Electronic Letters on Computer Vision and Image Analysis (ELCVIA - Universitat Autònoma de Barcelona)
Not a member yet
343 research outputs found
Sort by
Fast 3D-Vision System to Classify Metallic Coins by their Embossed Topography
Get PDFThis paper presents a security-related machine-vision solution for real-time classification of moving objects with highly reflective metallic surfaces and complex 3D-structures. As an application example of our so called Three-Color Selective Stereo Gradient Method (Three-Color SSGM) a classification system for three main coin denominations of Euro coins is presented. Such coins are quickly moving in a coin validation system. The objective is to decide only from comparison of specially measured and processed 3D-surface information with characteristic topographical data stored in a database whether a coin belongs to one of the reference classes or has to be rejected as a foreign or counterfeit coin. Under illumination from a three-color light emitting diode equipped ring a single image of the moving coin is captured by a digital color camera. Exploiting the spectral properties of the illumination sources, which correspond to the special spectral characteristics of the camera, three independent subimages can be extracted. Comparison between these subimages leads to a discrimination between a coin with real 3D-surface and a counterfeit coin based on a photographic image of a coin of the same type. After the coin has been located and segmented, grey value based rotation and translation invariant features are extracted froma normalized image. In combination with template matching methods, a coin can be classified. Classification results will be reported for the three main coin denominations of Euro coins
An Adaptive Color Image Segmentation
Get PDFA novel Adaptive Color Image Segmentation (ACIS) System for color image segmentation is presented. The proposed ACIS system uses a neural network with architecture similar to the multilayer perceptron (MLP) network. The main difference is that neurons here uses a multisigmoid activation function. The multisigmoid function is the key for segmentation. The number of steps i.e. thresholds in the multisigmoid function are dependant on the number of clusters in the image. The threshold values for detecting the clusters and their labels are found automatically from the first order derivative of histograms of saturation and intensity in the HSV color space. Here, the main use of neural network is to detect the number of objects automatically from an image. The advantage of this method is that no a priori knowledge is required to segment the color image. ACIS label the objects with their mean colors. The algorithm is found to be reliable and works satisfactorily on different kinds of color images. Experimental results show that the performance of ACIS is robust on noisy images also
Hand Tracking and Gesture Recognition for Human-Computer Interaction
Get PDFThe proposed work is part of a project that aims for the control of a videogame based on hand gesture recognition. This goal implies the restriction of real-time response and unconstrained environments. In this paper we present a real-time algorithm to track and recognise hand gestures for interacting with the videogame. This algorithm is based on three main steps: hand segmentation, hand tracking and gesture recognition from hand features. For the hand segmentation step we use the colour cue due to the characteristic colour values of human skin, its invariant properties and its computational simplicity. To prevent errors from hand segmentation we add a second step, hand tracking. Tracking is performed assuming a constant velocity model and using a pixel labeling approach. From the tracking process we extract several hand features that are fed to a finite state classifier which identifies the hand configuration. The hand can be classified into one of the four gesture classes or one of the four different movement directions. Finally, using the system’s performance evaluation results we show the usability of the algorithm in a video game environment
An Experimental Investigation about the Integration of Facial Dynamics in Video-Based Face Recognition
Get PDFRecent psychological and neural studies indicate that when people talk their changing facial expressions and head movements provide a dynamic cue for recognition. Therefore, both fixed facial features and dynamic personal characteristics are used in the human visual system (HVS) to recognize faces. However, most automatic recognition systems use only the static information as it is unclear how the dynamic cue can be integrated and exploited. The few works attempting to combine facial structure and its dynamics do not consider the relative importance of these two cues. They rather combine the two cues in an adhoc manner. But what is the relative importance of these two cues separately? Does combining them enhance systematically the recognition performance? To date, no work has extensively studied these issues. In this article, we investigate these issues by analyzing the effects of incorporating the dynamic information in video-based automatic face recognition. We consider two factors (face sequence length and image quality) and study their effects on the performance of video-based systems that attempt to use a spatio-temporal representation instead of one based on a still image. We experiment with two different databases and consider HMM (the temporal hidden Markov model) and ARMA (the auto-regressive and moving average model) as baseline methods for the spatio-temporal representation and PCA and LDA for the image-based one. The extensive experimental results show that motion information enhances also automatic recognition but not in a systematic way as in the HVS
Multiorder polygonal approximation of digital curves
Get PDFIn this paper, we propose a quick threshold-free algorithm, which computes the angular shape of a 2D object from the points of its contour. For that, we have extended the method defined in [4, 5] to a multiorder analysis. It is based on the arithmetical definition of discrete lines [11] with variable thickness. We provide a framework to analyse a digital curve at different levels of thickness. The extremities of a segment provided at a high resolution are tracked at lower resolution in order to refine their location. The method is thresholdfree and automatically provides a partitioning of a digital curve into its meaningful parts
Handwritten Document Analysis for Automatic Writer Recognition
Get PDFIn this paper, we show that both the writer identification and the writer verification tasks can be carried out using local features such as graphemes extracted from the segmentation of cursive handwriting. We thus enlarge the scope of the possible use of these two tasks which have been, up to now, mainly evaluated on script handwritings. A textual based Information Retrieval model is used for the writer identification stage. This allows the use of a particular feature space based on feature frequencies. Image queries are handwritten documents projected in this feature space. The approach achieves 95% correct identification on the PSI_DataBase and 86% on the IAM_DataBase. Then writer hypothesis retrieved are analysed during a verification phase. We call upon a mutual information criterion to verify that two documents may have been produced by the same writer or not. Hypothesis testing is used for this purpose. The proposed method is first scaled on the PSI_DataBase then evaluated on the IAM_DataBase. On both databases, similar performance of nearly 96% correct verification is reported, thus making the approach general and very promising for large scale applications in the domain of handwritten document querying and writer verification
Simultaneous and Causal Appearance Learning and Tracking
Get PDFA novel way to learn and track simultaneously the appearance of a previously non-seen face without intrusive techniques can be found in this article. The presented approach has a causal behaviour: no future frames are needed to process the current ones. The model used in the tracking process is refined with each input frame thanks to a new algorithm for the simultaneous and incremental computation of the singular value decomposition (SVD) and the mean of the data. Previously developed methods about iterative computation of SVD are taken into account and an original way to extract the mean information from the reduced SVD of a matrix is also considered. Furthermore, the results are produced with linear computational cost and sublinear memory requirements with respect to the size of the data. Finally, experimental results are included, showing the tracking performance and some comparisons between the batch and our incremental computation of the SVD with mean information
Bayesian Network Enhanced Prediction for Multiple Facial Feature Tracking
Get PDFIt is challenging to track multiple facial features simultaneously in video while rich facial expressions are presented in a human face. To accurately predict the positions of multiple facial features\u27 contours is important and difficult. This paper proposes a multi-cue prediction model based tracking algorithm. In the prediction model, CAMSHIFT is used to track the face in video in advance, and facial features\u27 spatial constraint is utilized to roughly obtain the positions of facial features. Second order autoregressive process (ARP) based dynamic model is combined with graphical model (Bayesian network) based dynamic model.Incorporating ARP\u27s quickness into graphical model\u27s accurateness, we obtain the fusion of the prediction. Finally the prediction model and the measurement model are integrated into the framework of Kalman filter. The experimental results show that our algorithm can accurately track multiple facial features with varied facial expressions
Self-supervised adaptation for on-line script text recognition
Get PDFWe have recently developed in our lab a text recognizer for on-line texts written on a touch-terminal. We present in this paper several strategies to adapt this recognizer in a self-supervised way to a given writer and compare them to the supervised adaptation scheme. The baseline system is based on the activation-verification cognitive model. We have designed this recognizer to be writer-independent but it may be adapted to be writer-dependent in order to increase the recognition speed and rate. The classification expert can be iteratively modified in order to learn the particularities of a writer. The best self-supervised adaptation strategy is called prototype dynamic management and gets good results, close to those of the supervised methods. The combination of supervised and self-supervised strategies increases accuracy again. Results, presented on a large database of 90 texts (5,400 words) written by 38 different writers are very encouraging with an error rate lower than 10~\%
Prior Knowledge Based Motion Model Representation
Get PDFThis paper presents a new approach for human walking modeling from monocular image sequences. A kinematics model and a walking motion model are introduced in order to exploit prior knowledge. The proposed technique consists of two steps. Initially, an efficient feature point selection and tracking approach is used to compute feature points\u27 trajectories. Peaks and valleys of these trajectories are used to detect key frames-frames where both legs are in contact with the floor. Secondly, motion models associated with each joint are locally tuned by using those key frames. Differently than previous approaches, this tuning process is not performed at every frame, reducing CPU time. In addition, the movement\u27s frequency is defined by the elapsed time between two consecutive key frames, which allows handling walking displacement at different speed. Experimental results with different video sequences are presented