Recently, maximum spatial eigenfiltering allowed significant improvements on the estimation performance of likelihood direction-of-arrival (DOA) estimators for closely spaced sources. However, that eigenfilter may greatly attenuate widely sources as SNR decreases, leading to severe degradation. Since differential spectrum shows prominent spectral peaks around true DOA angles even at very low SNR, we originally propose using it derive two eigenvalue-based finite-impulse response filters...

Face recognition methods are computationally very expensive and use too much memory processing time. An example of a method that allocates many computer resources is the Principal Component Analysis (PCA). In order to reduce time, was developed in this paper using only genetic algorithms perform face comparison PCA obtains higher accurate rates less

This work deals with object tracking, on live video, through the boundary detection by Hough transform and color maps transform. It is presented an experiment where trace of human iris tested variations texture contour measurements are stored discussed. Furthermore, it done a breakdown techniques used methodology applied.

The Compressive Sensing (CS) allows the acquisition of signals already compressed and posterior reconstruction with much less number measures than minimum required by Nyquist theorem. A subarea CS which improves performance at stage is named Model-Based CS. Some works have been developed within this subarea. However, most them consider only noise generated sparse approximation, disregarding quantization its influence on efficiency robustness objective study to investigate in for images...

This paper describes a fast thinning algorithm (FTA). possesses two main steps, that are repeated until the obtained image approaches medium axis of original image. In first step contour is marked and in second analysed to verify which pixels-on belonging this should be deleted. The FTA maintains connectivity preserves end points final skeleton. It practically immune noise and, extreme importance, has shorter execution time. Several tests carried out results very satisfactory.

The development of Intelligent Tutoring Systems has been studied for decades and requires multi interdisciplinary knowledge. In this context, article presents a hybrid intelligent tutoring system in which the teaching strategies are based on neural networks, Self-Organizing Maps (SOM), knowledge specialist teacher. This model reactive adaptive characteristic that provides personalized individual instruction student, by promoting necessary guidance transfer didactic contents. paper process...

The field orientation and the binarization in an image are often used fingerprint identification authentication, analysis of textures. proposed algorithm reuses additions multiplications calculation using switching property uses Digital Differential Analyzer (DDA) generation convolution masks for images. performance processing time result compared to algorithms that use convolutions were satisfactory other found literature.

Computer systems are used to support breast cancer diagnosis, with decisions taken from measurements carried out in regions of interest (ROIs). We show that obtained square or rectangular ROIs can include background different behavior healthy diseased tissues. In this study, the were identified as Partial Pixels (PP), a multilevel method segmentation based on maximum entropy. The behaviors healthy, and partial tissues quantified by fractal dimension multiscale lacunarity, calculated through...

This work presents a multi-dimensional similarity modeling strategy and relevance feedback technique for minimizing the semantic gap intrinsic problem of CBIR systems by allowing users to customize their queries according requirements preferences. We propose composite using multi-dimensional, vectorial, spatially clustered, relevance-ordered approach. Given set k features which represents images in an image database, measure between query another from collection is analyzed components, are...

This work aims to evaluate how does the performance of iris localization algorithms is influenced by severe compression eye images used in an automatic recognition system. The correct region essential guarantee credibility biometric recognition. As it first processing stage system, that can have its efficiency depredated quality loss be processed. Fractal and JPEG2000 were applied a public database. Two traditional methods which localize circular borders method for detection eyelids...

The visualization of image structures in 3D obtained from computerized tomography examinations aids the medical professional analysis images and consequently, provides a more accurate diagnoses. As these (slices) are spaced apart it becomes necessary to fill empty spaces show structure 3D. use virtual slice between real slices following restoration is new approach realizing interpolation aimed at visualization. goal this article develop method which produces with few regions and, through an...

O tema da inclusao digital tem crescido em importância para a social na medida que relacao entre tecnologia e sociedade torna-se cada dia mais complexa. Esta pesquisa se propoe estudar o avanco das novas tecnologias informacao comunicacao (NTICs) persistente abismo existente no Brasil. Verifica-se resposta apresentada pelo Estado brasileiro as mudancas tecnologicas, com analise politica publica de digital, tendo como base acoes/programas/projetos desenvolvidos, implantados disponibilizados...