This thesis examines an automated method to detect surface deformations on automotive panels for the purpose of quality control along a manufacturing assembly line. Automation in the automotive manufacturing industry is becoming more prominent, but quality control is still largely performed by human workers. Quality control is important in the context of automotive body panels as deformations can occur along the assembly line such as inadequate handling of parts or tools around a vehicle during assembly, rack storage, and shipping from subcontractors. These defects are currently identified and marked, before panels are either rectified or discarded. This work attempts to develop an automated system to detect deformations to alleviate the dependence on human workers in quality control and improve performance by increasing speed and accuracy. Some techniques make use of an ideal CAD model behaving as a master work, and panels scanned on the assembly line are compared to this model to determine the location of deformations. This thesis presents a solution for detecting deformations of various scales without a master work. It also focuses on automated analysis requiring minimal intuitive operator-set parameters and provides the ability to classify the deformations as dings, which are deformations that protrude from the surface, or dents, which are depressions into the surface. A complete automated deformation detection system is proposed, comprised of a feature extraction module, segmentation module, and classification module, which outputs the locations of deformations when provided with the 3D mesh of an automotive panel. Two feature extraction techniques are proposed. The first is a general feature extraction technique for 3D meshes using octrees for multi-resolution analysis and evaluates the amount of surface variation to locate deformations. The second is specifically designed for the purpose of deformation detection, and analyzes multi-resolution cross-sections of a 3D mesh to locate deformations based on their estimated size. The performance of the proposed automated deformation detection system, and all of its sub-modules, is tested on a set of meshes which represent differing characteristics of deformations in surface panels, including deformations of different scales. Noisy, low resolution meshes are captured from a 3D acquisition, while artificial meshes are generated to simulate ideal acquisition conditions. The proposed system shows accurate results in both ideal situations as well as non-ideal situations under the condition of noise and complex surface curvature by extracting only the deformations of interest and accurately classifying them as dings or dents.

Acoustic Event Detection / The detection of the Acoustic Events (AEs) naturally produced in a meeting room may help to describe the human and social activity. The automatic description of interactions between humans and environment can be useful for providing: implicit assistance to the people inside the room, context-aware and content-aware information requiring a minimum of human attention or interruptions, support for high-level analysis of the underlying acoustic scene, etc. On the other hand, the recent fast growth of available audio or audiovisual content strongly demands tools for analyzing, indexing, searching and retrieving the available documents. Given an audio document, the first processing step usually is audio segmentation (AS), i.e. the partitioning of the input audio stream into acoustically homogeneous regions which are labelled according to a predefined broad set of classes like speech, music, noise, etc. Acoustic event detection (AED) is the objective of this thesis work. A variety of features coming not only from audio but also from the video modality is proposed to deal with that detection problem in meeting-room and broadcast news domains. Two basic detection approaches are investigated in this work: a joint segmentation and classification using Hidden Markov Models (HMMs) with Gaussian Mixture Densities (GMMs), and a detection-by-classification approach using discriminative Support Vector Machines (SVMs). For the first case, a fast one-pass-training feature selection algorithm is developed in this thesis to select, for each AE class, the subset of multimodal features that shows the best detection rate. AED in meeting-room environments aims at processing the signals collected by distant microphones and video cameras in order to obtain the temporal sequence of (possibly overlapped) AEs that have been produced in the room. When applied to interactive seminars with a certain degree of spontaneity, the detection of acoustic events from only the audio modality alone shows a large amount of errors, which is mostly due to the temporal overlaps of sounds. This thesis includes several novelties regarding the task of multimodal AED. Firstly, the use of video features. Since in the video modality the acoustic sources do not overlap (except for occlusions), the proposed features improve AED in such rather spontaneous scenario recordings. Secondly, the inclusion of acoustic localization features, which, in combination with the usual spectro-temporal audio features, yield a further improvement in recognition rate. Thirdly, the comparison of feature-level and decision-level fusion strategies for the combination of audio and video modalities. In the later case, the system output scores are combined using two statistical approaches: weighted arithmetical mean and fuzzy integral. On the other hand, due to the scarcity of annotated multimodal data, and, in particular, of data with temporal sound overlaps, a new multimodal database with a rich variety of meeting-room AEs has been recorded and manually annotated, and it has been made publicly available for research purposes.

In this dissertation, we focus on two related parts of a 3D face recognition system with wireless transportation. In the first part, the core components of the system, namely, the feature extraction and classification component, are introduced. In the feature extraction component, range images are taken as inputs and processed in order to extract features. The classification component uses the extracted features as inputs and makes classification decisions based on trained classifiers. In the second part, we consider the wireless transportation problem of range images, which are captured by scattered sensor nodes from target objects and are forwarded to the core components (i.e., feature extraction and classification components) of the face recognition system. Contrary to the conventional definition of being a transducer, a sensor node can be a person, a vehicle, etc. The wireless transportation component not only brings flexibility to the system but also makes the “proactive” face recognition possible. For the feature extraction component, we first introduce the 3D Morphable Model. Then a 3D feature extraction algorithm based on the 3D Morphable Model is presented. The algorithm is insensitive to facial expression. Experimental results show that it can accurately extract features. Following that, we discuss the generic face warping algorithm that can quickly extract features with high accuracy. The proposed algorithm is robust to holes, facial expressions and hair. Furthermore, our experimental results show that the generated features can highly differentiate facial images. For the classification component, a classifier based on Mahalanobis distance is introduced. Based on the classifier, recognition performances of the extracted features are given. The classification results demonstrate the advantage of the features from the generic face warping algorithm. For the wireless transportation of the captured images, we consider the location-based wireless sensor networks (WSN). In order to achieve efficient routing perfor¬mance, a set of distributed stateless routing protocols (PAGER) are proposed for wireless sensor networks. The loop-free and delivery-guaranty properties of the static version (PAGER-S) are proved. Then the performance of PAGER protocols are compared with other well-known routing schemes using network simulator 2 (NS2). Simulation results demonstrate the advantages of PAGER.

In this study a fully automatic and scale invariant feature extractor which does not require manual initialization or special equipment is proposed. Face location and size is extracted using skin segmentation and ellipse fitting. Extracted face region is scaled to a predefined size, later upper and lower facial templates are used for feature extraction. Template localization and template parameter calculations are carried out using Principal Component Analysis. Changes in facial feature coordinates between analyzed image and neutral expression image are used for expression classification. Performances of different classifiers are evaluated. Performance of proposed feature extractor is also tested on sample video sequences. Facial features are extracted in the first frame and KLT tracker is used for tracking the extracted features. Lost features are detected using face geometry rules and they are relocated using feature extractor. As an alternative to feature based technique an available holistic method which analyses face without partitioning is implemented. Face images are filtered using Gabor filters tuned to different scales and orientations. Filtered images are combined to form Gabor jets. Dimensionality of Gabor jets is decreased using Principal Component Analysis. Performances of different classifiers on low dimensional Gabor jets are compared. Feature based and holistic classifier performances are compared using JAFFE and AF facial expression databases.

In this thesis study, a new method for automatic recognition of intrapulse modulations has been proposed. This new method deals the problem of modulation recognition with a feature-based approach. The features used to recognize the modulation type are Instantaneous Frequency, Instantaneous Bandwidth, Amplitude Modulation Depth, Box Dimension and Information Dimension. Instantaneous Bandwidth and Instantaneous Frequency features are extracted via Autoregressive Spectrum Modeling. Amplitude Modulation Depth is used to express the depth of amplitude change on the signal. The other features, Box Dimension and Information Dimension, are extracted using Fractal Theory in order to classify the modulations on signals depending on their shapes. A modulation database is used in association with Fractal Theory to decide on the modulation type of the analyzed signal, by means of a distance metric among fractal dimensions. Utilizing these features in a hierarchical flow, the new modulation recognition method is achieved. The proposed method has been tested for various intrapulse modulation types. It has been observed that the method has acceptably good performance even for low SNR cases and for signals with small PW.

This thesis presents a novel aspect and polarization invariant electromagnetic target recognition technique in resonance region based on use of MUSIC algorithm for the extraction of natural-resonance related target features. In the suggested method, the feature patterns called &ldquo / MUSIC Spectrum Matrices (MSMs)&rdquo / are constructed for each candidate target at each reference aspect angle using targets&rsquo / scattered data at different late-time intervals. These individual MSMs correspond to maps of targets&rsquo / natural-resonance related power distributions. All these patterns are first used to obtain optimal late-time interval for classifier design and a &ldquo / Fused MUSIC Spectrum Matrix (FMSM)&rdquo / is generated over this interval for each target by superposing MSMs. The resulting FMSMs include more complete information for target resonances and are almost insensitive to aspect and polarization. In case of multiple target recognition, the relative locations of a multi-target group and separation distance between targets are also important factors. Therefore, MSM features are computed for each multi-target group at each &ldquo / reference aspect/topology&rdquo / combination to determine the optimum late-time interval. The FMSM feature of a given multi-target group is obtained by the superposition of all these aspect and topology dependent MSMs. In both single and multiple target recognition cases, the resulting FMSM power patterns are main target features of the designed classifier to be used during real-time decisions. At decision phase, the unknown test target is classified either as one of the candidate targets or as an alien target by comparing correlation coefficients computed between MSM of test signal and FMSM of each candidate target.

In this thesis work, automatic recognition algorithms for digital modulated signals are surveyed. Feature extraction and classification algorithm stages are the main parts of a modulation recognition system. Performance of the modulation recognition system mainly depends on the prior knowledge of some of the signal parameters, selection of the key features and classification algorithm selection. Unfortunately, most of the features require some of the signal parameters such as carrier frequency, pulse shape, time of arrival, initial phase, symbol rate, signal to noise ratio, to be known or to be extracted. Thus, in this thesis, features which do not require prior knowledge of the signal parameters, such as the number of the peaks in the envelope histogram and the locations of these peaks, the number of peaks in the frequency histogram, higher order moments of the signal are considered. Particularly, symbol rate and signal to noise ratio estimation methods are surveyed. A method based on the cyclostationarity analysis is used for symbol rate estimation and a method based on the eigenvector decomposition is used for the estimation of signal to noise ratio. Also, estimated signal to noise ratio is used to improve the performance of the classification algorithm. Two methods are proposed for modulation recognition: 1) Decision tree based method 2) Bayesian based classification method A method to estimate the symbol rate and carrier frequency offset of minimum-shift keying (MSK) signal is also investigated.

The polarization of light was introduced last ten years ago in the field of imaging system is a physical phenomenon that can be controlled for the purposes of the vision system. As that found in the human eyes, in general the imaging sensors are not under construction which is sensitive to the polarization of light. These properties can be measured by adding optical components on a conventional camera. The purpose of this thesis is to develop an imaging system that is sensitive both to the stereoscopic and to the state of polarization. As well as the visual system on a various of insects in nature such as bees, that are have capability to move in space by extracted relevant information from the polarization. The developed prototype should be possible to reconstruct threedimensional of points of interest with the issues associated with a set of parameters of the state of polarization. The proposed system consists of two cameras, each camera equipped with liquid crystal components to obtain two images with different directions of polarization. For each acquisition, four images are acquired: two for each camera. Raised by the key of main capability to return polarization information from two different cameras. After an initial calibration step; geometric and photometric, the mapping of points of interest process is made difficult because of the optical components placed in front of different lenses. A detailed study of different methods of mapping was used to select sensitivity to the polarization effects. Once points are mapped, the polarization parameters of each point are calculated from the four values from four images acquired. The results on real scenes show the feasibility and desirability of this imaging system for robotic applications.

No description available.

Time series data arise in many applications, from motion capture, environmental monitoring, temperatures in data centers, to physiological signals in health care. In the thesis, I will focus on the theme of learning and mining large collections of co-evolving sequences, with the goal of developing fast algorithms for finding patterns, summarization, and anomalies. In particular, this thesis will answer the following recurring challenges for time series: 1. Forecasting and imputation: How to do forecasting and to recover missing values in time series data? 2. Pattern discovery and summarization: How to identify the patterns in the time sequences that would facilitate further mining tasks such as compression, segmentation and anomaly detection? 3. Similarity and feature extraction: How to extract compact and meaningful features from multiple co-evolving sequences that will enable better clustering and similarity queries of time series? 4. Scale up: How to handle large data sets on modern computing hardware? We develop models to mine time series with missing values, to extract compact representation from time sequences, to segment the sequences, and to do forecasting. For large scale data, we propose algorithms for learning time series models, in particular, including Linear Dynamical Systems (LDS) and Hidden Markov Models (HMM). We also develop a distributed algorithm for finding patterns in large web-click streams. Our thesis will present special models and algorithms that incorporate domain knowledge. For motion capture, we will describe the natural motion stitching and occlusion filling for human motion. In particular, we provide a metric for evaluating the naturalness of motion stitching, based which we choose the best stitching. Thanks to domain knowledge (body structure and bone lengths), our algorithm is capable of recovering occlusions in mocap sequences, better in accuracy and longer in missing period. We also develop an algorithm for forecasting thermal conditions in a warehouse-sized data center. The forecast will help us control and manage the data center in a energy-efficient way, which can save a significant percentage of electric power consumption in data centers.