Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

VIP - A framework-based approach to robot vision

: Utz, H.; Kaufmann, U.; Mayer, G.; Kraetzschmar, G.K.

International Journal of Advanced Robotics Systems 3 (2006), No.1, pp.067-072
ISSN: 1729-8806
ISSN: 1729-8814
Journal Article
Fraunhofer IAIS ()

For robot perception, video cameras are very valuable sensors, but the computer vision methods applied to extract information from camera images are usually computationally expensive. Integrating computer vision methods into a robot control architecture requires to balance exploitation of camera images with the need to preserve reactivity and robustness. We claim that better software support is needed in order to facilitate and simplify the application of computer vision and image processing methods on autonomous mobile robots. In particular, such support must address a simplified specification of image processing architectures, control and synchronization issues of image processing steps, and the integration of the image processing machinery into the overall robot control architecture. This paper introduces the video image processing (VIP) framework, a software framework for multithreaded control flow modeling in robot vision.