Knowledge representation scheme for computer vision systems
Main researcher
: RIBARIĆ, SLOBODAN (112300) Assistants
SUŽNJEVIĆ, VLADO (13105)
LUKŠA, DARKO (158913)
MARTINČIĆ, KRUNOSLAV (163073)
MATIĆ, NINOSLAV (169735)
BUDIN, ANDREA (190582)
KALAFATIĆ, ZORAN (190593)
PERKOVIĆ, ŽELJKA (190602)
STANISAVLJEVIĆ, VLADIMIR (190714)
Type of research: basic Duration from: 02/01/91. to 02/01/94. Papers on project (total): 77
Papers on project quoted in Current Contents: 3
Institution name: Fakultet elektrotehnike i računarstva, Zagreb (36) Department/Institute: Department for Electronics, Microelectronics, Computer and Intelligent Systems Address: Avenija Vukovar 39 City: 10000 - Zagreb, Croatia
Communication
Phone: 385 (0)1 6129 935
Fax: 385 (0)1 6129 653
E-mail: Slobodan.Ribaric@fer.hr
Summary: The goal of the project is the developement and the
verification of an integral knowledge representation scheme for declarative
and procedural knowledge in intelligent information (computer) systems. The
achieved research results will serve as a theoretical and practical base
for designing complex systems with visual feedback in intelligent
industrial and robot systems. The KRP knowledge representation scheme
based on Petri Net Theory has been developed. The scheme is used for
realization of the semantic level and the rule-based level in the
hierarchical knowledge base for robot vision systems. The scheme, which
represents declarative and procedural knowledge within a unique framework,
enables dynamical interaction between them and gives appropriate mechanisms
for mantaining consistent deductions in a relatively predictable working
environment. The reasoning procedures (inheritance and recognition) have
been developed and tested.
Research goals: The goal of the project is the development and the
verification of an integral knowledge representation scheme for declarative
and procedural knowledge in intelligent information (computer) systems.
The KRP knowledge representation scheme based on Petri Net Theory has been
developed. The scheme is used for realization of the semantic level and the
rule-based level in the hierarchical knowledge base for robot vision
systems. The scheme, which represents declarative and procedural knowledge
within a unique framework, enables dynamical interaction between them and
gives appropriate mechanisms for maintaining consistent deductions in a
relatively predictable working environment. The reasoning procedures
(inheritance and recognition) have been developed and tested. The achieved
research results are theoretical and practical base for designing complex
systems with visual feedback in intelligent industrial and robot vision
systems. A prototype of an integral system which includes visual subsystem
and robot arm is developed and tested. The knowledge representation scheme
KRP is used as a base of this integration. Other information about the project.