TRAPPER - ELIMINATING PERFORMANCE BOTTLENECKS IN A PARALLEL EMBEDDED APPLICATION

The Trapper software-engineering environment for parallel and distribu ted systems comprises tools for software design, hardware configuratio n, mapping, monitoring, visualization, and performance tuning of paral lel applications and systems. This article's authors used Trapper as t he main software-development tool to realize a complex industrial rese arch application within the Prometheus collision-avoidance project at Daimler-Benz. In this project, they outfitted a Mercedes 500 SEL resea rch vehicle with 18 cameras and approximately 60 computing nodes for v arious image processing and control algorithms such as lane following, Obstacle detection, and traffic-sign recognition. This article seeks to analyze the behavior and improve the performance of the most time-c onsuming task of the whole software traffic-sign recognition. The auth ors show how Trapper effectively detected and eliminated performance b ottlenecks of this parallel application.