Non-human primates (NHPs) are critical in neuroscience research, where researchers implant electrodes into their brains to observe neuronal activity. However, the high post-operative mortality rate of NHPs remains a significant concern, primarily due to inadequate post-surgical monitoring. Thus, it is essential to develop a monitoring system for assessing their physiological status. Conventional wearable monitoring devices are often removed by NHPs, making it challenging to to monitor their well-being. To address this issue, we propose a video-based monitoring system that utilizes Eulerian video magnification techniques and principal component analysis to detect and estimate the respiratory rate remotely. The results show that our method has a difference of less than 1 breathing rate per minute compared to a commercial monitor and works well in experiments involving different species, demonstrating the accuracy and effectiveness of our method. We found that Gaussian pyramid representation has lower requirements for the ROI selection, making it more robust for practical applications. The proposed system has the potential to improve animal welfare and facilitate new experiments in neuroscience research.

Respiratory rate estimation,Eulerian motion magnification,Non-contact measurement,Non-human Primates