Event-based image processing represents an emerging research field within computer vision and is gaining increasing relevance in modern imaging technologies. In contrast to conventional frame-based approaches, neuromorphic image sensors record only positive and negative changes in light intensity at the pixel level. This enables data acquisition with temporal resolution in the microsecond range, thereby opening up novel approaches for the real-time capture and analysis of highly dynamic scenes. This Paper investigates event-based imaging (EBI) as a novel sensing technology for high-precision object tracking in a high-speed cable-driven robot. First, a fundamental overview of the operating principles and characteristic properties of event-based image sensors is provided, and their potential for capturing highly dynamic scenes is discussed. Subsequently, a systematic comparison with conventional frame-based image processing methods is conducted, particularly with respect to temporal resolution, latency, and robustness under real operating conditions. Building on these considerations, the development and implementation of an event-based sensor system for real-time position estimation in the cable-driven robot are presented. The performance of the system is analyzed and evaluated through experimental investigations, with a focus on rapid object detection and the precise capture of dynamic motion patterns.
Karic et al. (Thu,) studied this question.