Cosimo Distante

Shadow detection for moving objects based on texture analysis

This paper presents a new approach for shadow detection of moving objects in visual surveillance environment, improving localization, segmentation, tracking and classification of detected objects. An automatic segmentation procedure based on adaptive background difference is performed in order to detect potential shadow points so that, for all moving pixels, the approach evaluates the compatibility of photometric properties with shadow characteristics. The shadow detection approach is improved by evaluating the similarity between little textured patches, since shadow regions present same textural characteristics in each frame and in the corresponding adaptive background model. In this work we suggest a new approach to describe textural information in terms of redundant systems of functions. The algorithm is designed to be unaffected by scene type, background type or light conditions. Experimental results validate the algorithms performance on a benchmark suite of indoor and outdoor video sequences.

Support vector machines for olfactory signals recognition

Pattern recognition techniques have widely been used in the context of odor recognition. The recognition of mixtures and simple odors as separate clusters is an untractable problem with some of the classical supervised methods. Recently, a new paradigm has been introduced in which the detection problem can be seen as a learning from examples problem. In this paper, we investigate odor recognition in this new perspective and in particular by using a novel learning scheme known as support vector machines (SVM) which guarantees high generalization ability on the test set. We illustrate the basics of the theory of SVM and show its performance in comparison with radial basis network and the error backpropagation training method. The leave-one-out procedure has been used for all classifiers, in order to finding the near-optimal SVM parameter and both to reduce the generalization error and to avoid outliers.

On the study of feature extraction methods for an electronic nose

In this study, we analyzed the transient of microsensors based on tin oxide sol–gel thin film. A novel method to this research field for data analysis and discrimination among different volatile organic compounds is presented. Moreover; several feature extraction methods have been considered, both steady-state (fractional change, relative, difference and log) and transient (Fourier and wavelet descriptors, integral and derivatives) information. Feature extraction methods have been validated qualitatively (by using principal component analysis) and quantitatively on the classification rate (by using a radial basis function neural network). # 2002 Elsevier Science B.V. All rights reserved.

Quantitative and qualitative analysis of VOCs mixtures by means of a microsensors array and different evaluation methods

Abstract In this work we show the capability of a sol–gel based electronic nose to be used in qualitative and quantitative analysis with the aim to recognize common volatile compounds usually present in the headspace of foods. Acetone, hexanal and 2-pentanone were chosen for this kind of measurements, performed both in dry air and in mixture of 50% humidity, just to simulate the experimental set-up in real applications. Moreover, different mixtures of 2-pentanone and hexanal at low concentration were investigated. We show how linear technique, such as principal component analysis (PCA) algorithm can be used for inspecting data distribution in simple cases like cluster discrimination. Moreover, non-linear techniques for difficult regression problems, such as artificial neural networks (multi-layer perceptron (MLP) and radial basis function (RBF)) are needed in particular for the prediction of different concentrations.

A shadow elimination approach in video-surveillance context

Moving objects tracking is an important problem in many applications such as video-surveillance. Monitoring systems can be improved using vision-based techniques able to extract and classify objects in the scene. However, problems arise due to unexpected shadows because shadow detection is critical for accurate objects detection in video stream, since shadow points are often misclassified as object points causing errors in localization, segmentation, measurements, tracking and classification of moving objects. The paper presents a new approach for removing shadows from moving objects, starting from a frame-difference method using a grey-level textured adaptive background. The shadow detection scheme uses photometric properties and the notion of shadow as semi-transparent region which retains a reduced-contrast representation of the underlying surface pattern and texture. We analyze the problem of representing texture information in terms of redundant systems of functions for texture identification. The method for discriminating shadows from moving objects is based on a Pursuit scheme using an over-complete dictionary. The basic idea is to use the simple but powerful Matching Pursuit algorithm (MP) for representing texture as linear combination of elements of a big set of functions. Particularly, MP selects the best little set of atoms of 2D Gabor dictionary for features selection representative of properties of the texture in the image. Experimental results validate the algorithms performance.

Randomized circle detection with isophotes curvature analysis

Circle detection is a critical issue in image analysis and object detection. Although Hough transform based solvers are largely used, randomized approaches, based on the iterative sampling of the edge pixels, are object of research in order to provide solutions less computationally expensive. This work presents a randomized iterative work-flow, which exploits geometrical properties of isophotes in the image to select the most meaningful edge pixels and to classify them in subsets of equal isophote curvature. The analysis of candidate circles is then performed with a kernel density estimation based voting strategy, followed by a refinement algorithm based on linear error compensation. The method has been applied to a set of real images on which it has also been compared with two leading state of the art approaches and Hough transform based solutions. The achieved results show how, discarding up to 57% of unnecessary edge pixels, it is able to accurately detect circles within a limited number of iterations, maintaining a sub-pixel accuracy even in the presence of high level of noise. HighlightsAn iterative randomized circle detection algorithm is proposed.Curvature of isophotes is used to reduce the number of necessary iterations.Edges on circumference(s) are selected by a Kernel density estimation strategy.Sub-pixel accuracy was maintained on real images even in presence of high noise levels.Isophotes analysis reduces the influence of the used edge map on the results.

Human posture recognition using active contours and radial basis function neural network

In this paper an automated video surveillance system for human posture recognition using active contours and neural networks is presented. Localization of moving objects in the scene and human posture estimation are key features of the proposed architecture. The system architecture consists of five sequential modules that include the moving target detection process, two levels of segmentation process for interested element localization, features extraction of the object shape and a human posture classification system based on the radial basis functions neural network. Moving objects are detected by using an adaptive background subtraction method with an automatic background adaptation speed parameter and a new fast gradient vector flow snake algorithm for the elements segmentation is proposed. The developed system has been tested for the classification of three different postures such as standing, bending and squatting considering different kinds of feature. Results are promising and the architecture is also useful for the discrimination of human activities.

An Investigation on the Feasibility of Uncalibrated and Unconstrained Gaze Tracking for Human Assistive Applications by Using Head Pose Estimation

This paper investigates the possibility of accurately detecting and tracking human gaze by using an unconstrained and noninvasive approach based on the head pose information extracted by an RGB-D device. The main advantages of the proposed solution are that it can operate in a totally unconstrained environment, it does not require any initial calibration and it can work in real-time. These features make it suitable for being used to assist human in everyday life (e.g., remote device control) or in specific actions (e.g., rehabilitation), and in general in all those applications where it is not possible to ask for user cooperation (e.g., when users with neurological impairments are involved). To evaluate gaze estimation accuracy, the proposed approach has been largely tested and results are then compared with the leading methods in the state of the art, which, in general, make use of strong constraints on the people movements, invasive/additional hardware and supervised pattern recognition modules. Experimental tests demonstrated that, in most cases, the errors in gaze estimation are comparable to the state of the art methods, although it works without additional constraints, calibration and supervised learning.

Facial expression recognition and histograms of oriented gradients: a comprehensive study

Automatic facial expression recognition (FER) is a topic of growing interest mainly due to the rapid spread of assistive technology applications, as human–robot interaction, where a robust emotional awareness is a key point to best accomplish the assistive task. This paper proposes a comprehensive study on the application of histogram of oriented gradients (HOG) descriptor in the FER problem, highlighting as this powerful technique could be effectively exploited for this purpose. In particular, this paper highlights that a proper set of the HOG parameters can make this descriptor one of the most suitable to characterize facial expression peculiarities. A large experimental session, that can be divided into three different phases, was carried out exploiting a consolidated algorithmic pipeline. The first experimental phase was aimed at proving the suitability of the HOG descriptor to characterize facial expression traits and, to do this, a successful comparison with most commonly used FER frameworks was carried out. In the second experimental phase, different publicly available facial datasets were used to test the system on images acquired in different conditions (e.g. image resolution, lighting conditions, etc.). As a final phase, a test on continuous data streams was carried out on-line in order to validate the system in real-world operating conditions that simulated a real-time human–machine interaction.

Dynamic Cluster Recognition with Multiple Self-Organising Maps

Abstract: A neural architecture, based on several self-organising maps, is presented which counteracts the parameter drift problem for an array of conducting polymer gas sensors when used for odour sensing. The neural architecture is named mSom, where m is the number of odours to be recognised, and is mainly constituted of m maps; each one approximates the statistical distribution of a given odour. Competition occurs both within each map and between maps for the selection of the minimum map distance in the Euclidean space. The network (mSom) is able to adapt itself to new changes of the input probability distribution by repetitive self-training processes based on its experience. This architecture has been tested and compared with other neural architectures, such as RBF and Fuzzy ARTMAP. The network shows long-term stable behaviour, and is completely autonomous during the testing phase, where re-adaptation of the neurons is needed due to the changes of the input probability distribution of the given data set.