Showing papers on "Feature extraction published in 1995"

Human and machine recognition of faces: a survey

Abstract: The goal of this paper is to present a critical survey of existing literature on human and machine recognition of faces. Machine recognition of faces has several applications, ranging from static matching of controlled photographs as in mug shots matching and credit card verification to surveillance video images. Such applications have different constraints in terms of complexity of processing requirements and thus present a wide range of different technical challenges. Over the last 20 years researchers in psychophysics, neural sciences and engineering, image processing analysis and computer vision have investigated a number of issues related to face recognition by humans and machines. Ongoing research activities have been given a renewed emphasis over the last five years. Existing techniques and systems have been tested on different sets of images of varying complexities. But very little synergism exists between studies in psychophysics and the engineering literature. Most importantly, there exists no evaluation or benchmarking studies using large databases with the image quality that arises in commercial and law enforcement applications In this paper, we first present different applications of face recognition in commercial and law enforcement sectors. This is followed by a brief overview of the literature on face recognition in the psychophysics community. We then present a detailed overview of move than 20 years of research done in the engineering community. Techniques for segmentation/location of the face, feature extraction and recognition are reviewed. Global transform and feature based methods using statistical, structural and neural classifiers are summarized. >

FastMap: a fast algorithm for indexing, data-mining and visualization of traditional and multimedia datasets

Abstract: A very promising idea for fast searching in traditional and multimedia databases is to map objects into points in k-d space, using k feature-extraction functions, provided by a domain expert [25]. Thus, we can subsequently use highly fine-tuned spatial access methods (SAMs), to answer several types of queries, including the 'Query By Example' type (which translates to a range query); the 'all pairs' query (which translates to a spatial join [8]); the nearest-neighbor or best-match query, etc.However, designing feature extraction functions can be hard. It is relatively easier for a domain expert to assess the similarity/distance of two objects. Given only the distance information though, it is not obvious how to map objects into points.This is exactly the topic of this paper. We describe a fast algorithm to map objects into points in some k-dimensional space (k is user-defined), such that the dis-similarities are preserved. There are two benefits from this mapping: (a) efficient retrieval, in conjunction with a SAM, as discussed before and (b) visualization and data-mining: the objects can now be plotted as points in 2-d or 3-d space, revealing potential clusters, correlations among attributes and other regularities that data-mining is looking for.We introduce an older method from pattern recognition, namely, Multi-Dimensional Scaling (MDS) [51]; although unsuitable for indexing, we use it as yardstick for our method. Then, we propose a much faster algorithm to solve the problem in hand, while in addition it allows for indexing. Experiments on real and synthetic data indeed show that the proposed algorithm is significantly faster than MDS, (being linear, as opposed to quadratic, on the database size N), while it manages to preserve distances and the overall structure of the data-set.

Artificial neural networks for feature extraction and multivariate data projection

Abstract: Classical feature extraction and data projection methods have been well studied in the pattern recognition and exploratory data analysis literature. We propose a number of networks and learning algorithms which provide new or alternative tools for feature extraction and data projection. These networks include a network (SAMANN) for J.W. Sammon's (1969) nonlinear projection, a linear discriminant analysis (LDA) network, a nonlinear discriminant analysis (NDA) network, and a network for nonlinear projection (NP-SOM) based on Kohonen's self-organizing map. A common attribute of these networks is that they all employ adaptive learning algorithms which makes them suitable in some environments where the distribution of patterns in feature space changes with respect to time. The availability of these networks also facilitates hardware implementation of well-known classical feature extraction and projection approaches. Moreover, the SAMANN network offers the generalization ability of projecting new data, which is not present in the original Sammon's projection algorithm; the NDA method and NP-SOM network provide new powerful approaches for visualizing high dimensional data. We evaluate five representative neural networks for feature extraction and data projection based on a visual judgement of the two-dimensional projection maps and three quantitative criteria on eight data sets with various properties. >

Texture segmentation using fractal dimension

Abstract: This paper deals with the problem of recognizing and segmenting textures in images. For this purpose the authors employ a technique based on the fractal dimension (FD) and the multi-fractal concept. Six FD features are based on the original image, the above average/high gray level image, the below average/low gray level image, the horizontally smoothed image, the vertically smoothed image, and the multi-fractal dimension of order two. A modified box-counting approach is proposed to estimate the FD, in combination with feature smoothing in order to reduce spurious regions. To segment a scene into the desired number of classes, an unsupervised K-means like clustering approach is used. Mosaics of various natural textures from the Brodatz album as well as microphotographs of thin sections of natural rocks are considered, and the segmentation results to show the efficiency of the technique. Supervised techniques such as minimum-distance and k-nearest neighbor classification are also considered. The results are compared with other techniques. >

Automatic Extraction of Man-Made Objects from Aerial and Space Images (II)

Abstract: General Topics and Scene Reconstruction- An Overview of DARPA's Research Program in Automatic Population of Geospatial Databases- A Testbed for the Evaluation of Feature Extraction Techniques in a Time Constrained Environment- The Role of Artificial Intelligence in the Reconstruction of Man-Made Objects from Aerial Images- Scene Reconstruction Research - Towards an Automatic System- Semantic Modelling of Man-Made Objects by Production Nets- From Large-Scale DTM Extraction to Feature Extraction- Building Detection and Reconstruction- 3-D Building Reconstruction with ARUBA: A Qualitative and Quantitative Evaluation- A System for Building Detection from Aerial Images- On the Reconstruction of Urban House Roofs from Aerial Images- Image-Based Reconstruction of Informal Settlements- A Model Driven Approach to Extract Buildings from Multi-View Aerial Imagery- Automated Building Extraction from Digital Stereo Imagery- Application of Semi-Automatic Building Acquisition- On the Integration of Object Modeling and Image Modeling in Automated Building Extraction from Aerial Images- TOBAGO - A Topology Builder for the Automated Generation of Building Models- Crestlines Constribution to the Automatic Building Extraction- Recognizing Buildings in Aerial Image- Above-Ground Objects in Urban Scenes from Medium Scale Aerial Imagery- Digital Surface Models for Building Extraction- Extracting Artificial Surface Objects from Airborne Laser Scanner Data- Interpretation of Urban Surface Models using 2D Building Information- Least Squares Matching for Three Dimensional Building Reconstruction- Assessment of the Effects of Resolution on Automated DEM and Building Extraction- Road Extraction- The Role of Grouping for Road Extraction- Artificial Intelligence in 3-D Feature Extraction- Updating Road Maps by Contextual Reasoning- Fast Robust Tracking of Curvy Partially Occluded Roads in Clutter in Aerial Images- Linear Feature Extraction with 3-D LSB-Snakes- Context-Supported Road Extraction- Map/GIS-Based Methods- Three-Dimensional Description of Dense Urban Areas using Maps and Aerial Images- MOSES: A Structural Approach to Aerial Image Understanding- An Approach for the Extraction of Settlement Areas- Extraction of Polygonal Features from Satellite Images for Automatic Registration: The ARCHANGEL Project- Visualisation- A Set of Visualization Data Needs in Urban Environmental Planning & Design for Photogrammetric Data- A Virtual Reality Model of a Major International Airport- Managing Large 3D Urban Database Contents Supporting Phototexture and Levels of Detail- List of Workshop Participants- Author Index

Finding faces in cluttered scenes using random labeled graph matching

Abstract: An algorithm for locating quasi-frontal views of human faces in cluttered scenes is presented. The algorithm works by coupling a set of local feature detectors with a statistical model of the mutual distances between facial features it is invariant with respect to translation, rotation (in the plane), and scale and can handle partial occlusions of the face. On a challenging database with complicated and varied backgrounds, the algorithm achieved a correct localization rate of 95% in images where the face appeared quasi-frontally. >

Optical recognition of motor vehicle license plates

Abstract: A system for the recognition of car license plates is presented The aim of the system is to read automatically the Italian license number of a car passing through a tollgate A CCTV camera and a frame grabber card are used to acquire a rear-view image of the vehicle The recognition process consists of three main phases First, a segmentation phase locates the license plate within the image Then, a procedure based upon feature projection estimates some image parameters needed to normalize the license plate characters Finally, the character recognizer extracts some feature points and uses template matching operators to get a robust solution under multiple acquisition conditions A test has been done on more than three thousand real images acquired under different weather and illumination conditions, thus obtaining a recognition rate close to 91% >

Local discriminant bases and their applications

Abstract: We describe an extension to the “best-basis” method to select an orthonormal basis suitable for signal/image classification problems from a large collection of orthonormal bases consisting of wavelet packets or local trigonometric bases The original best-basis algorithm selects a basis minimizing entropy from such a “library of orthonormal bases” whereas the proposed algorithm selects a basis maximizing a certain discriminant measure (eg, relative entropy) among classes Once such a basis is selected, a small number of most significant coordinates (features) are fed into a traditional classifier such as Linear Discriminant Analysis (LDA) or Classification and Regression Tree (CARTTM) The performance of these statistical methods is enhanced since the proposed methods reduce the dimensionality of the problem at hand without losing important information for that problem Here, the basis functions which are well-localized in the time-frequency plane are used as feature extractors We applied our method to two signal classification problems and an image texture classification problem These experiments show the superiority of our method over the direct application of these classifiers on the input signals As a further application, we also describe a method to extract signal component from data consisting of signal and textured background

Performance of the IBM large vocabulary continuous speech recognition system on the ARPA Wall Street Journal task

Abstract: In this paper we discuss various experimental results using our continuous speech recognition system on the Wall Street Journal task. Experiments with different feature extraction methods, varying amounts and type of training data, and different vocabulary sizes are reported.

Unsupervised texture segmentation of images using tuned matched Gabor filters

Abstract: Recent studies have confirmed that the multichannel Gabor decomposition represents an excellent tool for image segmentation and boundary detection. Unfortunately, this approach when used for unsupervised image analysis tasks imposes excessive storage requirements due to the nonorthogonality of the basis functions and is computationally highly demanding. In this correspondence, we propose a novel method for efficient image analysis that uses tuned matched Gabor filters. The algorithmic determination of the parameters of the Gabor filters is based on the analysis of spectral feature contrasts obtained from iterative computation of pyramidal Gabor transforms with progressive dyadic decrease of elementary cell sizes. The method requires no a priori knowledge of the analyzed image so that the analysis is unsupervised. Computer simulations applied to different classes of textures illustrate the matching property of the tuned Gabor filters derived using our determination algorithm. Also, their capability to extract significant image information and thus enable an easy and efficient low-level image analysis will be demonstrated. >

Automatic handwriting recognition using both static and dynamic parameters

Abstract: Methods and apparatus are disclosed for recognizing handwritten characters in response to an input signal from a handwriting transducer. A feature extraction and reduction procedure is disclosed that relies on static or shape information, wherein the temporal order in which points are captured by an electronic tablet may be disregarded. A method of the invention generates and processes the tablet data with three independent sets of feature vectors which encode the shape information of the input character information. These feature vectors include horizontal (x-axis) and vertical (y-axis) slices of a bit-mapped image of the input character data, and an additional feature vector to encode an absolute y-axis displacement from a baseline of the bit-mapped image. It is shown that the recognition errors that result from the spatial or static processing are quite different from those resulting from temporal or dynamic processing. Furthermore, it is shown that these differences complement one another. As a result, a combination of these two sources of feature vector information provides a substantial reduction in an overall recognition error rate. Methods to combine probability scores from dynamic and the static character models are also disclosed.

Automated recognition of partial discharges

Abstract: An overview of automated recognition of partial discharges (PD) is given. The selection of PD patterns, extraction of relevant information for PD recognition and the structure of a data base for PD recognition are discussed. Mathematical methods useful for the design of the data base are examined. Classification methods are interpreted from a geometrical point of view. Some problems encountered in the automation of PD recognition are addressed. >

A theory of specular surface geometry

Abstract: A theoretical framework is introduced for the perception of specular surface geometry. When an observer moves in three-dimensional space, real scene features, such as surface markings, remain stationary with respect to the surfaces they belong to. In contrast, a virtual feature, which is the specular reflection of a real feature, travels on the surface. Based on the notion of caustics, a novel feature classification algorithm is developed that distinguishes real and virtual features from their image trajectories that result from observer motion. Next, using support functions of curves, a closed-form relation is derived between the image trajectory of a virtual feature and the geometry of the specular surface it travels on. It is shown that in the 2D case where camera motion and the surface profile are coplanar, the profile is uniquely recovered by tracking just two unknown virtual features. Finally, these results are generalized to the case of arbitrary 3D surface profiles that are travelled by virtual features when camera motion is not confined to a plane. An algorithm is developed that uniquely recovers 3D surface profiles using a single virtual feature tracked from the occluding boundary of the object. All theoretical derivations and proposed algorithms are substantiated by experiments. >

A comparison of wavelet transform features for texture image annotation

Abstract: A comparison of different wavelet transform based texture features for content based search and retrieval is made. These include the conventional orthogonal and bi-orthogonal wavelet transforms, tree-structured decompositions, and the Gabor wavelet transforms. Issues discussed include image processing complexity, texture classification and discrimination, and suitability for developing indexing techniques.

A new efficient and direct solution for pose estimation using quadrangular targets: algorithm and evaluation

Abstract: Pose estimation is an important operation for many vision tasks. In this paper, the authors propose an algorithm for pose estimation based on the volume measurement of tetrahedra composed of feature-point triplets extracted from an arbitrary quadrangular target and the lens center of the vision system. The inputs to this algorithm are the six distances joining all feature pairs and the image coordinates of the quadrangular target. The outputs of this algorithm are the effective focal length of the vision system, the interior orientation parameters of the target, the exterior orientation parameters of the camera with respect to an arbitrary coordinate system if the target coordinates are known in this frame, and the final pose of the camera. The authors have also developed a shape restoration technique which is applied prior to pose recovery in order to reduce the effects of inaccuracies caused by image projection. An evaluation of the method has shown that this pose estimation technique is accurate and robust. Because it is based on a unique and closed form solution, its speed makes it a potential candidate for solving a variety of landmark-based tracking problems. >

Road extraction from aerial and satellite images by dynamic programming

Abstract: In this paper, we propose a semi-automatic road extraction scheme which combines the wavelet decomposition for road sharpening and a model-driven linear feature extraction algorithm based on dynamic programming. Semi-automatic means that a road is extracted automatically after some seed points have been given coarsely by the operator through activation of a mouse using a convenient interactive image-graphics user interface. With a wavelet transform interesting image structures can be enhanced and a multiresolution representation can be obtained by selection of a special wavelet. We have built a special wavelet for road sharpening, which has been implemented as a fast pyramidal algorithm. In the model-driven feature extraction scheme, a road is represented by a generic road model with six photometric and geometric properties. This model is formulated by some constraints and a merit function which embodies a notion of the “best road segment”, and evaluated by a “time-delayed” dynamic programming algorithm. The mathematical foundation and issues relating to its practical implementation are discussed in detail. This approach has been applied very successfully to extract complete road networks from single SPOT scenes and aerial images. Thereby the algorithm runs in a monoplotting mode, deriving X, Y, Z -coordinates of the roads, whereby the Z -component comes from real-time interpolation within an underlying DTM. Some experimental results are also given in this paper.

3D Multi-Modality Medical Image Registration Using Feature Space Clustering

Abstract: In this paper, 3D voxel-similarity-based (VB) registration algorithms that optimize a feature-space clustering measure are proposed to combine the segmentation and registration process. We present a unifying definition and a classification scheme for existing VB matching criteria and propose a new matching criterion: the entropy of the grey-level scatter-plot. This criterion requires no segmentation or feature extraction and no a priori knowledge of photometric model parameters. The effects of practical implementation issues concerning grey-level resampling, scatter-plot binning, parzen-windowing and resampling frequencies are discussed in detail and evaluated using real world data (CT and MRI).

Computer-aided method for image feature analysis and diagnosis in mammography

Abstract: A method for automated detection of abnormal anatomic regions, wherein a mammogram is digitized to produce a digital image and the digital image is processed using local edge gradient analysis and linear pattern analysis in addition to feature extraction routines to identify abnormal anatomic regions. Noise reduction filtering and pit-filling/spike-removal filtering techniques are also provided. Multiple difference imaging techniques are also used in which difference images employing different filter characteristics are obtained and processing results logically OR'ed to identify abnormal anatomic regions. In another embodiment the processing results with and without noise reduction filtering are logically AND'ed to improve detection sensitivity. Also, in another embodiment the wavelet transform is utilized in the identification and detection of abnormal regions. The wavelet transform is preferably used in conjunction with the difference imaging technique with the results of the two techniques being logically OR'ed.

Nonlinear manifold learning for visual speech recognition

Abstract: A technique for representing and learning smooth nonlinear manifolds is presented and applied to several lip reading tasks. Given a set of points drawn from a smooth manifold in an abstract feature space, the technique is capable of determining the structure of the surface and of finding the closest manifold point to a given query point. We use this technique to learn the "space of lips" in a visual speech recognition task. The learned manifold is used for tracking and extracting the lips, for interpolating between frames in an image sequence and for providing features for recognition. We describe a system based on hidden Markov models and this learned lip manifold that significantly improves the performance of acoustic speech recognizers in degraded environments. We also present preliminary results on a purely visual lip reader. >

Face detection by fuzzy pattern matching

Abstract: The paper describes an approach to detect faces whose size and position are unknown in an image with a complex background. The candidates of faces are detected by finding out "face like" regions in the input image using the fuzzy pattern matching method. The perceptually uniform color space is used in our research in order to obtain reliable results. The skin color that is used to detect face like regions, is represented by a model developed by us called skin color distribution function. The skin color regions are then extracted by estimating a measure that describes how well the color of a pixel looks like the skin color for each pixel in the input image. The faces which appear in images are modeled as several 2 dimensional patterns. The face like regions are extracted by a fuzzy pattern matching approach using these face models. The face candidates are then verified by estimating how well the extracted facial features fit a face model which describes the geometrical relations among facial features. >

Neural network models in EMG diagnosis

Abstract: In previous years, several computer-aided quantitative motor unit action potential (MUAP) techniques were reported. It is now possible to add to these techniques the capability of automated medical diagnosis so that all data can be processed in an integrated environment. In this study, the parametric pattern recognition (PPR) algorithm that facilitates automatic MUAP feature extraction and Artificial Neural Network (ANN) models are combined for providing an integrated system for the diagnosis of neuromuscular disorders. Two paradigms of learning for training ANN models were investigated, supervised, and unsupervised. For supervised learning, the back-propagation algorithm and for unsupervised learning, the Kohonen's self-organizing feature maps algorithm were used. The diagnostic yield for models trained with both procedures was similar and on the order of 80%. However, back propagation models required considerably more computational effort compared to the Kohonen's self-organizing feature map models. Poorer diagnostic performance was obtained when the K-means nearest neighbor clustering algorithm was applied on the same set of data. >

Iconic techniques for feature visualization

Abstract: This paper presents a conceptual framework and a process model for feature extraction and iconic visualization. Feature extraction is viewed as a process of data abstraction, which can proceed in multiple stages and corresponding data abstraction levels. The features are represented by attribute sets, which play a key role in the visualization process. Icons are symbolic parametric objects, designed as visual representations of features. The attributes are mapped to the parameters (or degrees of freedom) of an icon. We describe some generic techniques to generate attribute sets, such as volume integrals and medial axis transforms. A simple but powerful modeling language was developed to create icons, and to link the attributes to the icon parameters. We present illustrative examples of iconic visualization created with the techniques described, showing the effectiveness of this approach.

Performance evaluation of texture measures for ground cover identification in satellite images by means of a neural network classifier

Abstract: The performance of several feature extraction methods for classifying ground covers in satellite images is compared. Ground covers are viewed as texture of the image. Texture measures considered are: cooccurrence matrices, gray-level differences, texture-tone analysis, features derived from the Fourier spectrum, and Gabor filters. Some Fourier features and some Gabor filters were found to be good choices, in particular when a single frequency band was used for classification. A Thematic Mapper (TM) satellite image showing a variety of vegetations in central Colorado was used for the comparison. A related goal was to investigate the feasibility of extracting the main ground covers from an image. These ground covers may then form an index into a database. This would allow the retrieval of a set of images which are similar in contents. The results obtained in the indexing experiments are encouraging. >

Facial feature extraction method and apparatus for a neural network acoustic and visual speech recognition system

Abstract: A facial feature extraction method and apparatus uses the variation in light intensity (gray-scale) of a frontal view of a speaker's face. The sequence of video images are sampled and quantized into a regular array of 150×150 pixels that naturally form a coordinate system of scan lines and pixel position along a scan line. Left and right eye areas and a mouth are located by thresholding the pixel gray-scale and finding the centroids of the three areas. The line segment joining the eye area centroids is bisected at right angle to form an axis of symmetry. A straight line through the centroid of the mouth area that is at right angle to the axis of symmetry constitutes the mouth line. Pixels along the mouth line and the axis of symmetry in the vicinity of the mouth area form a horizontal and vertical gray-scale profile, respectively. The profiles could be used as feature vectors but it is more efficient to select peaks and valleys (maximas and minimas) of the profile that correspond to the important physiological speech features such as lower and upper lip, mouth corner, and mouth area positions and pixel values and their time derivatives as visual vector components. Time derivatives are estimated by pixel position and value changes between video image frames. A speech recognition system uses the visual feature vector in combination with a concomitant acoustic vector as inputs to a time-delay neural network.

Vanishing point detection by line clustering

Abstract: This paper presents a new method for the detection of vanishing points based on sub-pixel line descriptions which recognizes the existence of errors in feature detection and which does not rely on supervision or the arbitrary specification of thresholds. Image processing and image analysis are integrated into a coherent scheme which extracts straight line structure from images, develops a measure of line quality for each line, estimates the number of vanishing points and their approximate orientations, and then computes optimal vanishing point estimates through combined clustering and numerical optimization. Both qualitative and quantitative evaluation of the algorithms performance is included in the presentation. >

Feature selection in the pattern classification problem of digital chest radiograph segmentation

Abstract: In pattern classification problems, the choice of variables to include in the feature vector is a difficult one. The authors have investigated the use of stepwise discriminant analysis as a feature selection step in the problem of segmenting digital chest radiographs. In this problem, locally calculated features are used to classify pixels into one of several anatomic classes. The feature selection step was used to choose a subset of features which gave performance equivalent to the entire set of candidate features, while utilizing less computational resources. The impact of using the reduced/selected feature set on classifier performance is evaluated for two classifiers: a linear discriminator and a neural network. The results from the reduced/selected feature set were compared to that of the full feature set as well as a randomly selected reduced feature set. The results of the different feature sets were also compared after applying an additional postprocessing step which used a rule-based spatial information heuristic to improve the classification results. This work shows that, in the authors' pattern classification problem, using a feature selection step reduced the number of features used, reduced the processing time requirements, and gave results comparable to the full set of features. >