Emiliano Ferreira Castejon

Using neural networks and cellular automata for modelling intra-urban land-use dynamics

Empirical models designed to simulate and predict urban land‐use change in real situations are generally based on the utilization of statistical techniques to compute the land‐use change probabilities. In contrast to these methods, artificial neural networks arise as an alternative to assess such probabilities by means of non‐parametric approaches. This work introduces a simulation experiment on intra‐urban land‐use change in which a supervised back‐propagation neural network has been employed in the parameterization of several biophysical and infrastructure variables considered in the simulation model. The spatial land‐use transition probabilities estimated thereof feed a cellular automaton (CA) simulation model, based on stochastic transition rules. The model has been tested in a medium‐sized town in the Midwest of São Paulo State, Piracicaba. A series of simulation outputs for the case study town in the period 1985–1999 were generated, and statistical validation tests were then conducted for the best results, based on fuzzy similarity measures.

Image Fusion for Remote Sensing Applications

Remote Sensing systems, particularly those deployed on satellites, provide a repetitive and consistent view of the Earth (Schowengerdt, 2007). To meet the needs of different remote sensing applications the systems offer a wide range of spatial, spectral, radiometric and temporal resolutions. Satellites usually take several images from frequency bands in the visual and non-visual range. Each monochrome image is referred to as a band and a collection of several bands of the same scene acquired by a sensor is called multispectral image (MS). A combination of three bands associated in a RGB (Red, Green, Blue) color system produce a color image. The color information in a remote sensing image by using spectral band combinations for a given spatial resolution increases information content which is used in many remote sensing applications. Otherwise, different targets in a single band may appear similar which makes difficult to distinguish them. Different bands can be acquired by a single multispectral sensor or by multiple sensors operating at different frequencies. Complementary information about the same scene can be available in the following cases (Simone et al., 2002):  Data recorded by different sensors;  Data recorded by the same sensor operating in different spectral bands;  Data recorded by the same sensor at different polarization;  Data recorded by the same sensor located on platforms flying at different heights. In general, sensors with high spectral resolution, characterized by capturing the radiance from different land covers in a large number of bands of the electromagnetic spectrum, do not have an optimal spatial resolution, that may be inadequate to a specific identification task despite of its good spectral resolution (Gonzalez-Audicana, 2004). On a high spatial resolution panchromatic image (PAN), detailed geometric features can easily be recognized, while the multispectral images contain richer spectral information. The capabilities of the images can be enhanced if the advantages of both high spatial and spectral resolution can be integrated into one single image. The detailed features of such an integrated image thus can be easily recognized and will benefit many applications, such as urban and environmental studies (Shi et al., 2005).

Digital Image Processing in Remote Sensing

Imaging systems, particularly those on board satellites, provide a repetitive and consistent view of the earth that has been used in many remote sensing applications such as urban growth, deforestation and crop monitoring, weather prediction, land use mapping, land cover mapping and so on. For each application it is necessary to develop a specific methodology to extract information from the image data. To develop a methodology it is necessary to identify a procedure based on image processing techniques that is more adequate to the problem solution. In spite of the application complexity, some basic techniques are common in most of the remote sensing applications named as image registration, image fusion, image segmentation and classification. Hence, this paper aims to present an overview about the use of image processing techniques to solve a general problem on remote sensing application. A case study on an urban application is provided to illustrate the use of remote sensing technologies for solving the problem.

The Divide and Segment Method for Parallel Image Segmentation

Remote sensing images with large spatial dimensions are usual. Besides, they also include a diversity of spectral channels, increasing the volume of information. To obtain valuable information from remote sensing data, computers need higher amounts of memory and more efficient processing techniques. The first process in image analysis is segmentation, which identifies regions in images. Therefore, segmentation algorithms must deal with large amounts of data. Even with current computational power, certain image sizes may exceed the memory limits, which ask for different solutions. An alternative to overcome such limits is to employ the well-known divide and conquer strategy, by splitting the image into chunks, and segmenting each one individually. However, it arises the problem of merging neighboring chunks and keeping the homogeneity in such regions. In this work, we propose an alternative to divide the image into chunks by defining noncrisp borders between them. The noncrisp borders are computed based on Dijkstra algorithm, which is employed to find the shortest path between detected edges in the images. By applying our method, we avoid the postprocessing of neighboring regions, and therefore speed up the final segmentation.

Improvements in Sample Selection Methods for Image Classification

Traditional image classification algorithms are mainly divided into unsupervised and supervised paradigms. In the first paradigm, algorithms are designed to automatically estimate the classes’ distributions in the feature space. The second paradigm depends on the knowledge of a domain expert to identify representative examples from the image to be used for estimating the classification model. Recent improvements in human-computer interaction (HCI) enable the construction of more intuitive graphic user interfaces (GUIs) to help users obtain desired results. In remote sensing image classification, GUIs still need advancements. In this work, we describe our efforts to develop an improved GUI for selecting the representative samples needed to estimate the classification model. The idea is to identify changes in the common strategies for sample selection to create a user-driven sample selection, which focuses on different views of each sample, and to help domain experts identify explicit classification rules, which is a well-established technique in geographic object-based image analysis (GEOBIA). We also propose the use of the well-known nearest neighbor algorithm to identify similar samples and accelerate the classification.

MELHORIA DO PROCESSO DE CORREÇÃO GEOMÉTRICA DE IMAGENS CBERS-CCD PELO USO DE AMOSTRAS GEORREFERENCIADAS CLASSIFICADAS

As imagens da serie de satelites CBERS sao distribuidas gratuitamente, mas para que seja possivel utiliza-las, e necessario aplicar um metodo de correcao geometrica. E proposta uma melhoria do processo de correcao automatica de forma a selecionar as melhores amostras de referencia a partir das quais e possivel definir pontos de controle usados para o calculo dos parâmetros do modelo usado na correcao. Para demonstrar a eficacia, o metodo proposto e aplicado em um conjunto de imagens CBERS usando as amostras de imagens selecionadas.