Electronic Letters on Computer Vision and Image Analysis (ELCVIA - Universitat Autònoma de Barcelona)
Not a member yet
343 research outputs found
Sort by
A Comparison Framework for Walking Performances using aSpaces
Get PDFIn this paper, we address the analysis of human actions by comparing different performances of the same action executed by different actors. Specifically, we present a comparison procedure applied to the walking action, but the scheme can be applied to other different actions, such as bending, running, etc. To achieve fair comparison results, we define a novel human body model based on joint angles, which maximizes the differences between human postures and, moreover, reflects the anatomical structure of human beings. Subsequently, a human action space, called aSpace, is built in order to represent each performance (i.e., each predefined sequence of postures) as a parametric manifold. The final human action representation is called p-action, which is based on the most characteristic human body postures found during several walking performances. These postures are found automatically by means of a predefined distance function, and they are called key-frames. By using key-frames, we synchronize any performance with respect to the p-action. Furthermore, by considering an arc length parameterization, independence from the speed at which performances are played is attained. As a result, the style of human walking can be successfully analysed by establishing the differences of the joints between female and male walkers
Detecting human heads with their orientations
Get PDFWe propose a two-step method for detecting human heads with their orientations. In the first step, the method employs an ellipse as the contour model of human-head appearances to deal with wide variety of appearances. Our method then evaluates the ellipse to detect possible human heads. In the second step, on the other hand, our method focuses on features inside the ellipse, such as eyes, the mouth or cheeks, to model facial components. The method evaluates not only such components themselves but also their geometric configuration to eliminate false positives in the first step and, at the same time, to estimate face orientations. Our intensive experiments show that our method can correctly and stably detect human heads with their orientations
Learning Model Structure from Data: An Application to On-Line Handwriting
Get PDFWe present a learning strategy for Hidden Markov Models that may be used to cluster handwriting sequences or to learn a character model by identifying its main writing styles. Our approach aims at learning both the structure and parameters of a Hidden Markov Model (HMM) from the data. A byproduct of this learning strategy is the ability to cluster signals and identify allograph. We provide experimental results on artificial data that demonstrate the possibility to learn from data HMM parameters and topology. For a given topology, our approach outperforms in some cases that we identify standard Maximum Likelihood learning scheme. We also apply our unsupervised learning scheme on on-line handwritten signals for allograph clustering as well as for learning HMM models for handwritten digit recognition
SKCS-A Separable Kernel Family with Compact Support to improve visual segmentation of handwritten data
Get PDFExtraction of pertinent data from noisy gray level document images with various and complex backgrounds such as mail envelopes, bank checks, business forms, etc... remains a challenging problem in character recognition applications. It depends on the quality of the character segmentation process. Over the last few decades, mathematical tools have been developed for this purpose. Several authors show that the Gaussian kernel is unique and offers many beneficial properties. In their recent work Remaki and Cheriet proposed a new kernel family with compact supports (KCS) in scale space that achieved good performance in extracting data information with regard to the Gaussian kernel. In this paper, we focus in further improving the KCS efficiency by proposing a new separable version of kernel family namely (SKCS). This new kernel has also a compact support and preserves the most important properties of the Gaussian kernel in order to perform image segmentation efficiently and to make the recognizer task particularly easier. A practical comparison is established between results obtained by using the KCS and the SKCS operators. Our comparison is based on the information loss and the gain in time processing. Experiments, on real life data, for extracting handwritten data, from noisy gray level images, show promising performance of the SKCS kernel, especially in reducing drastically the processing time with regard to the KCS
A Fast Fractal Image Compression Method Based Entropy
Get PDFFractal image compression gives some desirable properties like resolution independence, fast decoding, and very competitive rate-distortion curves. But still suffers from a (sometimes very) high encoding time, depending on the approach being used. This paper presents a method to reduce the encoding time of this technique by reducing the size of the domain pool based on the Entropy value of each domain block. Experimental results on standard images show that the proposed method yields superior performance over conventional fractal encoding
Texture Segmentation using LBP embedded Region Competition
Get PDFIn this paper, we modify the region competition method to segment textures. First, local Binary pattern (LBP) histogram is adopted to capture the texture information. Then, considering the specific goal of texture segmentation, we propose new assumption about region competition and rewrite the energy function based on LBP histograms. We also develop the two-stage iterative algorithm to make our energy converge to a local minimum. Because of the fast LBP operator and nonparametric histogram model, we can simplify the step of parameter estimating, which is always the most time-consuming. Besides, LBP` s high performance for texture characterization helps to make our method more suitable for texture segmentation problem. Experiments show that the performance of our proposed method is promising, and a robust and fast segmentation of texture images is obtained
Improvement of Modal Matching Image Objects in Dynamic Pedobarography using Optimization Techniques
Get PDFThis paper presents an approach for matching objects in dynamic pedobarography image sequences, based on finite element modeling and modal analysis. The determination of correspondences between nodal objects is improved using optimization techniques and, because the number of elements of each object is not necessary the same, a new algorithm to match excess nodes is proposed. This new matching algorithm uses a neighborhood criterion and can overcome some disadvantages of the usual ";one to one"; matching. The proposed approach allows the determination of correspondences between 2D or 3D objects. Here, a special application to dynamic pedobarography images is presented
Robustness of a Blind Image Watermark Detector Designed by Orthogonal Projection
Get PDFBlind digital watermarking, which can detect watermark without using the original image, is a key technique practical intellectual property protecting systems and concealment correspondence systems. In this paper, we discussed a blind detection method for the digital image watermark. The theories research show that the orthogonal projection sequence of a digital image is one-to-one correspondence with this digital image. To make use of this conclusion, we designed and realized a kind of blind watermark detector with the good performance. To calculate the correlation value between the image and watermark, the intensity information of digital image is not adopted, but the orthogonal projection sequence of this image is adopted. Experiment results show that this watermark detector not only to have very strong resistant ability to translation and rotation attacks, but also to have the good robustness to Gaussian noise. Performance of this watermark detector is better than general detector designed by making use of the intensity information directly. The conclusions obtained by experiments are useful to the research in the future
Presentation of ELCVIA
No full textThis paper wants to be a presentation of the e-journal Electronic Letters on Computer Vision and Image Analysis (ELCVIA). ELCVIA is a new electronic journal whose aim is to provide a forum where Computer Vision and Image Analysis researchers can discuss about methods, applications and experiences in this field. ELCVIA tries to offer to authors electronical means (images, video or exe files) to show in a better form their results, which are more difficult to demonstrate in paper format. ELCVIA tries to make the usual journal process (article submission, referee\u27s revision and revisions reception) through WWW, by means of Web forms in the ELCVIA Web site http://www.cvc.uab.es/elcvia. ELCVIA offers two types of contributions: DRAFT, with an open discussion of all the people, and ARTICLE, with a close revision of Editorial Board and reviewers. This paper also contains information to the authors about the article or draft style format to be send to ELCVIA
Optimization of Weights in a Multiple Classifier Handwritten Word Recognition System Using a Genetic Algorithm
Get PDFAutomatic handwritten text recognition by computer has a number of interesting applications. However, due to a great variety of individual writing styles, the problem is very difficult and far from being solved. Recently, a number of classifier creation methods, known as ensemble methods, have been proposed in the field of machine learning. They have shown improved recognition performance over single classifiers. For the combination of these classifiers many methods have been proposed in the literature. In this paper we describe a weighted voting scheme where the weights are obtained by a genetic algorithm