Author Archives: admin

Joint laboratory MICC – Thales

MICC, Media Integration and Communication Center of the University of Florence, and Thales Italy have established a partnership to create a joint laboratory between university and company in order to research and develop innovative solutions per safety, sensitive sites, critical infrastructure and transport.

MICC - Thales joint lab demo at Thales Technoday 2011

In particular the technology program is mainly focused (but not limited) on surveillance through video analysis, employing computer vision and pattern recognition technologies.

A current active filed of research, continued from 2009 to 2011 was that of studying how to increase the effectiveness of classic video surveillance systems using active sensors (Pan Tilt Zoom cameras) and obtain higher resolution images of tracked targets.

MICC - Thales joint lab projects

The collaboration allowed to start studying the inherent complexities of PTZ camera setting and algorithms for target tracking and was focused on the study and verification of a set of basic video analysis functionalities.

Thales

In 2011 the joint lab led to two important demos at two main events: Festival della Creatività, October 2010 in Florence (Italy) and Thales Technoday 2011 in January 2011 in Paris (France). In the latter the PTZ Tracker has been nominated as VIP Demo (Very ImPortant Demo).

Some videos about this events:

Onna: a natural interface system for virtual reconstruction

Leave a reply

A technology transfer project for an international exhibition about the story of Onna, an italian town near to L’Aquila, which was affected by the earthquake during 2009.

A natural interaction based system is designed and developed in order to present a large number of multimedia contents (videos, images and audio) collected and created by the curators of the exhibition.

Interactive system scenario in the Infobox at Onna

The project involves the study and development of an interactive system which adopts the paradigm of the natural interaction in order to allow users to access and consult multimedia contents related to different areas of the town of Onna, an italian town near to L’Aquila, which was affected by the earthquake during 2009.

The concept proposed for the user-interface is inspired by the educational game book published in the seventies about the devastation of the town of Pompei after the eruption of Vesuvius in 79 B.C. The pages of the book are composed of images of the destroyed Pompei which can be overlapped with images of the town before the eruption.

Our idea is to recreate a similar mode of interaction and use a background picture of the town of Onna after the earthquake so that the user can interact with some areas of the image and see them as they were before the earthquake. In addition for each area is possible to visualize multimedia contents about history, architecture and life before the earthquake.

The user-interface will be optimized in the environment of the exhibition in order to allow multiple users to interact independently with the system.

MAC-GEO: the effects of geothermal power in Tuscany

1 Reply

A technology transfer project for the Regione Toscana in order to provide a solution to predict the effects of geothermal power both in the same basin and the surrounding environment in some areas of Tuscany.

Geotherm viewer data visualization

A web application in a GIS environment is designed and developed in order to visualize and query simulation data and their scenarios stored in a geospatial database.

Geotherm Viewer

Project URL: http://macgeo.math.unifi.it/~macgeo/

The MAC-GEO project aims to give the Regione Toscana some simulation tools to predict the effects of geothermal power both in the same basin and the surrounding environment.

The project was funded by Regione Toscana CIPE funds, following the announcement Framework Programme Agreement Research and technology transfer for the production system – Supplementary Agreement III.

The project started in September 2008 and lasts two years.

MICC research staff designed and developed the Geotherm Viewer, a web application in a GIS environment for viewing and querying simulation data and their scenarios stored in geospatial database.

The system was designed for users who need to quickly view the results of the generated simulations in order to assess which groups of data for later analysis and deepen the effects of geothermal both in the same basin and the surrounding environment. In addition, the user-interface provides a set of interactive features for geographic data visualization, data from web map services and information generated by Google Maps.

The Geotherm Viewer has been developed with open source technologies.

MICC demo at Technoday

Leave a reply

The MICC partecipated to the 2011 Technoday in Paris (France) presenting a demo about multi-target tracking with a PTZ camera developed in the joint lab with Thales Italy. The demo has been awarded as VIP (Very ImPortant).

The ninth TechnoDay event has provided an opportunity to view more than 90 demonstrations showcasing innovative and proactive capabilities in optronics, telecommunicationsand information systems for defence and security applications.

Innovation – which lies at the heart of Thales’s strategy – is placed squarely within the defence-security continuum for the very first time at TechnoDay 2011. The demonstrations clearly have illustrated potential synergies, despite the initiative only being in its infancy.

Having moved the event to a new, larger location, the Cité de l’Eau, has given the opportunity to highlight an even greater range of innovations, grouped around four key topics:

Resilient networks and information systems
Radio communications
Protection & detection
Battlegroup C4I

This ninth TechnoDay event has also provided an insight into how the focus has been mantained on innovation in new technologies in a more difficult business environment.

TechnoDay offered a unique, not-to-be-missed opportunity to discover latest technological advances and meet technical teams.

Technoday 2011

Technoday 2011 - MICC staff

Technoday 2011 - the location

Programming in pictures

Leave a reply

Programming in pictures (or filmification of methods) is an approach where pictures and moving pictures are used as (super-) characters to represent algorithms. Within this approach algorithms are considered as activities in 4-D space-time where some “data space” is traversed by a “front of computation” and necessary operations are performed during this traversal process. There are compound pictures to define algorithmic steps (called Algorithmic CyberFrames) and generic pictures to define the contents of compound pictures. Compound pictures are assembled into special series to represent predefined algorithmic features. A number of the series is assembled into an Algorithmic CyberFilm.

Programming in images

In this presentation, the concept and fundamental features of CyberFilm programming technology will be explained and examples of programs will be demonstrated. An idea of programs as media for human (programmers) communication will also be discussed.

Introducing Graph Words

Leave a reply

In the context of content based image retrieval, one of the most common approach nowadays is the Bag of Words (BoW) approach applied to local features such as textons, SIFT or SURF points etc. A dictionary is built by clustering the local features of a learning set of images. Then, for a test image each feature extracted is quantified according to the dictionary yielding a distribution of the features according to the dictionary. This method does not take into account any spatial relations between the features.

Delaunay diagrams

We introduce a semi-local feature approach called Delaunay graph feature. We use SURF points as nodes of a graph built be a Delaunay triangulation and then compute a dictionary graph words by a two pass hierarchic agglomerative clustering. The presentation will posit the premises and first experimental results of this work.

Enrich

Leave a reply

Enrich is a targeted project funded under the eContentPlus programme. Its objective is to provide seamless access to distributed digital representations of old documentary heritage from various european cultural institutions in order to create a shared virtual research environment especially for study of manuscripts, but also incunabula, rare old printed books, and other historical documents. It builds on the Manuscriptorium Digital Library (http://www.manuscriptorium.eu) that has already managed to aggregate data from 46 collections from the Czech Republic and abroad.

Manuscripts found by Enrich system

The project groups together almost 85% currently digitized manuscripts in the national libraries in Europe. These collections will be enhanced by substantial amount of data from university libraries and other types of institutions. The consortium will make available more than five million digitized pages.

Manuscriptorium is a result of 15 years of work and development carried jointly by two important czech institutions: AIP Beroun LTD and the National Library of the Czech Republic. It is the richest digital manuscript resource in Europe putting at disposal more than one million digitized pages, having a safe digital archive, enjoying of state support for digitization, and talking in czech and english languages. Ca. 50% of its users come from abroad and it operates a special clone for support of teaching and learning in secondary schools. Its origins are in the Memory of the World programme of UNESCO; therefore, the National Library of the Czech Republic received the UNESCO world Jikji award in 2005. The Manuscriptorium-related digitization knowledge and know-how have been taught and shared in many countries of the world.

Manuscriptorium builds on a robust xml schema the most important part of which is the european master format for electronic description of manuscripts based on TEI.

The basic corpus of data is made available from the digital storage facilities operated by AIP Beroun LTD in the Czech Republic. The Enrich project will integrate even more data from remote foreign digital libraries. The metadata records for the central database will be preferably collected via the OAI protocol; they will contain links to images stored in remote image databanks. Necessary transformation routines will be created and tuned for each partner. Specialized on-line tools will be also developed to enable Manuscriptorium schema compatible metadata structuring and output validation for those partners that have digital data with no presentation tools and will like to make them available.

Enrich target user groups are content owners/holders, libraries, museums and archives, researchers and students, policy makers, and general interest users. The project will allow them to search and access documents which would otherwise be hardly accessible. Besides images, it will be also able to offer access to TEI-structured historical full texts, research resources, other types of illustrative data (audio and video files) or large images of historical maps. The Enrich consortium will closely cooperate with TEL (The European Library) and will become a component part of the European Digital Library when this becomes reality.

The users will get tools that will assist them to create their own documents and personal digital libraries in Manuscriptorium using any analytical objects of which the documents consist. Tools for usage of more languages will be offered for application in Manuscriptorium as well as multilingual ontologies enabling search in local languages and retrieval of data in source languages.

The Enrich consortium consists of 18 partners and the project is also supported by a number of other institutions among which there are many important content owners.

The project is coordinated by the National Library of the Czech Republic together with other two Czech partners: AIP Beroun LTD and Crossczech Prague Inc.

The interest to cooperate has been expressed, among others, also by the national libraries of Hungary, Kazakhstan, Moldova, Poland, Romania, Serbia, Sweden, and Turkey as well as by the university libraries in Bratislava, Bucharest and Heidelberg. The list of associated partners will grow during the project duration.

DanThe. Digital and Tuscan heritage

Leave a reply

DanThe is an online service, developed and implemented by Tuscany Region and MICC – Media Integration and Communication Center – University of Florence, to promote the resources related to digital cultural heritage of Tuscany. DanThe provides a direct access to collections, databases, regional museums, libraries and catalogues of cultural heritage.

Danthe. Digital and Tuscan Heritage

The “catalogue of collections” provides a selection of the Tuscan digital documents (texts, images, video, audio) concerning cultural heritage (artistic, historical, archaeological, ethno-anthropological, archival and bibliographical). These collections were previously catalogued by the Michael Project (Multilingual Inventory of Cultural Heritage in Europe). Within the universe of European digital resources Danthe identifies a Tuscan subset of collections and make it accessible to different users.

The “catalogue of cultural heritage” is designed to collect and make available descriptive and informative catalogue records – in accordance with local owners (municipalities, museums, institutions etc.) – and realized using the standards ICCD and the CART free software, produced by the Tuscany Region and carried out by the Scuola Normale Superiore in Pisa.

In accordance with the guidelines of Michael Project, Danthe selected digital collections regarding people, works and cultural heritage of Tuscany, promoted, financed or produced by Tuscan institutions and regional offices of national or foreign institutions. The collections presented in Danthe, as those ones published on the European portal of Michael, are available online and/or offline, are free, or available on the registration fee and may be completed, planned or in progress.

In future the system will enable an innovative technological regional application to provide services that manage digital contents regarding cultural heritage and it will be able for the integration with the Regional Cooperation Application of Tuscany – CART

In this perspective, the Regional Information System will be able to:

hold copies of data on cultural heritage (regional databases and third party) and their metadata in a centralized repository;
allow automatic updating of data through the application infrastructure for cooperation;
interoperate with Information systems of external structures (Michael, CulturItalia, Europeana) using appropriate protocols and server (OAI-PMH);
allow the use of such digital contents by the web interface unified Danthe.

Danthe. Digital and Tuscan heritage

DanThe is part of a national strategy (portal CulturItalia) and (Europeana) which aims to combine and enhance digital assets in Europe, to consolidate descriptive standards and to set minimum standards in the development of resources of various nature and non-traditional format, generated using innovative technologies and for heterogeneous users.

Because of the deep analysis of the existent carried on digital resources relating to cultural heritage, the European project Michael was the basis to fill and populate DanThe database. The close cooperation between national European project of Michael, the regional cataloguers and the editors of DanThe allowed to use the expertise and knowledge of the local census of digital collections and to personalize the presentation and the fruition of the digital resources themselves.

Even though DanThe borrowed from Michael all the entities structured and divided according to the descriptive standard of the project, the Tuscany recourses are presented in a singular and personal way. In particular, it were not censed catalogues of libraries when they gave no services other than paper, to consider as digital collections not only the resources derived from analogical source (books, art works and objects of art, etc..) but also those formed from digital material (videos, Word documents, audio files, etc..) and to give a direct access to the collections online providing, where possible, the specific URL of the resource rather than to the generic website or the catalogue containing it.

PASCAL VOC 2010: semantic object segmentation and action recognition in still images

Leave a reply

Last year I presented the work we at the Computer Vision Center (CVC) did on semantic image segmentation in context of the PASCAL 2009 Visual Object Classification (VOC) challenge. This year, we again fielded teams in several competitions of the PASCAL 2010 VOC challenge.

In this talk, I will discuss the extensions we have made to our approach to semantic image segmentation. I will show how the results of object detectors and spatial priors can be naturally integrated into our hierarchical conditional random field (HCRF) approach based on the harmony potential. The addition of these extra cues, as well as class-specific normalization of classifier outputs, significantly improves segmentation quality.

Semantic segmentation in still images

I will also discuss our approach to human action recognition in still images. Action recognition from still images is a new, “taster” competition in this year’s VOC competition. It requires participants to identify the action being performed in individual images and the task is further complicated by the lack of large quantities of training data. Our approach is based on a spatial pyramids over a classical
bag-of-visual-words approach with extensive, class-specific cross validation used for feature selection.

Human action recognition in still images

Our results on semantic object class segmentation show that our approach obtains state-of-the-art results on three challenging datasets: PASCAL VOC 2009, PASCAL VOC 2010 and MSRC-21. In the PASCAL 2010 challenge, our approach won eleven gold medals, taking first place in the segmentation challenge. In action classification, our approach won three gold medals and jointly won the first place award along with INRIA LEAR and University of Surrey.

PASCAL VOC 2010: semantic object segmentation and action recognition in still images

PointAt system at Palazzo Medici Riccardi

Leave a reply

Palazzo Medici Riccardi is one of the most important museums in Florence: in its small chapel, it hosts the famous fresco “La cavalcata dei magi” (“The Journey of the Magi”) by Benozzo Gozzoli (1421–1497).

The PointAt system’s goal is to stimulate the visitors to interact with a digital version of the fresco and, at the same time, make them interact in the same way they will in the chapel, reinforcing their real experience with the fresco. That is to use information technology to make teaching attractive and effective.

PointAt at Palazzo Medici Riccardi

Visitors are invited to stand in front of the screens and indicate with their hand the part of the painting that interests them. Two digital cameras analyse the visitors’ pointing action and a computer vision algorithm calculates the screen location where they’re pointing. The system then provides audio information about the subject.

In designing the system, we considered the following issues:

Easy and simple interaction. Visitors don’t need any instruction or have to wear any special device.
High-resolution display. The fresco is displayed on large screens so that visitors can appreciate even small particulars (almost invisible in the real chapel).
Interactivity for different categories of visitors. Interaction should be satisfactory for visitors who just want an idea about the fresco, for those who are attracted by particular characters and for those who want to have complete information on the whole fresco.
Not intrusive setting. The physical setting must host both active and passive visitors (for example, the relatives of the person who’s actually interacting with the system and those interested in listening but not in being active).
Pleasant look & feel. The interactive environment is integrated within the museum and it respects the visitors’ whole experience.

PointAt is considered to be a good vanguard experiment in the field of museum didactics, and has been functioning successfully since 2004.