Rede neural não supervisionada baseada em transformada Karhunen-Loève para processamento, compressão e reconhecimento de imagens

The merit of the Karhunen-Loève transform is well known. Since its basis is the eigenvector set of the covariance matrix, a statistical, not functional, representation of the variance in pattern ensembles is generated. By using the Karhunen-Loève transform coefficients as a natural feature representation of a character image, the eigenvector set can be regarded as an feature extractor for a classifier.

Detecção de pistas de aeroporto em imagens orbitais utilizando morfologia matemática

The Brazilian Cartography presents great deficiency in cartographic products updating. This form, Remote Sensins techniques together Digital Processing Images - DPI, are contributing to improve this problem. The Mathematical Morphology theory was used in this work. The principal function was the pruning operator. With its were extracted the interest features that can be used in cartographic process updating. The obtained results are positives and showed the use potential of mathematical morphology theory in cartography, mainly in updating.

Indicadores de floresta urbana a partir de imagens aéreas multiespectrais de alta resolução

The present work evaluated urban forest indicators, acquired through airborne high-resolution multiespectral images, on the quality of the urban design and its vegetative fraction, in special its trees, in nine neighborhoods of Piracicaba, SP. There were made supervised classifications for characterization of intra-urban elements and the proportions obtained, as exposed soil, tree cover, lawns, asphalt, concrete pavements and roofs. They were studied for the measurement of the urban forest in each place. These variables were related to each other, as well as with the independent variables: population density, people with more than fifteen years of study and family heads with income above twenty minimum wages, obtained through population census. Through the analysis of linear regression variables were identified for intra-urban areas evaluation. Correlations were made and linear regressions among the data obtained from the image and among the proposed indicators. Negative correlations were obtained among population density and arboreal covering and the evaluated indices, in accordance with the predicted in the literature. Composite indicators are proposed, as: the proportion between arboreous space on waterproof space (PAW) and the proportion between arboreous space on building space (PAB). It is concluded by the possibility of the use of those indicators for evaluation of the urban forest and definition of priorities in the execution of ordinances to the improvement of the urban forestry, being prioritized the application of resources in the most lacking neighborhoods.

Extração de feições rodoviárias em imagens digitais: Metodologias desenvolvidas pelo GP-F&VC

Several kinds of research in road extraction have been carried out in the last 6 years by the Photogrammetry and Computer Vision Research Group (GPF&VC - Grupo de Pesquisa em Fotogrametria e Visão Computacional). Several semi-automatic road extraction methodologies have been developed, including sequential and optimizatin techniques. The GP-F&VC has also been developing fully automatic methodologies for road extraction. This paper presents an overview of the GP-F&VC research in road extraction from digital images, along with examples of results obtained by the developed methodologies.

Extração semi-automática de rodovias em um par de imagens aéreas digitais utilizando geometria epipolar e programação dinâmica

The purpose of this paper is to introduce a methodology for semi-automatic road extraction from aerial digital image pairs by using dynamic programming and epipolar geometry. The method uses both images from where each road feature pair is extracted. The operator identifies the corresponding road featuresand s/he selects sparse seed points along them. After all road pairs have been extracted, epipolar geometry is applied to determine the automatic point-to-point correspondence between each correspondent feature. Finally, each correspondent road pair is georeferenced by photogrammetric intersection. Experiments were made with rural aerial images. The results led to the conclusion that the methodology is robust and efficient, even in the presence of shadows of trees and buildings or other irregularities.

Orientação semi-automática de imagens CBERS usando rodovias como controle de campo

In this paper is proposed a methodology for semiautomatic CBERS image orientation using roads as ground control. It is based on an iterative strategy involving three steps. In the first step, an operator identifies on the image the ground control roads and supplies along them a few seed points, which could be sparsely and coarsely distributed. These seed points are used by the dynamic programming algorithm for extracting the ground control roads from the image. In the second step, it is established the correspondences between points describing the ground control roads and the corresponding ones extracted from the image. In the last step, the corresponding points are used to orient the CBERS image by using the DLT (Direct Linear Transformation). The two last steps are iterated until the convergence of the orientation process is verified. Experimental results showed that the proposed methodology was efficient with several test images. In all cases the orientation process converged. Moreover, the estimated orientation parameters allowed the registration of check roads with pixel accuracy or better.

Um modelo para orientação de imagens orbitais baseado em feições retas

The aim of this paper is to present a model for orientation of pushbroom sensors that allows estimating the polynomial coefficients describing the trajectory of the platform, using linear features as ground control. Considering that pushbroom image acquisition is not instantaneous, six EOP (Exterior Orientation Parameters) for each scanned line must be estimated. The sensor position and attitude parameters are modeled with a time dependent polynomial. The relationship between object and image space is established through a mathematical model based on the equivalence between the vector normal to the projection plane in the image space and to the vector normal to the rotated projection plane in the object space. The equivalence property between planes was adapted to consider the pushbroom geometry. Some experiments with simulated data corresponding to CBERS scene (China-Brazil Earth Resource Satellite) were accomplished in order to test the developed model using straight lines. Moreover, experiments with points ground with the model based on collinearity equations adapted to the pushbroom geometry were also accomplished. The obtained results showed that the proposed model can be used to estimate the EOP of pushbroom images with suitable accuracy.

Detecção de bordas em imagens digitais através de um processo de difusão anisotrópica não linear

The edges detection model by a non-linear anisotropic diffusion, consists in a mathematical model of smoothing based in Partial Differential Equation (PDE), alternative to the conventional low-pass filters. The smoothing model consists in a selective process, where homogeneous areas of the image are smoothed intensely in agreement with the temporal evolution applied to the model. The level of smoothing is related with the amount of undesired information contained in the image, i.e., the model is directly related with the optimal level of smoothing, eliminating the undesired information and keeping selectively the interest features for Cartography area. The model is primordial for cartographic applications, its function is to realize the image preprocessing without losing edges and other important details on the image, mainly airports tracks and paved roads. Experiments carried out with digital images showed that the methodology allows to obtain the features, e.g. airports tracks, with efficiency.

Proposição de rotina morfológica para detecção de malha viária em imagens orbitais

The outdating of cartographic products affects planning. It is important to propose methods to help detect changes in surface. Thus, the combined use of remote sensing image and techniques of digital image processing has contributed significantly to minimize such outdating. Mathematical morphology is an image processing technique which describes quantitatively geometric structures presented in the image and provides tools such as edge detectors and morphological filters. Previous studies have shown that the technique has potential on the detection of significant features. Thus, this paper proposes a routine of morphological operators to detect a road network. The test area corresponds to an excerpt Quickbird image and has as a feature of interest an avenue of the city of Presidente Prudente, SP. In the processing, the main morphological operators used were threshad, areaopen, binary and erosion. To estimate the accuracy with which the linear features were detected, it was done the analysis of linear correlation between vectors of the features detected and the corresponding topographical map of the region. The results showed that the mathematical morphology can be used in cartography, aiming to use them in conventional cartographic updating processes.

Técnicas de correlação no delineamento de feições rodoviárias em imagens de baixa e média resolução

This paper presents a method for the sequential road feature delineation from digital images. It is based on a feedback loop between extrapolation and refinement steps of a given road centerline point, using in both steps correlation techniques. Firstly, a previously extracted road centerline point is linearly extrapolated, resulting in an approximate position. Secondly, this approximate position is corrected by comparing gray level profiles extracted perpendicularly to the extrapolation direction. This strategy is then repeated to allow the entire road centerline to be extracted or a stop point to be found. In order to initialize the extraction process, the operator needs to supply a starting point plus direction and width. Experimental results obtained from the application of the method to real image data are presented and discussed in this paper.

Analisando imagens: Um ensaio sobre a criação de fontes narrativas para compreender os Grupos Escolares

This paper is an essay on how photos can be analysed and used to create narratives which may serve as resources for historical studies about school practices. As an exercise, we deal with six old photographs taken in Grupos Escolares, a Brazilian educational institution founded in the last decade of the 19th Century and extinguished in the 1970's. According to some authors, these schools represented the beginning of the public educational system in Brazil.

Classificação de áreas de favelas a partir de imagens IKONOS: Viabilidade de uso de uma abordagem orientada a objeto

The growth of large cities is usually accelerated and disorganized, which causes social, economical and infrastructural conflicts and frequently, occupation in illegal areas. For a better administration of these areas, the public manager needs information about their location. This information can be obtained through land utilization and land cover maps, where orbital images of remote sensing are used as one of the most traditional sources of data. In this context, the present work tested the applicability of the object-based classification to categorize two slum areas, taking into account the structure of the streets, size of the huts, distance between the houses, among other parameters. These area combinations of physical aspects were analyzed using the image IKONOS II and the software eCognition. Slum areas tend to be, to the contrary of the planned areas, disarranged, with narrow streets, small houses built with a variety of materials and without definition of blocks. The results of land cover classification for slum areas are encouraging because they are accurate and little ambiguous in the classification process. Thus, it would allow its utilization by urban managers.

Um filtro gaussiano adaptativo para análise comparativa de imagens de radares e satélites meteorológicos

Meteorological satellite and radar data comparative analysis allows to correlate the precipitation structures observed in both images. Such analysis would make feasible the extension of the range of ground-based meteorological radars. In addition to the different spatial and temporal resolution of these images this comparative analysis presents difficulties due to the effects of rotation and distortion, besides the different formats, projections, and coordinate systems. This work employed an approach based on a Gaussian adaptive filter in order to compare such images. The statistical results obtained from the comparison of the images are matched to those produced by other methods.

Extração de vias combinando métodos de detecção de regiões e linhas em imagens de intensidade de pulso laser

This paper aims at extracting street centerlines from previously isolated street regions by using the image of laser scanning intensity. In this image, streets are easily identified, since they manifest as dark, elongate ribbons contrasting with background objects. The intensity image is segmented by using the region growing technique, which generates regions representing the streets. From these regions, the street centerlines are extracted in two manners. The first one is through the Steger lines detection method combined with a line length thresholding by which lines being shorter than a minimum length are removed. The other manner is by combining the skeletonization method of regions based on the Medial Axis Transform and with a pruning process to eliminate as much as possible the ramifications. Experiments showed that the Steger-based method provided results better than the method based on skeletonization.