Home

The amount of digital data on the internet has been growing exponentially over the last decade. Information freely available to users include text, graphical data etc. Search engines make it easier to search and retrieve relevant information. These engines take a textual query as input to find the relevant document on the web corresponding to the occurrence of the words within the query...

Recent development of sophisticated technologies for creating and acquiring 3D data, have resulted in the increasing amount of 3D graphical data available on the web as well as in corporate repositories. This facilitates the process of designing new 3D models by providing the possibility to reuse available data.
However while a textual query appears to be a natural description to search for documents it proves to be very efficient when retrieving 3D information. Files containing 3D models cannot be found in a standard manner, and using textual information related to them, including file names, meta-information, is not sufficient.
The problem of shape retrieval becomes ever more challenging due to the increasing number of 3D models available in Internet. As a consequence the need of methods capable to define and describe a shape arises. Shape descriptors should efficiently describe a shape but at the same time should be compact enough to allow efficient retrieval. A good solution could be a descriptor, which could decompose a model into components corresponding to meaningful parts of the shape, extract the most distinctive features and merge them into one structure.

This structure can be then used to describe the connectivity relations between the features represented by geometrical or morphological characteristics. However, it can be difficult to describe the shape of all components in a unique way. As a consequence, Graphitech has developed a number of different software tools whose aim is to define the semantics of a shape and facilitate its retrieval.
This approach provides a higher, semantic-level solution to image retrieval problems, as opposed to existing traditional methods and techniques dealing with low-level image retrieval issues. Techniques and algorithms from CAD and pattern recognition have been tightly integrated to produce a flexible, accurate and fast image retrieval method using sketch and spatial information. This research activity reflects and deals with various low level aspects such as:

• Shape alignment.
• Shape segmentation.
• Shape matching.
• Shape descriptors.

These research activities have demonstrated how the use of PDM systems, coupled with semantic Web technologies, can improve inter-working activities and interdisciplinary knowledge sharing. Perhaps the most relevant of all projects in this field is AIM@SHAPE, a Network of Excellence on “Advanced and Innovative Models and Tools for the development of Semantic-based systems for Handling, Acquiring, and processing knowledge Embedded in multidimensional digital objects”.

This project has brought to the development of a number of shape descriptors together with a number of software prototypes performing intelligent search retrieval, based on shape semantics. Developments were integrated within the so-called Digital Shape Workbench (DSW) a set of tools and services for modelling, processing and interpreting digital shapes.