Inhaltszusammenfassung

This paper presents a framework for object detection of different types of data. This framework is based on semantic tech- nologies as an ontology which gives a structure to store and organize the knowledge for object detection. This framework is com- posed of four modules which provide an automatic and adaptive object detection. Each module provides a tool and all these tools are combined to determine and apply a detection method depending on the data type, the scene context and the targeted...This paper presents a framework for object detection of different types of data. This framework is based on semantic tech- nologies as an ontology which gives a structure to store and organize the knowledge for object detection. This framework is com- posed of four modules which provide an automatic and adaptive object detection. Each module provides a tool and all these tools are combined to determine and apply a detection method depending on the data type, the scene context and the targeted object(s). The knowledge module and the reasoning module allow the determination of atomic algorithms which are necessary for the detec- tion execution. Another module executes these algorithms and another aim is to learn in order to improve the object detection. This framework aims to provide a flexible and accurate tool for object detection in 2D as in 3D. » weiterlesen» einklappen

Autoren

Klassifikation

DDC Sachgruppe:
Informatik