Guilherme Lucio Abelha Mota

Face detector combining eigenfaces, neural network and bootstrap

A critical issue in an automatic face recognition system is the determination of the region containing a face in an image with a cluttered background. The paper presents a method that optimizes the detection task through the use of eigenfaces, neural networks and a bootstrap algorithm. The main component of the proposed method is a nonlinear operator that detects the presence of a well-framed face image in 20x20 pixel windows. To detect faces larger than the window size the input image is successively reduced by a factor of 1.2 and the procedure is applied at each scale. To obtain a compact representation of the face images, the method applies principal component analysis directly to the pixel intensities of face images. Each image window analyzed by the detection algorithm is then projected upon the n principal components, the so-called eigenfaces. The dimensionality reduction thus achieved implies in a reconstruction error, the DFFS-distance from features space. The patterns representing an image window are formed by the n projections plus the DFFS.

Dealing with double vagueness in DEM morphometric analysis

ABSTRACT A method is presented to explicitly incorporate spatial and scale vagueness – double vagueness – into geomorphometric analyses. Known limitations of usual practices include using a single fixed set of crisp thresholds for morphometric classification and the imposition of a single arbitrary number of scales of analysis to the entire digital elevation model (DEM). Among the advantages of the proposed method are: fuzzification of morphometric classification rules, scale-dependent adaptive fuzzy set parametrization and an objective definition of maximum scale of analysis on a cell-by-cell basis. The method was applied to several DEMs ranging from the ocean floor to surface landscapes of both Earth and Mars. The result was evaluated with respect to modal morphometric features and to characteristic scales, suggesting a more robust method for deriving both morphometric classifications and terrain attributes. We argue that the method would be preferable to any single-scale crisp approach, at least in the context of preliminary hands-off morphometric analyses of DEMs.

Automatic Selection of Training Samples for Multitemporal Image Classification

The present work presents and evaluates a method to automatically select training samples of medium resolution satellite images within a supervised object oriented classification procedure. The method first takes a pair of images of the same area acquired in different dates and segments them in homogeneous regions on both images. Then a change detection algorithm takes stable segments as training samples. In experiments using Landsat images of an area in Southwest Brazil taken at three consecutive years the performance of the proposed method was close to the performance associated to the manual selection of training samples.

Query by image similarity using a fuzzy logic approach

In this paper we propose a new model for query by image similarity. The model utilizes a fuzzy logic approach to cluster intrinsic image characteristics, which are extracted from subregions of the image. The clustering process provides a set of parameters that are used to compare a target image with a group of images. As a result, the system provides the images in the data set which are similar to the target image. We present as an example some queries by similarity on an image database composed of 20 types of animals. The main objective of this model is to develop an intelligent image query system that can be applied on the web and image databases.

The E-Foto Project and the Research to Implement a GNU/GPL Open source Educational Digital Photogrammetric Workstation

Photogrammetry is the science devoted to the geometric reconstruction of object space, based on photographic images. Among its techniques, stereophotogrammetry, which employs a pair of images to precisely estimate 3D coordinates in the object space, should be highlighted. The state of the art of photogrammetric equipment is represented by digital photogrammetric workstations (DPW), that are primarily hardware and software dedicated to photogrammetric purposes. The present chapter is devoted to the E-Foto project, which, to the best of our knowledge and belief, implements the only open source GNU/GPL digital photogrammetric workstation presently available. This project is originally concerned on education and is underlain on the synergy between development, teaching and research devoted to this field of science. It encompasses a team of undergraduate and Master students as well as Professors of The Rio de Janeiro State University. Besides the DPW, the E-Foto project products are: the only digital photogrammetry book edited in Brazil, the project’s home page, the software usage tutorials, master dissertations, graduation projects and research papers.1

Estimating Class Dynamics for Fuzzy Markov Chain-Based Multitemporal Cascade Classification

The key component of a fuzzy Markov chain (FMC)-based multitemporal cascade classifier is the transition possibility matrix (TPM). Such matrix represents the temporal dynamics of the land use/land cover classes in the target site in a given time period. The choice of the TPM estimation approach is a crucial step in the design of FMC-based classifiers, as it strongly influences the final classification accuracy. Moreover, the task of collecting training data may involve considerable effort, since the number of transitions to be represented grows with the square of the number of classes in the application. In spite of their relevance, the TPM estimation has only been addressed superficially in previous publications about FCM-based classification methods. In this letter, we concern some of those aspects and investigate alternative ways of the TPM estimation. Experimental analysis on a multitemporal data set covering a 20-year period sheds light on the conditions under which those alternative estimation approaches may be used, as well as on their impact over the classification performance.