The performance of an image processing system is significantly impacted by the quality of the images, which, in turn, depends upon the vehicle environment, i.e. the scene, and upon the imaging process. Variations of the image processing performance can only be understood in the light of these factors. Without their knowledge, neither an optimization nor a quantitative evaluation of the image processing algorithms can reasonably be done. The objective of the project documented here was to derive a quantitative description of image quality as a function of scene parameters and of imaging errors. Scene parameter estimation as an analysis tool for the performance of video-based environment sensing for road vehicles has been implemented. The dimensions of the scene parameter space are light intensity, weather conditions, and light distribution. In addition to the scene parameters, image errors (blooming, synchronization errors and noise) are recognized. Scene parameter estimation is a key tool also for the development of a test strategy in view of a safety validation of Advanced Driver Assistance Systems.
Scene parameter estimation for the performance analysis of image processing systems
Szene-Parameter-Abschätzung zur Betriebsanalyse eines Bildverarbeitungssystems
2008
5 Seiten, 9 Bilder, 3 Tabellen, 2 Quellen
Conference paper
English
CRUISE SCENE LEARNING DEVICE, CRUISE SCENE ESTIMATION DEVICE, AND CRUISE SCENE ESTIMATION SYSTEM
European Patent Office | 2019
|