Uma contribuição ao estudo das categorias internas e de sua proliferação em redes ARTMAP

ART networks present some advantages: online learning; convergence in a few epochs of training; incremental learning, etc. Even though, some problems exist, such as: categories proliferation, sensitivity to the presentation order of training patterns, the choice of a good vigilance parameter, etc. Among the problems, the most important is the category proliferation that is probably the most critical. This problem makes the network create too many categories, consuming resources to store unnecessarily a large number of categories, impacting negatively or even making the processing time unfeasible, without contributing to the quality of the representation problem, i. e., in many cases, the excessive amount of categories generated by ART networks makes the quality of generation inferior to the one it could reach. Another factor that leads to the category proliferation of ART networks is the difficulty of approximating regions that have non-rectangular geometry, causing a generalization inferior to the one obtained by other methods of classification. From the observation of these problems, three methodologies were proposed, being two of them focused on using a most flexible geometry than the one used by traditional ART networks, which minimize the problem of categories proliferation. The third methodology minimizes the problem of the presentation order of training patterns. To validate these new approaches, many tests were performed, where these results demonstrate that these new methodologies can improve the quality of generalization for ART networks

Seleção de features guiada por atenção visual em imagens com fóvea

Visual attention is a very important task in autonomous robotics, but, because of its complexity, the processing time required is significant. We propose an architecture for feature selection using foveated images that is guided by visual attention tasks and that reduces the processing time required to perform these tasks. Our system can be applied in bottom-up or top-down visual attention. The foveated model determines which scales are to be used on the feature extraction algorithm. The system is able to discard features that are not extremely necessary for the tasks, thus, reducing the processing time. If the fovea is correctly placed, then it is possible to reduce the processing time without compromising the quality of the tasks outputs. The distance of the fovea from the object is also analyzed. If the visual system loses the tracking in top-down attention, basic strategies of fovea placement can be applied. Experiments have shown that it is possible to reduce up to 60% the processing time with this approach. To validate the method, we tested it with the feature algorithm known as Speeded Up Robust Features (SURF), one of the most efficient approaches for feature extraction. With the proposed architecture, we can accomplish real time requirements of robotics vision, mainly to be applied in autonomous robotics

Correspondência estéreo usando imagens em multiresolução com profundidade variável

We propose a multi-resolution, coarse-to-fine approach for stereo matching, where the first matching happens at a different depth for each pixel. The proposed technique has the potential of attenuating several problems faced by the constant depth algorithm, making it possible to reduce the number of errors or the number of comparations needed to get equivalent results. Several experiments were performed to demonstrate the method efficiency, including comparison with the traditional plain correlation technique, where the multi-resolution matching with variable depth, proposed here, generated better results with a smaller processing time

Implementação e análise de desempenho dos protocolos de criptografia neural e Diffie-Hellman em sistemas RFID utilizando uma plataforma embarcada

RFID (Radio Frequency Identification) identifies object by using the radio frequency which is a non-contact automatic identification technique. This technology has shown its powerful practical value and potential in the field of manufacturing, retailing, logistics and hospital automation. Unfortunately, the key problem that impacts the application of RFID system is the security of the information. Recently, researchers have demonstrated solutions to security threats in RFID technology. Among these solutions are several key management protocols. This master dissertations presents a performance evaluation of Neural Cryptography and Diffie-Hellman protocols in RFID systems. For this, we measure the processing time inherent in these protocols. The tests was developed on FPGA (Field-Programmable Gate Array) platform with Nios IIr embedded processor. The research methodology is based on the aggregation of knowledge to development of new RFID systems through a comparative analysis between these two protocols. The main contributions of this work are: performance evaluation of protocols (Diffie-Hellman encryption and Neural) on embedded platform and a survey on RFID security threats. According to the results the Diffie-Hellman key agreement protocol is more suitable for RFID systems

Técnicas visuais de localização e mapeamento simultâneos sem extração de primitivas geométricas da imagem

In Simultaneous Localization and Mapping (SLAM - Simultaneous Localization and Mapping), a robot placed in an unknown location in any environment must be able to create a perspective of this environment (a map) and is situated in the same simultaneously, using only information captured by the robot s sensors and control signals known. Recently, driven by the advance of computing power, work in this area have proposed to use video camera as a sensor and it came so Visual SLAM. This has several approaches and the vast majority of them work basically extracting features of the environment, calculating the necessary correspondence and through these estimate the required parameters. This work presented a monocular visual SLAM system that uses direct image registration to calculate the image reprojection error and optimization methods that minimize this error and thus obtain the parameters for the robot pose and map of the environment directly from the pixels of the images. Thus the steps of extracting and matching features are not needed, enabling our system works well in environments where traditional approaches have difficulty. Moreover, when addressing the problem of SLAM as proposed in this work we avoid a very common problem in traditional approaches, known as error propagation. Worrying about the high computational cost of this approach have been tested several types of optimization methods in order to find a good balance between good estimates and processing time. The results presented in this work show the success of this system in different environments

Localização de um robô móvel usando odometria e marcos naturais

Several methods of mobile robot navigation request the mensuration of robot position and orientation in its workspace. In the wheeled mobile robot case, techniques based on odometry allow to determine the robot localization by the integration of incremental displacements of its wheels. However, this technique is subject to errors that accumulate with the distance traveled by the robot, making unfeasible its exclusive use. Other methods are based on the detection of natural or artificial landmarks present in the environment and whose location is known. This technique doesnt generate cumulative errors, but it can request a larger processing time than the methods based on odometry. Thus, many methods make use of both techniques, in such a way that the odometry errors are periodically corrected through mensurations obtained from landmarks. Accordding to this approach, this work proposes a hybrid localization system for wheeled mobile robots in indoor environments based on odometry and natural landmarks. The landmarks are straight lines de.ned by the junctions in environments floor, forming a bi-dimensional grid. The landmark detection from digital images is perfomed through the Hough transform. Heuristics are associated with that transform to allow its application in real time. To reduce the search time of landmarks, we propose to map odometry errors in an area of the captured image that possesses high probability of containing the sought mark

Estudo da desidratação de tomates em pedaços com pré-tratamento osmótico

Vegetables drying plays an important role in the field of food dehydration, being a very old practice that was originated from sun drying items of food in order to preserve them to be consumed during the periods of scarcity. One of these vegetables is the tomato, that was originally grown in South America. Tomatoes are easily perishable after being picked up from the tree and this makes the process of tomato dehydration a challenge due to the high amount of water (95%) contained in them. The present research work was mainly intended to develop alternative processes for tomato conservation, by drying slices of skinned and unskinned tomatoes in the in natura form or in the osmotically pre-dehydrated form. Firstly, the best conditions of the osmotic pre-dehydration process were defined including temperature, immersion time and concentration of the osmotic solution, based on the results of water loss, solids gain and weigh reduction of the pre-dehydration tomatoes at different processing conditions. The osmotic solution used was made up of NaCl (5 and 10%) and sucrose (25 and 35%) at different combinations. For a fixed conditions of osmotic pre-dehydration, the drying tests of the pre-processed and in natura tomatoes were carried out in a stove with air circulation and a convective dryer with trays, at two levels of temperature. The sensorial analysis of the osmotically pre-treated and unskinned dehydrated tomatoes was carried out as well as a study on the their shelf-live. The results obtained showed that the drying of the tomatoes took place as a result of the internal control of the water transport, and did not show a constant rate, while two distinct periods of the decreasing phase were observed. The osmotic pre treatment substancially reduced the initial amount of humidity in the tomatoes, thus reducing the necessary time for the product to attain levels of intermediate humidity. The impermeability of the tomato skin was identified as well as the unfavorable influence of the pre-treatment on the unskinned tomatoes, whose solid gain brought about a decrease in the water activity with subsequent reduction of the drying rate. Despite the various simplifications carried out during the development of this study, the proposed diffusive model adjusted to the experimental data satisfactorily, thus making it possible to determine the effective coefficients of diffusion, whose results were consistent and compatible with those found in the current literature. Concerning the higher rates of evaporation and the lowest processing time, the best results were obtained in the drying of the unskinned, in natura tomatoes and of the skinned, pre-dehydrated tomatoes, at 60ºC, both processed in the convective drier. The results of the sensorial analysis of the unskinned and pre-treated product did not prove to be satisfactory. Regarding the shelf-live of the tomatoes, for a period of 45 days, no physicochemical or microbiological alteration of the product was noted

Secagem de grãos de girassol em leito fixo e em leito de jorro

This work studies two methods for drying sunflower grains grown in the western region of Rio Grande do Norte, in the premises of the Instituto Federal de Educação, Ciência e Tecnologia do Rio Grande do Norte - IFRN - Campus Apodi. This initiative was made because of the harvested grain during the harvest, being stored in sheds without any control of temperature, humidity etc. Therewith, many physical, chemical and physiological characteristics are compromised and grains lose much quality for oil production as their germination power. Taking into account that most of the stored grain is used for replanting, the studied methods include drying of grains in a thin layer using an oven with air circulation (fixed bed) and drying in a spouted bed. It was studied the drying of grains in natura, i.e., newly harvested. The fixed bed drying was carried out at temperatures of 40, 50, 60 and 70°C. Experiments in spouted bed were performed based on an experimental design, 2² + 3, with three replications at the central point, where the independent variables were grains load (1500, 2000 and 2500 g) and the temperature of the inlet air (70, 80, and 90 °C), obtaining the drying and desorption equilibrium isotherms. Previously, the characteristic curves of the bed were obtained. Both in the fixed bed as in the spouted bed, drying and desorption curves were obtained by weighing the grains throughout the experiments and measurements of water activity, respectively. The grains drying in the spouted bed showed good results with significant reduction of processing time. The models of FICK and PAGE were fitted to the experimental data, models which will represent the drying of grains both in the fixed bed as in the spouted bed. The desorption curves showed no influence of the processing temperature in the hygroscopic characteristics of the grains. The models of GAB, OSWIN and LUIKOV could well represent the desorption isotherms

Electronic structure and optical properties of BaMoO4 powders

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Recovery of silver residues from dental amalgam

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Electric load forecasting using a fuzzy ART&ARTMAP neural network

This work presents a neural network based on the ART architecture ( adaptive resonance theory), named fuzzy ART& ARTMAP neural network, applied to the electric load-forecasting problem. The neural networks based on the ARTarchitecture have two fundamental characteristics that are extremely important for the network performance ( stability and plasticity), which allow the implementation of continuous training. The fuzzy ART& ARTMAP neural network aims to reduce the imprecision of the forecasting results by a mechanism that separate the analog and binary data, processing them separately. Therefore, this represents a reduction on the processing time and improved quality of the results, when compared to the Back-Propagation neural network, and better to the classical forecasting techniques (ARIMA of Box and Jenkins methods). Finished the training, the fuzzy ART& ARTMAP neural network is capable to forecast electrical loads 24 h in advance. To validate the methodology, data from a Brazilian electric company is used. (C) 2004 Elsevier B.V. All rights reserved.

Detection and classification of voltage disturbances using a Fuzzy-ARTMAP-wavelet network

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Fault section estimation in electric power systems using an optimization immune algorithm

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Sistemas distribuídos aplicados à compressão e recuperação de imagens

Digital image processing is a field that demands great processing capacity. As such it becomes relevant to implement software that is based on the distribution of the processing into several nodes divided by computers belonging to the same network. Specifically discussed in this work are distributed algorithms of compression and expansion of images using the discrete cosine transform. The results show that the savings in processing time obtained due to the parallel algorithms in comparison to its sequential equivalents is a function that depends on the resolution of the image and the complexity of the involved calculation; that is efficiency is greater the longer the processing period is in terms of the time involved for the communication between the network points.

Alternative proposal for Modal Representation of a Non-transposed Three-Phase Transmission Line with a Vertical Symmetry Plane

The objective of this paper is to show an alternative representation in time domain of a non-transposed three-phase transmission line decomposed in its exact modes by using two transformation matrices. The first matrix is Clarke's matrix that is real, frequency independent, easily represented in computational transient programs (EMTP) and separates the line into quasi-modes a, b and zero. After that, Quasi-modes a and zero are decomposed into their exact modes by using a modal transformation matrix whose elements can be synthesized in time domain through standard curve-fitting techniques. The main advantage of this alternative representation is to reduce the processing time because a frequency dependent modal transformation matrix of a three-phase line has nine elements to be represented in time domain while a modal transformation matrix of a two-phase line has only four elements. This paper shows modal decomposition process and eigenvectors of a non-transposed three-phase line with a vertical symmetry plane whose nominal voltage is 440 kV and line length is 500 km.