TU Delft Image Processing Platform

The Program

The Tensor Image Processing Platform (TIPP) is a software platform that is currently being developed at the faculty of aerospace engineering. The platform seeks to become a unified platform in which researchers will create plugins which can combine various research techniques and improve the conservation of written code. It is also meant to provide a usable and understandable interface for researchers that have less experience with programming so that they too are able to use powerful plugins and analysis methods created by their co-researchers.

Project Methodology

Development on this platform was started in 2015, with the following list of goals in mind:

: When using a piece of commercial software it is often unknown what methods the software uses exactly whilst performing analysis, this is due to the fact that the source code of commercially created software falls under copyright law and is thus not freely available. With TIPP the general idea is to always have the source code of the plugins fully available so that the user is always aware of the exact algorithms being performed.
: For the various research methods found throughout the NDT lab at the faculty of aerospace engineering there is not yet a unified way to simultaneously perform analysis of a single test sample within one program. TIPP aims to become this program where the end goal is full data fusion of multiple research methods.
: Since 2012 Python (2.7) is the standard programming language taught to all students at the faculty of aerospace engineering, however because a lot of the researchers still use Matlab there is a large gap between new students and researcher. With TIPP this gap will be narrowed by connecting the fresh students with experienced researchers that have spent years performing research in their respective fields and thus have a lot of knowledge to offer.
: A big problem with code written by a large group of researchers is that once they are done with their research the code is often shelved and lost, which is a bad thing because if a new researcher wants to continue the work they often need to read trough the paper and reconstruct the code from there. With TIPP the code can be conserved in the platform and the next researcher will instantly be at the level where the previous researcher left off. They can then take the processed data and perform more analysis on it using their own plugin, which will again be conserved within the platform.

Plugins

TIPP makes use of a plugin manager to manage all the plugins. These plugins can be installed by simply dragging and dropping them into the Plug-Ins folder.