基于模型的位姿估计中优化方法研究

介绍了基于模型的位姿估计中所使用的一些优化方法。为了提高位姿估计的精度,摄像机的标定参数必须足够精确,这就对标定过程的非线性优化算法提出了很高的要求,采用了一种新的优化目标函数,用来最小化控制点间的三维重建误差,从而使标定参数是全局最优;在双像机位姿估计中,引入了实时遗传算法进行全局搜索,加快了算法的收敛速度。最后的实验证明了这些方法的正确性并显示出这些方法在精度上比传统方法有了较大程度的提高。

Towards an Example-Based Image Compression Architecture for Video-Conferencing

This paper consists of two major parts. First, we present the outline of a simple approach to very-low bandwidth video-conferencing system relying on an example-based hierarchical image compression scheme. In particular, we discuss the use of example images as a model, the number of required examples, faces as a class of semi-rigid objects, a hierarchical model based on decomposition into different time-scales, and the decomposition of face images into patches of interest. In the second part, we present several algorithms for image processing and animation as well as experimental evaluations. Among the original contributions of this paper is an automatic algorithm for pose estimation and normalization. We also review and compare different algorithms for finding the nearest neighbors in a database for a new input as well as a generalized algorithm for blending patches of interest in order to synthesize new images. Finally, we outline the possible integration of several algorithms to illustrate a simple model-based video-conference system.

Inferring 3D Structure with a Statistical Image-Based Shape Model

We present an image-based approach to infer 3D structure parameters using a probabilistic "shape+structure'' model. The 3D shape of a class of objects may be represented by sets of contours from silhouette views simultaneously observed from multiple calibrated cameras. Bayesian reconstructions of new shapes can then be estimated using a prior density constructed with a mixture model and probabilistic principal components analysis. We augment the shape model to incorporate structural features of interest; novel examples with missing structure parameters may then be reconstructed to obtain estimates of these parameters. Model matching and parameter inference are done entirely in the image domain and require no explicit 3D construction. Our shape model enables accurate estimation of structure despite segmentation errors or missing views in the input silhouettes, and works even with only a single input view. Using a dataset of thousands of pedestrian images generated from a synthetic model, we can perform accurate inference of the 3D locations of 19 joints on the body based on observed silhouette contours from real images.

Robust and Efficient 3D Recognition by Alignment

Alignment is a prevalent approach for recognizing 3D objects in 2D images. A major problem with current implementations is how to robustly handle errors that propagate from uncertainties in the locations of image features. This thesis gives a technique for bounding these errors. The technique makes use of a new solution to the problem of recovering 3D pose from three matching point pairs under weak-perspective projection. Furthermore, the error bounds are used to demonstrate that using line segments for features instead of points significantly reduces the false positive rate, to the extent that alignment can remain reliable even in cluttered scenes.

BoostMap: A Method for Efficient Approximate Similarity Rankings

This paper introduces BoostMap, a method that can significantly reduce retrieval time in image and video database systems that employ computationally expensive distance measures, metric or non-metric. Database and query objects are embedded into a Euclidean space, in which similarities can be rapidly measured using a weighted Manhattan distance. Embedding construction is formulated as a machine learning task, where AdaBoost is used to combine many simple, 1D embeddings into a multidimensional embedding that preserves a significant amount of the proximity structure in the original space. Performance is evaluated in a hand pose estimation system, and a dynamic gesture recognition system, where the proposed method is used to retrieve approximate nearest neighbors under expensive image and video similarity measures. In both systems, BoostMap significantly increases efficiency, with minimal losses in accuracy. Moreover, the experiments indicate that BoostMap compares favorably with existing embedding methods that have been employed in computer vision and database applications, i.e., FastMap and Bourgain embeddings.

Indexing Methods for Efficient Multiclass Recognition

Many real world image analysis problems, such as face recognition and hand pose estimation, involve recognizing a large number of classes of objects or shapes. Large margin methods, such as AdaBoost and Support Vector Machines (SVMs), often provide competitive accuracy rates, but at the cost of evaluating a large number of binary classifiers, thus making it difficult to apply such methods when thousands or millions of classes need to be recognized. This thesis proposes a filter-and-refine framework, whereby, given a test pattern, a small number of candidate classes can be identified efficiently at the filter step, and computationally expensive large margin classifiers are used to evaluate these candidates at the refine step. Two different filtering methods are proposed, ClassMap and OVA-VS (One-vs.-All classification using Vector Search). ClassMap is an embedding-based method, works for both boosted classifiers and SVMs, and tends to map the patterns and their associated classes close to each other in a vector space. OVA-VS maps OVA classifiers and test patterns to vectors based on the weights and outputs of weak classifiers of the boosting scheme. At runtime, finding the strongest-responding OVA classifier becomes a classical vector search problem, where well-known methods can be used to gain efficiency. In our experiments, the proposed methods achieve significant speed-ups, in some cases up to two orders of magnitude, compared to exhaustive evaluation of all OVA classifiers. This was achieved in hand pose recognition and face recognition systems where the number of classes ranges from 535 to 48,600.

Learning a Family of Detectors

Object detection and recognition are important problems in computer vision. The challenges of these problems come from the presence of noise, background clutter, large within class variations of the object class and limited training data. In addition, the computational complexity in the recognition process is also a concern in practice. In this thesis, we propose one approach to handle the problem of detecting an object class that exhibits large within-class variations, and a second approach to speed up the classification processes. In the first approach, we show that foreground-background classification (detection) and within-class classification of the foreground class (pose estimation) can be jointly solved with using a multiplicative form of two kernel functions. One kernel measures similarity for foreground-background classification. The other kernel accounts for latent factors that control within-class variation and implicitly enables feature sharing among foreground training samples. For applications where explicit parameterization of the within-class states is unavailable, a nonparametric formulation of the kernel can be constructed with a proper foreground distance/similarity measure. Detector training is accomplished via standard Support Vector Machine learning. The resulting detectors are tuned to specific variations in the foreground class. They also serve to evaluate hypotheses of the foreground state. When the image masks for foreground objects are provided in training, the detectors can also produce object segmentation. Methods for generating a representative sample set of detectors are proposed that can enable efficient detection and tracking. In addition, because individual detectors verify hypotheses of foreground state, they can also be incorporated in a tracking-by-detection frame work to recover foreground state in image sequences. To run the detectors efficiently at the online stage, an input-sensitive speedup strategy is proposed to select the most relevant detectors quickly. The proposed approach is tested on data sets of human hands, vehicles and human faces. On all data sets, the proposed approach achieves improved detection accuracy over the best competing approaches. In the second part of the thesis, we formulate a filter-and-refine scheme to speed up recognition processes. The binary outputs of the weak classifiers in a boosted detector are used to identify a small number of candidate foreground state hypotheses quickly via Hamming distance or weighted Hamming distance. The approach is evaluated in three applications: face recognition on the face recognition grand challenge version 2 data set, hand shape detection and parameter estimation on a hand data set, and vehicle detection and estimation of the view angle on a multi-pose vehicle data set. On all data sets, our approach is at least five times faster than simply evaluating all foreground state hypotheses with virtually no loss in classification accuracy.

Rao-blackwellized particle filter for human appearance and position tracking

In human motion analysis, the joint estimation of appearance, body pose and location parameters is not always tractable due to its huge computational cost. In this paper, we propose a Rao-Blackwellized Particle Filter for addressing the problem of human pose estimation and tracking. The advantage of the proposed approach is that Rao-Blackwellization allows the state variables to be splitted into two sets, being one of them analytically calculated from the posterior probability of the remaining ones. This procedure reduces the dimensionality of the Particle Filter, thus requiring fewer particles to achieve a similar tracking performance. In this manner, location and size over the image are obtained stochastically using colour and motion clues, whereas body pose is solved analytically applying learned human Point Distribution Models.

2D silhouette and 3D skeletal models for human detection and tracking

In this paper we propose a statistical model for detection and tracking of human silhouette and the corresponding 3D skeletal structure in gait sequences. We follow a point distribution model (PDM) approach using a Principal Component Analysis (PCA). The problem of non-lineal PCA is partially resolved by applying a different PDM depending of pose estimation; frontal, lateral and diagonal, estimated by Fisher's linear discriminant. Additionally, the fitting is carried out by selecting the closest allowable shape from the training set by means of a nearest neighbor classifier. To improve the performance of the model we develop a human gait analysis to take into account temporal dynamic to track the human body. The incorporation of temporal constraints on the model increase reliability and robustness.

An Investigation Into Features For Multi-View Lipreading

For the first time in this paper we present results showing the effect of speaker head pose angle on automatic lip-reading performance over a wide range of closely spaced angles. We analyse the effect head pose has upon the features themselves and show that by selecting coefficients with minimum variance w.r.t. pose angle, recognition performance can be improved when train-test pose angles differ. Experiments are conducted using the initial phase of a unique multi view Audio-Visual database designed specifically for research and development of pose-invariant lip-reading systems. We firstly show that it is the higher order horizontal spatial frequency components that become most detrimental as the pose deviates. Secondly we assess the performance of different feature selection masks across a range of pose angles including a new mask based on Minimum Cross-Pose Variance coefficients. We report a relative improvement of 50% in Word Error Rate when using our selection mask over a common energy based selection during profile view lip-reading.

Kinect and Episodic Reasoning for Human Action Recognition

This paper presents a method for rational behaviour recognition that combines vision-based pose estimation with knowledge modeling and reasoning. The proposed method consists of two stages. First, RGB-D images are used in the estimation of the body postures. Then, estimated actions are evaluated to verify that they make sense. This method requires rational behaviour to be exhibited. To comply with this requirement, this work proposes a rational RGB-D dataset with two types of sequences, some for training and some for testing. Preliminary results show the addition of knowledge modeling and reasoning leads to a significant increase of recognition accuracy when compared to a system based only on computer vision.

Sistema de visão para aterragem automática de UAV

Dissertação para obtenção do grau de Mestre em Engenharia Electrotécnica Ramo de Automação e Electrónica Industrial

Probabilistic Egomotion for Stereo Visual Odometry

We present a novel approach of Stereo Visual Odometry for vehicles equipped with calibrated stereo cameras. We combine a dense probabilistic 5D egomotion estimation method with a sparse keypoint based stereo approach to provide high quality estimates of vehicle’s angular and linear velocities. To validate our approach, we perform two sets of experiments with a well known benchmarking dataset. First, we assess the quality of the raw velocity estimates in comparison to classical pose estimation algorithms. Second, we added to our method’s instantaneous velocity estimates a Kalman Filter and compare its performance with a well known open source stereo Visual Odometry library. The presented results compare favorably with state-of-the-art approaches, mainly in the estimation of the angular velocities, where significant improvements are achieved.

Video-based analysis of Gait pathologies

L’analyse de la marche a émergé comme l’un des domaines médicaux le plus im- portants récemment. Les systèmes à base de marqueurs sont les méthodes les plus fa- vorisées par l’évaluation du mouvement humain et l’analyse de la marche, cependant, ces systèmes nécessitent des équipements et de l’expertise spécifiques et sont lourds, coûteux et difficiles à utiliser. De nombreuses approches récentes basées sur la vision par ordinateur ont été développées pour réduire le coût des systèmes de capture de mou- vement tout en assurant un résultat de haute précision. Dans cette thèse, nous présentons notre nouveau système d’analyse de la démarche à faible coût, qui est composé de deux caméras vidéo monoculaire placées sur le côté gauche et droit d’un tapis roulant. Chaque modèle 2D de la moitié du squelette humain est reconstruit à partir de chaque vue sur la base de la segmentation dynamique de la couleur, l’analyse de la marche est alors effectuée sur ces deux modèles. La validation avec l’état de l’art basée sur la vision du système de capture de mouvement (en utilisant le Microsoft Kinect) et la réalité du ter- rain (avec des marqueurs) a été faite pour démontrer la robustesse et l’efficacité de notre système. L’erreur moyenne de l’estimation du modèle de squelette humain par rapport à la réalité du terrain entre notre méthode vs Kinect est très prometteur: les joints des angles de cuisses (6,29◦ contre 9,68◦), jambes (7,68◦ contre 11,47◦), pieds (6,14◦ contre 13,63◦), la longueur de la foulée (6.14cm rapport de 13.63cm) sont meilleurs et plus stables que ceux de la Kinect, alors que le système peut maintenir une précision assez proche de la Kinect pour les bras (7,29◦ contre 6,12◦), les bras inférieurs (8,33◦ contre 8,04◦), et le torse (8,69◦contre 6,47◦). Basé sur le modèle de squelette obtenu par chaque méthode, nous avons réalisé une étude de symétrie sur différentes articulations (coude, genou et cheville) en utilisant chaque méthode sur trois sujets différents pour voir quelle méthode permet de distinguer plus efficacement la caractéristique symétrie / asymétrie de la marche. Dans notre test, notre système a un angle de genou au maximum de 8,97◦ et 13,86◦ pour des promenades normale et asymétrique respectivement, tandis que la Kinect a donné 10,58◦et 11,94◦. Par rapport à la réalité de terrain, 7,64◦et 14,34◦, notre système a montré une plus grande précision et pouvoir discriminant entre les deux cas.

Statistical Object Recognition

Two formulations of model-based object recognition are described. MAP Model Matching evaluates joint hypotheses of match and pose, while Posterior Marginal Pose Estimation evaluates the pose only. Local search in pose space is carried out with the Expectation--Maximization (EM) algorithm. Recognition experiments are described where the EM algorithm is used to refine and evaluate pose hypotheses in 2D and 3D. Initial hypotheses for the 2D experiments were generated by a simple indexing method: Angle Pair Indexing. The Linear Combination of Views method of Ullman and Basri is employed as the projection model in the 3D experiments.