Publications by Roberto Vezzani

Computer vision and ubiquitous multimedia access nowadays make feasible the development of a mostly automated system for human-behavior analysis. In this context, our proposal is to analyze human behaviors by classifying the posture of the monitored person and, consequently, detecting corresponding events and alarm situations, like a fall. To this aim, our approach can be divided in two phases: for each frame, the projection histograms (Haritaoglu et al., 1998) of each person are computed and compared with the probabilistic projection maps stored for each posture during the training phase; then, the obtained posture is further validated exploiting the information extracted by a tracking module in order to take into account the reliability of the classification of the first phase. Moreover, the tracking algorithm is used to handle occlusions, making the system particularly robust even in indoors environments. Extensive experimental results demonstrate a promising average accuracy of more than 95% in correctly classifying human postures, even in the case of challenging conditions.

In this paper we address the problem of human posture classification, in particular focusing to an indoor surveillance application. The approach was initially inspired to a previous works of Haritaoglou et al. [5] that uses histogram projections to classify people’s posture. Projection histograms are here exploited as the main feature for the posture classification, but, differently from [5], we propose a supervised statistical learning phase to create probability maps adopted as posture templates. Moreover, camera calibration and homography are included to solve perspective problems and to improve the precision of the classification. Furthermore, we make use of a finite state machine to detect dangerous situations as falls and to activate a suitable alarm generator. The system works on-line on standard workstations with network cameras.

This work presents a novel people tracking approach, able to cope with frequent shape changes and large occlusions. In particular, the tracks are described by means of probabilistic masks and appearance models. Occlusions due to other tracks or due to background objects and false occlusions are discriminated. The tracking system is general enough to be applied with any motion segmentation module, it can track people interacting each other and it maintains the pixel assignment to track even with large occlusions. At the same time, the update model is very reactive, so as to cope with sudden body motion and silhouette's shape changes. Due to its robustness, it has been used in many experiments of people behavior control in indoor situations.

This paper describes our approach to real-time detection of camera motion and moving object segmentation in videos acquired from moving cameras. As far as we know, none of the proposals reported in the literature are able to meet real-time requirements. In this work, we present an approach based on a color segmentation followed by a region-merging on motion through Markov Random Fields (MRFs). The technique we propose is inspired to a work of Gelgon and Bouthemy (Pattern Recognition 33 (2000) 725-40), that has been modified to reduce computational cost in order to achieve a fast segmentation (about 10 frame per second). To this aim a modified region matching algorithm (namely Partitioned Region Matching) and an innovative arc-based MRF optimization algorithm with a suitable definition of the motion reliability are proposed. Results on both synthetic and real sequences are reported to confirm validity of our solution.

We describe an integrated solution devised for inhouse video surveillance, to control the safety of people living in a domestic environment. The system is composed of robust moving object detection module, able to disregard shadows, a tracking module designed for large occlusion solution and of a posture detector. Shadows, large occlusions and deformable model of people are key features of inhouse surveillance. Moreover, the requirements of high speed reaction to dangerous situations and the need to implement a reliable and low cost televiewing system, led to the introduction of a new multimedia model of semantic transcoding, capable of supporting different user's requests and constraints of their devices (PDA, smart phones, ...). Our application context is the emerging area of domotics (from the Latin word domus that means "home" and informatics) and, in particular, indoor video surveillance of the house where people with some difficulties (elders and disabled people) can now live in a sufficient degree of autonomy, thanks to the strong interaction with the new technologies that can be distributed in the house with affordable costs and high reliability.

The paper presents an approach for a robust (semi-)automatic correction of radial lens distortion in images and videos. This method, based on the Hough transform, has the characteristics to be applicable also on videos from unknown cameras that, consequently, can not be a priori calibrated. We approximated the lens distortion by considering only the lower-order term of the radial distortion. Thus, the method relies on the assumption that pure radial distortion transforms straight lines into curves. The computation of the best value of the distortion parameter is performed in a multi-resolution way. The method precision depends on the scale of the multi-resolution and on the Hough space's resolution. Experiments are provided for both outdoor, uncalibrated camera and an indoor, calibrated one. The stability of the value found in different frames of the same video demonstrates the reliability of the proposed method.

In this paper we propose an approach to indoor environment surveillance and, in particular, to people behaviour control in home automation context. The reference application is a silent and automatic control of the behaviour of people living alone in the house and specially conceived for people with limited autonomy (e.g., elders or disabled people). The aim is to detect dangerous events (such as a person falling down) and to react to these events by establishing a remote connection with low-performance clients, such as PDA (Personal Digital Assistant). To this aim, we propose an integrated server architecture, typically connected in intranet with network cameras, able to segment and track objects of interest; in the case of objects classified as people, the system must also evaluate the people posture and infer possible dangerous situations. Finally, the system is equipped with a specifically designed transcoding server to adapt the video content to PDA requirements (display area and bandwidth) and to the user's requests. The main issues of the proposal are a reliable real-time object detector and tracking module, a simple but effective posture classifier improved by a supervised learning phase, and an high performance transcoding inspired on MPEG-4 object-level standard, tailored to PDA. Results on different video sequences and performance analysis are discussed.

In this paper we address the problem of human posture classification, in particular focusing to an indoor surveillance application. The approach was initially inspired to a previous works of Haritaoglou et al. [6] that uses histogram projections to classify people’s posture. Projection histograms are here exploited as the main feature for the posture classification, but, differently from [6], we propose a supervised statistical learning phase to create probability maps adopted as posture templates. Moreover, camera calibration and homography is included to resolve prospective problems and improve the precision of classification. Furthermore, we make use of a finite state machineto detect dangerous situations as falls and to activate a suitable alarm generator. The system works on line on standard workstation with network cameras.

In this paper we address the problem of fast segmenting moving objects in video acquired by moving camera or more generally with a moving background. We present an approach based on a color segmentation followed by a region-merging on motion through Markov Random Fields (MRFs). The technique we propose is inspired to a work of Gelgon and Bouthemy [6], that has been modified to reduce computational cost in order to achieve a fast segmentation (about ten frame per second). To this aim a modified region matching algorithm (namely Partitioned Region Matching) and an innovative arc-based MRF optimization algorithmwith a suitable definition of the motion reliability are proposed. Results on both synthetic and real sequences are reported to confirm validity of our solution.