Juan Humberto Sossa Azuela

A comparison between two robust techniques for segmentation of blood vessels

Image segmentation plays an important role in image analysis. According to several authors, segmentation terminates when the observers goal is satisfied. For this reason, a unique method that can be applied to all possible cases does not yet exist. In this paper, we have carried out a comparison between two current segmentation techniques, namely the mean shift method, for which we propose a new algorithm, and the so-called spectral method. In this investigation the important information to be extracted from an image is the number of blood vessels (BV) present in the image. The results obtained by both strategies were compared with the results provided by manual segmentation. We have found that using the mean shift segmentation an error less than 20% for false positives (FP) and 0% for false negatives (FN) was observed, while for the spectral method more than 45% for FP and 0% for FN were obtained. We discuss the advantages and disadvantages of both methods.

A Procedure to Select the Vigilance Threshold for the ART2 for Supervised and Unsupervised Training

The Adaptive Resonance Theory ART2 [1] is used as a non supervised tool to generate clusters. The clusters generated by an ART2 Neural Network (ART2 NN), depend on a vigilance threshold (ρ). If ρ is near to zero, then a lot of clusters will be generated; if ρ is greater then more clusters will be generated. To get a good performance, this ρ has to be suitable selected for each problem. Until now, no technique had been proposed to automatically select a proper ρ for a specific problem. In this paper we present a first way to automatically obtain the value of ρ, we also illustrate how it can be used in supervised and unsupervised learning. The goal to select a suitable threshold is to reach a better performance at the moment of classification. To improve classification, we also propose to use a set of feature vectors instead of only one to describe the objects. We present some results in the case of character recognition.

Improving Pattern Recognition Using Several Feature Vectors

Most pattern recognition systems use only one feature vector to describe the objects to be recognized. In this paper we suggest to use more than one feature vector to improve the classification results. The use of several feature vectors require a special neural network, a supervised ART2 NN is used [1]. The performance of a supervised or unsupervised ART2 NN depends on the appropriate selection of the vigilance threshold. If the value is near to zero, a lot of clusters will be generated, but if it is greater, then must clusters will be generated. A methodology to select this threshold was first proposed in [2]. The advantages to use several feature vectors instead of only one are shown on this work. We show some results in the case of character recognition using one and two feature vectors. We also compare the performance of our proposal with the multilayer perceptron.

Feature Matching Using Accumulation Spaces

A new way to solve the matching problem between model and image features is described in this paper. Matches between features accumulate in a region of an abstract space; a space similar to the Hough space. In such a space, found clusters determine possible 2D rotations and scale changes of the object in the image. Finally the relative position between model and image features is verified in each cluster. The use of a space of accumulation drastically reduces the complexity of matching. The proposed approach has been tested with several images with very promising results.

Impulse rejecting filter for efficient noise removal and fine detail preservation

A novel filtering algorithm applicable to image processing is presented. It was designed using rank-ordered mean (ROM) estimator to remove an outlier and robust local data activity estimators to detect the outliers. The proposed filter effectively remove impulse noise and preserve edge and fine details. The filter possesses good visual quality of the processed simulated images and good quantitative quality in comparison to the standard median filter. Recommendations to obtain best processing results by proper selection of the filter parameters are given. The designed filter is suitable for impulse noise removal in any image processing applications. One can use it at the first stage of image enhancement followed by any detail-preserving techniques such as the Sigma filter at the second stage.