Implementation of an Electronic Monitoring Device Using Computer Vision for Non-Invasive Respiratory Rate Measurement

An electronic monitoring circuit has been developed to detect a person's respiratory rate without requiring medical professionals or physical contact, thereby improving patient comfort. The Eulerian Video Magnification method and the Optical Flow technique were applied in computer vision to obtain, quantify and analyze respiratory movements using videos in an efficient manner that standard sensors might misread. The circuit can perform constant monitoring at low cost using a Raspberry pi camera OV5647 which has an intuitive interface to be used for the analysis. Comparative tests were conducted between the proposed device and a traditional thermistor-based nasal sensor. The findings showed that the non-invasive device achieved a high precision rate, closely matching the accuracy of the traditional method, with a margin of error below 1.5% in well-lit environments. The results indicate that computer vision can be implemented to develop efficient, low-cost medical monitoring solutions for clinical and homecare environments.