The ability to quickly screen people for symptoms of infectious disease is important to reduce disease transmission in long term care facilities and crowded public spaces. To achieve this goal, one option could be non-contact sensor arrays that measure vital signs. In this work, we present initial results for the assessment of heart rate through video magnification techniques applied to visible light Red/Green/Blue (RGB) video recordings of the face. The work specifically explores the effect of region of interest size on the accuracy of heart rate measurements. The visible skin on a person’s face can be obscured by hair or face masks, leading to a need for algorithms and methods that can compensate for these effects. The results show the potential for the combination of many small regions as an alternative to a single large region. The best performance for small regions is a mean absolute error of 9.9% while larger regions performed better with the best error performance of <3%. The work also shows that for larger regions covering most/all of the face, body motion reduces the performance more than the use of a face mask that obscures a portion of the face. This work provides the foundation for further development of robust non-contact health screening solutions.