Tania Mezzadri Centeno

General Type-2 Fuzzy Inference Systems: Analysis, Design and Computational Aspects

The aim of this work is to handle non-interval type-2 fuzzy logic systems (NIT2 FLS) in a simple manner. We retrieve an alternative representation of Type-2 fuzzy sets (T2 FS) that we call general footprint of uncertainty. Such representation, not only lets us easily visualize T2 FS in two-dimensions but also makes the understanding of basic operations and the inference procedure easier. We introduce the concept of supremum of a T2 FS and translation and cylindric extension for vertical slices to support the adopted inference mechanism, based on the scaled inference mechanism of Type-1 FIS. Finally, we propose a new defuzzification method, the vertical slice centroid type reduction, which requires low computational effort. Some calculations are presented to illustrate that the theory and simplifications proposed in this paper make NIT2 FLS, referred here as general type-2 fuzzy inference systems, much more accessible to FIS designers.

Land Cover Classification Based on General Type-2 Fuzzy Classifiers

This paper proposes a fuzzy classifier based on type-2 fuzzy sets to be applied in land cover classification. The classifier is built on the basis of the available data and considers the merging of information drawn from different experts. The data regard a thematic mapper representing the land cover of a real plain cultivated area. The experts are represented by different bands which classify the spectral sensor information. The new proposed method to design the classifier as well as the use of general type-2 fuzzy sets allows the modeling of input-output relations and minimizes the effects of uncertainties in the usual fuzzy rule-based classifiers. The experiments carried out attest to the efficiency of the proposed general type-2 fuzzy classifier.

An object detection and recognition system for weld bead extraction from digital radiographs

With base in object detection and recognition techniques, we developed and implemented a new methodology to perform the first head-function of a weld quality interpretation system: the weld bead extraction from a digital radiograph. The proposed methodology uses a genetic algorithm to manage the search for suitable parameters values (position, width, length, and angle) that best defines a window, in the radiographic image, matching with the model image of a weld bead sample. The search results are verified in a classification process that recognize true detections using image matching parameters also proposed in this work. To test the proposed methodology, two groups of images were used; one consisting of 110 radiographs from pipelines welded joints and the other containing 6 images with different numbers of radiographs per image. The tests results showed that, besides automatically check the number of weld beads per image, the proposed methodology is also able to supply the respective position, width, length, and angle of each weld bead, with an accurate rate of 94.4%. As a result, the detected weld beads are correctly extracted from the original image and made available to be inspected through others algorithms for failure detection and classification.

General type-2 fuzzy classifiers to land cover classification

This paper proposes a fuzzy classifier based on type-2 fuzzy sets to be applied in land cover classification. The classifier is built from the available data and considers the merging of information acquired from different experts. The data regards a thematic mapper representing the land cover of a real plain cultivated area. The experts are represented by different bands which discretize the spectral sensor information. The new method proposed to design the classifier as well as the use of general type-2 fuzzy sets allows the modeling of input-output relations and minimize the effects of uncertainties in the usual fuzzy rule-based classifiers. The experiments carried out attest the efficiency of the proposed general type-2 fuzzy classifier.

Particle Swarm Optimization for Object Recognition in Computer Vision

Particle Swarm Optimization (PSO) is an evolutionary computation technique frequently used for optimization tasks. This work aims at applying PSO for recognizing specific patterns in complex images. Experiments were done with gray level and color images, with and without noise. PSO was able to find predefined reference images, submitted to translation, rotation, scaling, occlusion, noise and change in the viewpoint in the landscape image. Several experiments were done to evaluate the performance of PSO. Results show that the proposed method is robust and very promising for real-world applications.

A soft computing-based approach to spatio-temporal prediction

This paper aims to incorporate intelligent mechanisms based on Soft Computing in Geographical Information Systems (GIS). The proposal here is to present a spatio-temporal prediction method of forestry evolution for a sequence of binary images by means of fuzzy inference systems (FIS), genetic algorithm (GA) and genetic programming (GP). The main inference is based on a fuzzy system which processes a set of crisp/fuzzy relations and infers a crisp relation representing the predicted image at a predefined date. The fuzzy system is formed by a fixed fuzzy rule base and a partition set that may be defined by an expert or optimized by means of a GA. Genetic programming may also be adopted to generate the size of predicted area used in the final stage of the inference process. The developed methodology is applied in regions of Venezuela, France and Guatemala to identify their forestry evolution trends. The proposed approaches are compared with other techniques to validate the system.

Object detection for computer vision using a robust genetic algorithm

This work is concerned with the development and implementation of an image pattern recognition approach to support computational vision systems when it is necessary to automatically check the presence of specific objects on a scene, and, besides, to describe their position, orientation and scale. The developed methodology involves the use of a genetic algorithm to find known 2D object views in the image. The proposed approach is fast and presented a robust performance in several test instances including multiobject scenes, with or without partial occlusion.

Genetic algorithms to automatic weld bead detection in double wall double image digital radiographs

This paper presents a pioneering approach for weld bead detection in radiographic images obtained by the Double Wall Double Image (DWDI) technique. Such task constitutes an essential step for several high level processes, such as fully automatic flaw identification on welded joints. Sets of sample pixels, corresponding to candidate solutions provided by a genetic algorithm (GA), are compared to pre-defined synthetic weld bead and pipe models in an image matching procedure. The fitness of each set (individual) is evaluated based on a linear combination of its genotype (evaluated by a heuristic function) and phenotype. The evolutionary process automatically selects the best individual in the population and, thus, provides information such as position, orientation and dimension of the detected object. The proposed approach successfully detects pipes and weld beads in radiographic images of different complexities, encouraging future works.

Spatio-Temporal Prediction by Means of a Fuzzy Rule-based Approach

Geographical information is a powerful tool for decision-making in Territorial-Physical Planning. By its turn, fuzzy systems seem to be an interesting technique to deal with imprecision and uncertainty in the large amount of heterogeneous data available in geographical information systems (GIS). This paper aims to incorporate intelligent mechanisms in GIS by means of fuzzy rule-based systems. The proposal here is to present a spatio-temporal prediction method of forestry evolution for a sequence of satellite images through the use of fuzzy inference systems.

Dimensional reduction of large image datasets using non-linear principal components

In this paper we apply a Neural Network (NN) to reduce image dataset, distilling the massive datasets down to a new space of smaller dimension. Due to the possibility of these data have nonlinearities, traditional multivariate analysis, like the Principal Component Analysis (PCA), may not represent reality. Alternatively, Nonlinear Principal Component Analysis (NLPCA) can be performed by a NN model to fulfill that deficiency. However, when the dimension of the image increases, NN may easily saturate. This work presents an original methodology associated with the use of a set of cascaded multi-layer NN with a bottleneck structure to extract nonlinear information of the large set of image data. We illustrate its good performance with a set of tests against comparisons using this methodology and PCA in the treatment of oceanographic data associated with mesoscale variability of an oceanic boundary current.