subtleties of facial expression. Such results can pave the way for a video-based method for quantitative analysis of facial expressions in clinical research of disorders that cause affective deficits.

All-hexahedral remeshing for the finite element analysis of metal forming processes

Hexahedral elements are computationally efficient for three-dimensional modelling of metal forming processes but its utilization is still below expectations due to significant barriers caused by remeshing. Remeshing of hexahedral meshes is generally time consuming when performed manually or semi-automatically, suffers from lack of robustness when performed automatically and is commonly regarded as the bottleneck of the finite element analysis of metal forming processes. This paper presents an algorithm for the automatic generation of all-hexahedral meshes that is capable of representing the deformed geometry of a workpiece during updated Lagrangian finite element calculations without loosing geometrical as well as mechanical consistency. Topics to be addressed include the extraction of the basic geometrical features of the deformed mesh, construction of the core mesh, generation of the boundary mesh linking the core with the surface of the workpiece, application of smoothing procedures to edges, surfaces and volumes and transfer of field variables from the old to the new mesh. The emphasis is focused on providing detailed description of the algorithms and programming solutions for the benefit of those who want to design similar softwares for supporting their metal forming finite element simulation systems. The effectiveness of the proposed algorithm is discussed through its application to a simple but illustrative example derived from the cylindrical side pressing of a rod. Additional and more complex metal forming examples are included to enrich the presentation and to illustrate its scope of application.

Segmentation of random data and extraction of geometric features based on normal vectors

The region growing method is applied in the research of segmentation of random data and extraction of quadric sur- faces geometric features.Firstly,seed regions are located auto- matically,then these regions are grown by recurrent processes composed by iteration and surface fitting.To enhance the reli- ability of this presented method,a direct quadric surface fitting algorithm based on normal vectors of random data points are proposed particularly by using linear least square technique to calculate geometric parameters of various quadric surfaces based on their characteristics.Above algorithm avoid the low efficiency and stability brought by inaccurate setting of initial geometric parameters,and the effect of it is proved by an ex- ample.

As-Built Modeling of Objects for Performance Assessment

The goal of “as-built” computational modeling is to incorporate the most representative geometry and material information for an (fabricated or legacy) object into simulations. While most engineering finite element simulations are based on an object’s idealized “as-designed” configuration with information obtained from technical drawings or computer-aided design models, as-built modeling uses nondestructive characterization and metrology techniques to provide the feature information. By incorporating more representative geometry and material features as initial conditions, the uncertainty in the simulation results can be reduced, providing a more realistic understanding of the event and object being modeled. In this paper, key steps and technology areas in the as-built modeling framework are: (1) inspection using nondestructive characterization and metrology techniques; (2) data reduction (signal and image processing including artifact removal, data sensor fusion, and geometric feature extraction); and (3) engineering and physics analysis using finite element codes. We illustrate the process with a cylindrical phantom and include a discussion of the key concepts and areas that need improvement. Our results show that reasonable as-built initial conditions based on a volume overlap criteria can be achieved and that notable differences between simulations of the as-built and as-designed configurations can be observed for a given load case. Specifically, a volume averaged difference of accumulated plastic strain of 3% and local spatially varying differences up to 10%. The example presented provides motivation and justification to engineering teams for the additional effort required in the as-built modeling of high value parts. Further validation of the approach has been proposed as future work.

Feature-based reverse modeling strategies

We presented two integrated solution schemes, sectional feature based strategy and surface feature based strategy, for modeling industrial components from point cloud to surfaces without using triangulation. For the sectional feature based strategy, slicing, curve feature recognition and constrained fitting are introduced. This strategy emphasizes the advanced feature architecture patterns from 2D to 3D in reverse engineering. The surface feature based strategy relies on differential geometric attributes estimation and diverse feature extraction techniques. The methods and algorithms such as attributes estimation based on 4D Shepard surface, symmetry plane extraction, quadric surface recognition and optimization, extruded and rotational surface extraction, and blend feature extraction with probability and statistic theory are proposed. The reliable three-dimensional feature fabricated the valid substratum of B-rep model faultlessly. All the algorithms are implemented in RE-SOFT, a reverse engineering software developed by Zhejiang University. The proposed strategies can be used to capture the original design intention accurately and to complete the reverse modeling process conveniently. Typical industrial components are used to illustrate the validation of our feature-based strategies.

Ear Biometrics Based on Geometrical Feature Extraction

Biometrics identification methods proved to be very efficient, more natural and easy for users than traditional methods of human identification. In fact, only biometrics methods truly identify humans, not keys and cards they posses or passwords they should remember. The future of biometrics will surely lead to systems based on image analysis as the data acquisition is very simple and requires only cameras, scanners or sensors. More importantly such methods could be passive, which means that the user does not have to take active part in the whole process or, in fact, would not even know that the process of identification takes place. There are many possible data sources for human identification systems, but the physiological biometrics seem to have many advantages over methods based on human behaviour. The most interesting human anatomical parts for such passive, physiological biometrics systems based on images acquired from cameras are face and ear. Both of those methods contain large volume of unique features that allow to distinctively identify many users and will be surely implemented into efficient biometrics systems for many applications. The article introduces to ear biometrics and presents its advantages over face biometrics in passive human identification systems. Then the geometrical method of feature extraction from human ear images in order to perform human identification is presented.

Some challenges of realistic flow simulations by unstructured grid CFD

AbstractThis paper discusses the unstructured grid method to compute flows around geometrically complex bodies having relative motions. To enhance the capability to treat complex geometries, a surface triangulation method using the advancing front method coupled with geometric feature extraction technique is described. Stereolithography (STL) data are adopted as an interface between a CAD system and the surface grid generator. Moving bodies are treated by the overset unstructured grid method. The capability of the method is demonstrated for simulations of an airplane separation process from a rocket booster and a hornet in flight. In the hornet simulation, the detailed components such as antennas, legs and a sting are all included in the computational grid. The flapping wings are treated by the overset unstructured grid method where a grid around the wing is overlapped on a stationary grid around the body of a hornet and moves with time to simulate the flapping motion. Copyright © 2003 John Wiley & Sons, Ltd.

Implementation techniques for geometric branch-and-bound matching methods

Algorithms for geometric matching and feature extraction that work by recursively subdividing transformation space and bounding the quality of match have been proposed in a number of different contexts and become increasingly popular over the last few years. This paper describes matchlist-based branch-and-bound techniques and presents a number of new applications of branch-and-bound methods, among them, a method for globally optimal partial line segment matching under bounded or Gaussian error, point matching under a Gaussian error model with subpixel accuracy and precise orientation models, and a simple and robust technique for finding multiple distinct object instances. It also contains extensive reference information for the implementation of such matching methods under a wide variety of error bounds and transformations. In addition, the paper contains a number of benchmarks and evaluations that provide new information about the runtime behavior of branch-and-bound matching algorithms in general, and that help choose among different implementation strategies, such as the use of point location data structures and space/time tradeoffs involving depth-first search.