1 / 23

GNORASI WORKSHOP Project overview

Ανάπτυξη ροών συνδυασμού γνώσης και αλγορίθμων επεξεργασίας σε εφαρμογές οργάνωσης και διαχείρισης δεδομένων τηλεπισκόπισης Knowledge and processing algorithms for remote sensing data management. GNORASI WORKSHOP Project overview. Tsampoulatidis Ioannis CERTH / ITI. What lead us to GNORASI.

marnin
Download Presentation

GNORASI WORKSHOP Project overview

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Ανάπτυξη ροών συνδυασμού γνώσης και αλγορίθμων επεξεργασίας σε εφαρμογές οργάνωσης και διαχείρισης δεδομένων τηλεπισκόπισης Knowledge and processing algorithms for remote sensing data management GNORASI WORKSHOPProject overview TsampoulatidisIoannis CERTH / ITI

  2. What lead us to GNORASI • Yet Another Remote Sensing Application • Existing approaches tend to materialize customized solutions in correspondence with the requirements and specifications posed by individual application scenarios • Poor extensibility and adaptability properties • Serious limitations in terms of modularity • Sharing and reuse of the accomplished results is severely hindered, necessitating significant re-engineering in order to address the different requirements featured in a new application context. • Significant cost in terms of time and resources in order to find the combination of algorithms and methodologies that successfully meets the specifications at hand, before proceeding to its implementation • GNORASI addresses the formalization of the semantic interpretation of remote sensing data through a modular, extensible system that enables the effective combination of • Knowledge • Reasoning • Imaging processing algorithms

  3. GNORASI goals • An integrated, modular framework for the seamless and effective development of remote sensing data interpretation services, • by enabling the expert user to select each time the appropriate image processing algorithms • to define the axiomatic knowledge that underlies the given problem • Investigation and research into novel algorithms for remote sensing data analysis • Research in knowledge modeling and inferencing methodologies, in order to further alleviate limitations challenging the state of the art • Mechanism to embed knowledge and thus allow the semantic interpretation of remote-sensing data • Development of a software platform to manage remote-sensing data • Maximum freedom • Multi-level adaptation according to user experience • Support multimodality and cover many domains • 1stpriority: Expandability, Reusability (DRY)

  4. GNORASI added value • GNORASI provides a generic and powerful framework • Enables the expert user to effectively put into practice its scientific expertise in an graphical, intuitive fashion • Either through its interaction with the already existing knowledge repository, i.e. the pool of already introduced analysis algorithms and knowledge structures • Or through the incorporation of new ones • Through the gradual use of GNORASI platform • Experts’ knowledge can be accumulated and made available to new users… • providing thereby a continuously growing knowledge repository appropriately equipped to address a wide range of remote sensing applications

  5. How to achieve maximum freedom • Avoid monolithic design • Involve as many people as possible • Build a robust technical infrastructure and strong documentation • Create large repository of modules ready to be used out-of-the-box • Provide solutions covering multiple domains • Make the project open-source • GNORASI is Free Software licensed under the GPLv2 • Free software IS NOT THE OPPOSITE ofCommercial software • Services • Training • Installation • Maintenance • Custom solutions / new algorithms

  6. Target Group • GNORASI results can be applied on • Agricultural planning • Environmental conservation • Hazards management • natural disasters • pollution incidents • rural development • Decision support systems • Land usage and conservation • Weather forecasts based on the monitoring of meteorological parameters • Research on climate changes • Public sector and specialized “environmental-oriented” private companies • Wherever satellite and aerial images could be used to extract knowledge • …and by remote sensing experts and researchers to build and test their algorithms, visually.

  7. Use Cases • GNORASI is used by end-users like Regional Authorities and companies providing Urban and Environmental Service to help recognize and quantify information like: • Burnt areas • Flooded areas • Urban green areas • Agricultural crop types • Illegal building in areas with residential restrictions • Leakage / contamination in aquatic systems (e.g. oil spills) • Equipment (e.g. high voltage posts, wind turbines, photo-voltaic parks, etc.)

  8. GNORASI Users 1 3 2 • End users • See only results • Direct use • Import / export • Use defaultworkflows / use case • Experienced users • Remote-sensing experts • Non programmers • Need to compose new algorithms or edit existing ones • Power users • Programmers • Community to implement new modules • Services • Maximumfreedom

  9. Design

  10. Data-flow paradigm

  11. Data-flow network in GNORASI • System Analysis Design – Data Flow Diagrams • Close to Unix Pipelining | • Successive actions one after the other • Graph Theory • Smarter than simple pipe-lining • Parallel processing • Connection validation • Multiple Input / Output (Ν:Ν relation) • Ability to support heterogeneous inputs on the same processor • Multi-instanceper processor (theoretically infinite number of same processors in singe data-flow) • Each other with different parameters

  12. Data-flow network in GNORASI • Data-flow networks consist of modular units, called processors, which encapsulate rendering and data processing algorithms. • A processor operates on input data it receives from its inportsand outputs the processed results via its outports. • The data-flow is established by unidirectional connections from outports to inports. • Processor ports are typed, i.e., each port transmits a certain type of data, such as Image, Vector, LabelMap or a collection of objects of these types. • Processors furthermore have properties for the parameterization of the encapsulated algorithms. • and widgets (Qt-based complex graphical windows, i.e. GNORASI Viewer) • The strength of the data-flow concept lies in the flexible combinability of components. • Since a processor's behaviour is solely determined by its input data and property configuration, processors can be arbitrarily joined into networks • Only restriction: connected ports need to be of equal type. • No distinction made between standard processors provided by default in GNORASI and custom processors created by external developers.

  13. Graphical representation of data-flow networks

  14. Graphical representation of data-flow networks Property widgets are automatically generated for each property in a network.

  15. Why we need data-flow networks

  16. Why we need visual data-flow networks? • User friendly – No need to be a programmer to create new algorithms • Remote sensing experts • Advances experimentation • Easy to apply their scientific expertise • Rapid development and direct results • Possibility to overview intermediate results • Easier to spot weaknesses in algorithms • Isolate part of the algorithm’s functionality • Debugging on data-flow level • Easy adaptation • Reduces • Resources • Time

  17. System Architecture

  18. Data-flow network UNIT • Data flow network unit is where users interact with the system. GUI to create network and interpretation of the flow takes place in this unit. • The unit composed of the: • Validation and compatibility control sub-unit • Knowledge connector sub-unit

  19. Data-flow interpretation UNIT • Backbone of GNORASI platform. • The mechanism behind the scene where graph theory is applied to rank and process each individual processor • Data-flow execution and validation sub-unitactually interpreters the network after the internal indexing that takes place in the data-flow network indexing sub-unit • Export and visualization sub-unit exports the results to be displayed on screen

  20. Knowledge Management Unit • Knowledge management unit handles the knowledge repository where semantic rules, ontologies and appropriate mechanisms to interact with external reasoners are relied. • All modules and processors are stored in that unit

  21. Interconnection with 3rd party systems • GNORASI platform supports web-services, wrapping their functionality in a GNORASI-compatible manner. • The advantage of WS is the use of data which are stored and processed remotely.

  22. Facts and numbers • During the last 3 years 30 people where actively involved in GNORASI (management, development, support) • 23 Deliverables • 779 commits and counting • Presented in 3 conferences • GNORASI is included in the external projects of OTB (OrfeoToolBox) by the Centre national d'étudesspatiales (CNES) (English: National Centre for Space Studies), the French government space agency • Open source community starts to involve dynamically in the implementation and further support of GNORASI • We believe it the potentiality of GNORASI platform

  23. Spread the word !!!

More Related