Novel Fingertip Image-Based Heart Rate Detection Methods for a Smartphone

Figure 1. Our developed smartphone-based heart rate detection application: (a) fingertip placement; and (b) graphic user interface (GUI) of our developed application.

Figure 2. Procedure of the fingertip curve line movement-based detection method.

Figure 3. An example of the edge detection procedure on a fingertip image obtained from a smartphone video camera: (a) greyscale image in YUV raw image; (b) preprocessed image; and (c) the edge detected image.

Figure 4. Flow chart of our proposed edge detection procedure.

Figure 5. An example of the smoothing procedure on a fingertip image obtained from a smartphone video camera: (a) after distance transform; (b) after diffusion; and (c) after binarization.

Figure 6. An example of the fingertip curved region detection procedure on a fingertip image obtained from a smartphone video camera: (a) non-trivial three regions (white) are found in an inverting smoothed image and (b) one representative region having the largest davg is found.

Figure 7. Procedure of the fingertip image intensity-based detection method.

Figure 8. An example of fingertip image intensity-based detection results obtained from a smartphone video camera.

Figure 9. Estimated heart rhythms of the proposed fingertip curve line movement-based algorithm and the conventional average intensity-based algorithm.

Figure 10. An example of the correlation between heart rhythms obtained from the proposed fingertip curve line movement-based and the conventional heart rate detection methods for subject 2. Here, the correlation line is denoted by the red solid line and the correlation index is 0.9590.

Figure 11. An example of the heart rate estimation of the proposed fingertip curve line movement-based and the conventional heart rate detection methods (frequency-domain estimation).

Figure 12. Heart rhythms obtained by the proposed (median-based) and conventional methods: (a) heart rhythms from the conventional RGB band-based method and the proposed method; and (b) peak detection results from the conventional red band-based method and the proposed method.