Evaluating and improving the robustness of a video-based PR/RR monitoring system for a clinical environment

摘要: With the advancement and use of low cost cameras, video-based remote vital-monitoring technologies have drawn more attention. However, many previously proposed techniques rely on assumption that video is acquired under well-controlled conditions. There has been no study reported around how various acquisition factors and/or noise will affect estimation accuracy. In this paper, we describe a systematic approach in understanding challenges extending technology beyond environment share solution paths making robust one specific application. The involves first data analysis process, identifying key algorithms then estimating bounds metrics explain variations these vitals. As part paths, our paper goes into details were improved to address findings current state. refinement includes getting stronger signals less controlled setting.