Roberto Rosas-Romero

Wavelet-based smoke detection in outdoor video sequences

In this paper an approach to detect smoke columns from outdoor forest video sequences is proposed. The approach follows three basic steps. The first step is an image pre-processing block which resizes the image by applying a bicubic interpolation algorithm. The image is then transformed to its intensity values with a gray-scale transformation and finally the image is grouped by common areas with an image indexation. The second step consists of a smoke detection algorithm which performs a stationary wavelet transform (SWT) to remove high frequencies on horizontal, vertical, and diagonal details. The inverse SWT is then implemented and finally the image is compared to a non-smoke scene in order to determine the possible regions of interest (ROI). In order to reduce the number of false alarms, the final step of the proposed approach consists on a smoke verification algorithm, which determines whether the ROI is increasing its area or not. These results are combined to reach a final decision for detecting a smoke column on a sequence of static images from an outdoor video. Experimental results show that multi-resolution wavelet analysis is more accurate than the traditional low-pass filters on this application.

A method to assist in the diagnosis of early diabetic retinopathy: Image processing applied to detection of microaneurysms in fundus images.

Diabetes increases the risk of developing any deterioration in the blood vessels that supply the retina, an ailment known as Diabetic Retinopathy (DR). Since this disease is asymptomatic, it can only be diagnosed by an ophthalmologist. However, the growth of the number of ophthalmologists is lower than the growth of the population with diabetes so that preventive and early diagnosis is difficult due to the lack of opportunity in terms of time and cost. Preliminary, affordable and accessible ophthalmological diagnosis will give the opportunity to perform routine preventive examinations, indicating the need to consult an ophthalmologist during a stage of non proliferation. During this stage, there is a lesion on the retina known as microaneurysm (MA), which is one of the first clinically observable lesions that indicate the disease. In recent years, different image processing algorithms, which allow the detection of the DR, have been developed; however, the issue is still open since acceptable levels of sensitivity and specificity have not yet been reached, preventing its use as a pre-diagnostic tool. Consequently, this work proposes a new approach for MA detection based on (1) reduction of non-uniform illumination; (2) normalization of image grayscale content to improve dependence of images from different contexts; (3) application of the bottom-hat transform to leave reddish regions intact while suppressing bright objects; (4) binarization of the image of interest with the result that objects corresponding to MAs, blood vessels, and other reddish objects (Regions of Interest-ROIs) are completely separated from the background; (5) application of the hit-or-miss Transformation on the binary image to remove blood vessels from the ROIs; (6) two features are extracted from a candidate to distinguish real MAs from FPs, where one feature discriminates round shaped candidates (MAs) from elongated shaped ones (vessels) through application of Principal Component Analysis (PCA); (7) the second feature is a count of the number of times that the radon transform of the candidate ROI, evaluated at the set of discrete angle values {0°, 1°, 2°, …, 180°}, is characterized by a valley between two peaks. The proposed approach is tested on the public databases DiaretDB1 and Retinopathy Online Challenge (ROC) competition. The proposed MA detection method achieves sensitivity, specificity and precision of 92.32%, 93.87% and 95.93% for the diaretDB1 database and 88.06%, 97.47% and 92.19% for the ROC database. Theory, results, challenges and performance related to the proposed MA detecting method are presented.

Segmentation of endocardium in ultrasound images based on sparse representation over learned redundant dictionaries

This paper considers the problem of segmenting the endocardium in 2-D short-axis echocardiographic images from rats by using the sparse representation of feature vectors over learned dictionaries during classification. We highlight important aspects of the application of the theory of sparse representation and dictionary learning to the problem of ultrasound image segmentation. Experiments were conducted following two directions for the generation of dictionaries for myocardium and blood pool regions; by manual extraction of image patches to build untrained dictionaries and by patch extraction followed by training of dictionaries. The results obtained from different learned dictionaries are compared. During classification of an image patch, instead of using features of the patch alone, features of neighboring patches are combined.

Forecasting of stock return prices with sparse representation of financial time series over redundant dictionaries

A new predictive model for time series within the financial sector is presented.The method is based on learned redundant dictionaries for sparse representation of financial time series.The overall return gain generated by the predictive model exceeds the gain generated by the market.Untrained dictionaries outperform dictionaries trained with the KSV-D method.Untrained dictionaries require a reduced number of atoms to achieve successful results. This paper presents the theory, methodology and application of a new predictive model for time series within the financial sector, specifically data from 20 companies listed on the U.S. stock exchange market. The main impact of this article is (1) the proposal of a recommender system for financial investment to increase the cumulative gain; (2) an artificial predictor that beats the market in most cases; and (3) the fact that, to the best of our knowledge, this is the first effort to predict time series by learning redundant dictionaries to sparsely reconstruct these signals. The methodology is conducted by finding the optimal set of predicting model atoms through two directions for dictionaries generation: the first one by extracting atoms from past daily return price values in order to build untrained dictionaries; and the second one, by atom extraction followed by training of dictionaries though K-SVD. Prediction of financial time series is a periodic process where each cycle consists of two stages: (1) training of the model to learn the dictionary that maximizes the probability of occurrence of an observation sequence of return values, (2) prediction of the return value for the next coming trading day. The motivation for such research is the fact that a tool, which might generate confidence of the potential benefits obtained from using formal financial services, would encourage more participation in a formal system such as the stock market. Theory, issues, challenges and results related to the application of sparse representation to the prediction of financial time series, as well as the performance of the method, are presented.

Remote detection of forest fires from video signals with classifiers based on K-SVD learned dictionaries

Abstract In this paper a method for remote detection of forest fires in video signals from surveillance cameras is presented. The idea is based on learned redundant dictionaries for sparse representation of feature vectors extracted from image patches on three different regions; smoke, sky and ground. A testing image patch is assigned to the region for which the corresponding dictionary gives the best sparse representation during segmentation. To further reduce the presence of misclassified patches, a spatio-temporal cuboid of patches is built around a classified patch to take a majority vote in the set of classes inside the cuboid. To reduce the number of false positives there is a verification process to determine if a region of interest is growing. Theory, results, issues and challenges related to the implementation of the forest fire monitoring system, and performance of the method are presented.

Detecting and Classifying Attacks in Computer Networks Using Feed-Forward and Elman Neural Networks

In this paper, we present an approach for detecting and classifying attacks in computer networks by using neural networks. Specifically, a design of an intruder detection system is presented to protect the hypertext transfer protocol (HTTP). We propose the use of an application-based model using neural networks to model properly non-linear data. The benefit of this perspective is to work directly on the causes of an attack, which are determined directly by the commands used in the protected application. The intruder detection system is designed by defining three different neural networks, which include two multi-layer feed-forward networks and the Elman recurrent network. The results reported in this paper show that the Elman recurrent network achieved a performance around ninety percent of good detection, which demonstrates the reliability of the designed system to detect and classify attacks in high-level network protocols.

A New Method for Designing Flat Shelving and Peaking Filters Based on Allpass Filters

Two most commonly IIR filters used in audio equalization are shelving filters and peaking filters. Traditional design of shelving and peaking filters is based on the design of analog filters, mainly Butterworth filters, and bilinear transformation. In this way, it is well known the design of first order shelving filters and second order peaking filters. Additionally, the resulting filter can be efficiently implemented using allpass filter structures with a low sensitivity to the filter quantization and a low noise level. In this paper, we present a direct design of high order shelving and peaking filters with flat magnitude response in both passband and stop-band. The design is reduced to the design of one digital allpass filter with real coefficients. Using this allpass filter, we obtain two stable and real allpass filters, which are used to implement the resulting shelving and peaking filters. Additionally, closed form equations for the pole/zero computations are given. In contrast with others proposed methods, the design parameters for the shelving filter are the gains KBdB and KcdB at omega=0 and omega=pi, respectively, the passband droop Ap, stopband attenuation As, passband frequency omegap, and stopband frequency omega s, while for the peaking filter we have the gains KBdB and KcdB, the passband and stopband attenuation Ap and As, passband width Wp, stopband width Ws, and the central frequency omega0. The proposed method is illustrated by means of examples. Finally, the appendix shows the MATLAB function ShelvingEq.m, which implements the proposed method for the design of shelving filters

Learning and approximation of chaotic time series using wavelet-networks

This paper presents a wavelet neural-network for learning and approximation of chaotic time series. Wavelet networks are a class of neural network that take advantage of good localization and approximation properties of multiresolution analysis. These networks use wavelets as activation functions in the hidden layer and a hierarchical method is used for learning. Comparisons are made between a wavelet network, tested with two different wavelets, and the typical feedforward network trained with the back-propagation algorithm. The results reported in this paper show that wavelet networks have better approximation properties than back-propagation networks.

Intrusion Detection and Classification of Attacks in High-Level Network Protocols Using Recurrent Neural Networks

This paper presents an application-based model for classifying and identifying attacks in a communications network and therefore guarantees its safety from HTTP protocol-based malicious commands. The proposed model is based on a recurrent neural network architecture and it is therefore suitable to work online and for analyzing non-linear patterns in real time to self-adjust to changes in its input environment. Three different neural network-based systems have been modelled and simulated for comparison purposes in terms of overall performance: a Feed-forward Neural Network, an Elman Network, and a Recurrent Neural Network. Simulation results show that the latter possesses a greater capacity than either of the others for the correct identification and classification of HTTP attacks, and it also reaches a result at a great speed, its somewhat taxing computing requirements notwithstanding.

Multi-modal medical image registration based on non-rigid transformations and feature point extraction by using wavelets

In order to correctly match two sets of images from different modalities, our method applies a non-rigid transformation to one set to get as close as possible to the other. This requires the estimation of the optimal similarity transformation between the two set of images. Estimation of the non-rigid deformation between the two sets of images is referred to as the deformation estimation between the pair of three-dimensional object extracted from both sets. We present a new methodology for image registration by first extracting objects from the set of images by reconstructing the object surfaces where this extraction supports semi-automatic segmentation of sets of 3-D medical images and then finding the best similarity transformation based on matching of two sets of surface points, but also incorporates the matching of two sets of feature points, and we have shown that deformation estimates based on simultaneous matching of surfaces and features are more accurate than those based on surface matching alone. This is especially true when the deformation involves physically realistic cases, such as those in human organs. Our technique uses free-form deformation models and applies the wavelet transform to extract feature points in the 3-D space. Feature point extraction also provides a means to compute the error in our estimates. We have applied our method to register sequences of MRI images to histology images of the carotid artery.