Giancarlo Iannizzotto

Fast and accurate edge-based segmentation with no contour smoothing in 2-D real images

In this paper we propose an edge-based segmentation algorithm built on a new type of active contour which is fast, has a low computational complexity and does not introduce unwanted smoothing on the retrieved contours. The contours are always returned as closed chains of points, resulting in a very useful base for subsequent shape representation techniques.

Quantitative Analysis of Pursuit Ocular Movements in Parkinson’s Disease by Using a Video-Based Eye Tracking System

The purpose of this study is to assess the efficacy and the tolerability of a new vision-based non-intrusive eye tracker in a population composed of normal controls and in patients affected by nonadvanced Parkinson’s disease (PD). PD patients characteristically have difficulty in sustaining repetitive motor actions. Previous studies showed a progressive bradykinesia and hypokinesia of pursuit ocular movements (POM) in advanced PD. We found that the values of POM were lower in PD patients than in normal controls (p < 0.001). In PD patients, the values correlated closely with Hoehn and Yahr stage and Unified Parkinson Disease Rating Scale motor subscore (p < 0.001, for both). Our data suggest that deficit in POM occurs also in nonadvanced PD patients and it is closely correlated with clinical scores. Thus, this vision-based system can be considered a new method to provide, noninvasively, measures of POM dysfunctions and can be used as reliable indices of disease severity in PD patients.

Badge3D for Visually Impaired

Conducting an independent life is probably the most important issue for visually impaired people. In this paper, we suggest a contribution to the solution of this problem using wearable computer technology. We present a visual support system that provides acoustic information about the objects in the surrounding environment, obtained by remotely reading barcode tags sticked on significant objects and surrounding elements, such as doors, windows and so on. The user, walking in an indoor environment, is informed in realtime about the location (direction, distance and pose) of the available objects. Barcode tags deployed in the environment can act as reliable stimuli that trigger local navigation behaviours to achieve global navigation objectives. The proposed system is expected to be useful in the real-time interaction with dynamic environments. To illustrate our work, we introduce a proof-of-concept multimodal, sensorbased application and discuss its implementation and the obtained experimental results.

Simulative assessments of IEEE 802.1 Ethernet AVB and Time-Triggered Ethernet for Advanced Driver Assistance Systems and in-car infotainment

Investigations into the usage of Ethernet in automobiles is in progress in academia, the car industry and companies producing automotive electronic devices. The interest in Ethernet is motivated by the high bandwidth and scalability provided. It is a well experienced technology with support for the Internet Protocol (IP) suite. Ethernet as in-car network is expected to breakthrough in Advanced Driver Assistance Systems (ADAS) involving cameras and in the multimedia domain. Both the IEEE Audio Video Bridging (AVB) standard and Time-Triggered Ethernet (TTE) are promising candidates. This paper presents a simulation study aimed to investigate the behavior of these technologies when supporting ADAS and multimedia traffic on star-based networks under varying workload. The performance under different operating conditions is presented and discussed.

A vision-based system for elderly patients monitoring

Remote patient monitoring can improve the quality of life of elderly and impaired people, while reducing the costs. Among the most interesting technologies being investigated, computer vision has proved to be very effective in several important scenarios in which conventional sensors fail or are impractical. We propose a computer vision-based wireless sensor system for people remote tracking and monitoring based on low-cost embedded systems able to visually track the patient and detect critical motion and posture patterns, associated with dangerous situations. Motivation for the work and experimental results are provided, showing the effectiveness and the validity of the presented approach.

Towards IEEE 802.1 Ethernet AVB for Advanced Driver Assistance Systems: A preliminary assessment

Advanced Driver Assistance Systems involving cameras and the multimedia domain are both quite challenging for automotive network technologies, due to the bandwidth requirements and latency/jitter constraints they impose. Ethernet as in-car network is expected to breakthrough in both domains and the IEEE Audio Video Bridging (AVB) standard is a promising candidate. This paper reports preliminary performance assessments of AVB for ADAS, multimedia and infotainment, obtained through the simulation of realistic traffic patterns.

Competitive Combination of Multiple Eye Detection and Tracking Techniques

Reliable detection and tracking of eyes is an important requirement for attentive user interfaces. In this paper, we present an innovative approach to the problem of eye tracking. A number of traditional eye detectors, chosen for their own properties, are combined by two different competitive schemes with the aim to obtain a higher degree of robustness and reliability. To illustrate our work and validate our competitive combination approach, we introduce a proof-of-concept single-camera remote eye tracker and discuss its implementation and the produced experimental results.

A multimodal perceptual user interface for video-surveillance environments

In this paper a perceptual user interface (PUI) for video-surveillance environments is introduced. This system provides a tool for a video-surveillance control-room, and exploits a novel multimodal user interaction paradigm based on hand gesture and perceptual user interfaces. The proposed system, being simple and intuitive, is expected to be useful in the control of large and dynamic environments. To illustrate our work, we introduce a proof-of concept multimodal, bare-hand gesture-based application and discuss its implementation and the obtained experimental results.

A fast, accurate method to segment and retrieve object contours in real images

We have developed a fast method to segment 2-D grey-level images, obtaining closed-line, continuous, smooth and analytical representations of object contours. The results are very good, even in presence of noisy or low contrast images, and the shape of objects are retrieved with very good accuracy.

VirtualBoard: real-time visual gesture recognition for natural human-computer interaction

The paper proposes a simple, economical and efficient real-time, visual hand gesture recognition system for human-computer interaction. The system is designed to work in critical lighting conditions, for example when an image is projected onto a wall (conferences, offices, etc). Given the low level of computational complexity, the system can be used instead of a mouse. The extreme naturalness of the interaction makes previous training unnecessary and the robustness of the algorithms used guarantees a high level of reliability. The architecture of the system is described, and experimental results obtained in a large number of tests with different, untrained users are presented and commented.