Analysis and implementation of image-based robust heart frequency monitoring algorithms in automotive environment

OData support
Supervisor:
Dr. Kiss Bálint
Department of Control Engineering and Information Technology

In my thesis I describe the main types of the contact and non-contact heart rate monitoring methods, slightly with more attention to the non-contact monitoring methods. I chose an image-based method for monitoring heart rate (EVM) that I explain in more details. This algorithm is developed by the Massachusetts Institute of Technology. I evaluate the robustness of this algorithm against the varying lighting conditions and the movements of the test person. During my work I used several sequences from three cameras, one color (Kinect) and two infrared (Tobii and PointGrey FlyCapture2) cameras.

Hereinafter I examined how much the tracking of the visual features will improve the robustness of the algorithm against the previously mentioned influences. Meanwhile I used many image- and signal processing algorithms, wherewith I was able to eliminate the high frequency components from the different signals and I could also stabilize a specified visual area on a video. At the end of my thesis I examine the future works and the further improvement potential.

Downloads

Please sign in to download the files of this thesis.